| History log of /src/sys/netinet |
| Revision | Date | Author | Comments |
| 1.32 | 28-Oct-2022 |
ozaki-r | Remove in_pcb_hdr.h
|
| 1.31 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.30 | 06-Sep-2018 |
maxv | Remove the network ATM code.
|
| 1.29 | 11-Jul-2018 |
kre | Fix build. pf_ioctl.c needs netinet/in_offload.h (after previous change).
Because this is in a module, apparently, that means that netinet_in_offload.h
needs to get installed in /usr/include, so do that as well.
Feel free to fix this in a better way...
|
| 1.28 | 16-Feb-2017 |
knakahara | branches: 1.28.12; 1.28.14;
add l2tp(4) L2TPv3 interface.
originally implemented by IIJ SEIL team.
|
| 1.27 | 13-Oct-2015 |
rjs | branches: 1.27.2; 1.27.4;
Add core networking support for SCTP.
|
| 1.26 | 10-Feb-2015 |
rjs | Add DCCP protocol support from KAME.
|
| 1.25 | 15-Sep-2012 |
plunky | branches: 1.25.14;
install header files from IPF 5.1.2 (sys/external/bsd/ipf) instead of
older IPF (sys/dist/ipf).
This adds ipf_rb.h
|
| 1.24 | 25-Jun-2012 |
christos | branches: 1.24.2;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.23 | 15-Feb-2012 |
riz | Back out the recent import of IPFilter 5.1.1 for the upcoming branch,
which will now have IPFilter 4.1.34. IPFilter 5.1.1 will be restored
post-branch.
ok: core, releng.
|
| 1.22 | 30-Jan-2012 |
darrenr | Patch to include ipf_rb.h missed from merge.
|
| 1.21 | 24-Sep-2011 |
christos | branches: 1.21.2; 1.21.6;
install the header.
|
| 1.20 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.19 | 05-Oct-2007 |
dyoung | branches: 1.19.14; 1.19.44; 1.19.50;
Work in progress: use a raw socket for GRE in IP encapsulation
instead of adding/subtracting our own IPv4 header.
There are many benefits: gre(4) needn't grok the outer encapsulation
header any longer, so this simplifies the gre(4) code. The IP
stack needn't grok GRE, so it is simplified, too. gre(4) will
benefit from optimizations in the socket code. Eventually, gre(4)
will gain an IPv6 encapsulation with very few new lines of code.
There is a small performance loss. A 133 MHz, 486-class AMD Elan
sinks/sources a TCP stream over GRE with about 93% the throughput
of the old code. TCP throughput on a 266 MHz, 586-class AMD Geode
is about 96% the throughput of the old code. A 175-MHz ADM5120
(MIPS) only sinks a TCP stream over GRE at about 90% of the old
code; I am still investigating that.
I produced stripped-down versions of sosend() and soreceive() for
gre(4) to use. They are guaranteed not to block, so they can be
called from a software interrupt and from a socket upcall,
respectively.
A kernel thread is no longer necessary for socket transmit/receive,
but I didn't get around to removing it, yet.
Thanks to Matt Thomas for suggesting the use of stripped-down socket
code and software interrupts, and to Andrew Doran for advice and
answers concerning software interrupts, threads, and performance.
|
| 1.18 | 02-May-2007 |
dyoung | branches: 1.18.6; 1.18.8; 1.18.10;
Remove obsolete files netinet/in_route.[ch].
|
| 1.17 | 09-Dec-2006 |
dyoung | branches: 1.17.2; 1.17.6; 1.17.8;
Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.16 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.15 | 18-May-2006 |
liamjfoy | branches: 1.15.8; 1.15.10;
Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.14 | 11-Dec-2005 |
christos | branches: 1.14.4; 1.14.6; 1.14.8; 1.14.12;
merge ktrace-lwp.
|
| 1.13 | 09-Jul-2005 |
xtraeme | Move ipl.h into the ipfilter block, which is the right place.
|
| 1.12 | 01-May-2005 |
martti | branches: 1.12.2;
Install netinet/ipl.h (bin/30095)
|
| 1.11 | 22-Feb-2005 |
peter | branches: 1.11.2;
Add MKIPFILTER; if set to no, don't build and install the ipf(4) programs,
headers and LKM.
Add MKPF; if set to no, don't build and install the pf(4) programs,
headers, LKM and spamd.
Both options default to yes, so nothing changed in the default build.
Reviewed by lukem.
|
| 1.10 | 05-Oct-2004 |
yamt | branches: 1.10.4; 1.10.6;
move ipf headers and add a comment.
|
| 1.9 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.8 | 04-Sep-2004 |
manu | IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.7 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.6 | 28-Mar-2004 |
martti | branches: 1.6.2;
Sync with official IPFilter
|
| 1.5 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.4 | 26-Nov-2002 |
lukem | branches: 1.4.6;
Remove KDIR=, since SYS_INCLUDE=symlinks and KDIR are not supported any more.
|
| 1.3 | 19-Apr-2000 |
itojun | branches: 1.3.6;
add net/if_stf.h and netinet/ip_encap.h (almost noone will include them though)
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 12-Jun-1998 |
cgd | branches: 1.1.10; 1.1.12;
Rework the way kernel include files are installed. In the new method,
as with user-land programs, include files are installed by each directory
in the tree that has includes to install. (This allows more flexibility
as to what gets installed, makes 'partial installs' easier, and gives us
more options as to which machines' includes get installed at any given
time.) The old SYS_INCLUDES={symlinks,copies} behaviours are _both_
still supported, though at least one bug in the 'symlinks' case is
fixed by this change. Include files can't be build before installation,
so directories that have includes as targets (e.g. dev/pci) have to move
those targets into a different Makefile.
|
| 1.1.12.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.1.12.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.1.12.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.1.10.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.3.6.1 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.4.6.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.4.6.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.4.6.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.4.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.4.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.4.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.6.2.1 | 13-Aug-2004 |
jmc | branches: 1.6.2.1.2;
Pullup rev 1.7 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.6.2.1.2.2 | 13-May-2005 |
riz | Pull up revision 1.12 via patch (requested by martti in ticket #1495):
Install netinet/ipl.h (bin/30095)
|
| 1.6.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up patch (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.10.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.10.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.11.2.2 | 24-Jul-2005 |
tron | Pull up revision 1.13 (requested by peter in ticket #612):
Move ipl.h into the ipfilter block, which is the right place.
|
| 1.11.2.1 | 01-May-2005 |
tron | Pull up revision 1.12 (requested by martti in ticket #231):
Install netinet/ipl.h (bin/30095)
|
| 1.12.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.12.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.12.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.12.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.14.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.14.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.14.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.14.4.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.14.4.1 | 02-Feb-2006 |
rpaulo | in_pcb_hdr.h is gone.
|
| 1.15.10.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.15.8.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.15.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.17.8.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.17.6.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.17.6.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.17.2.1 | 07-May-2007 |
yamt | sync with head.
|
| 1.18.10.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.18.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.18.6.1 | 07-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.19.50.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.19.44.1 | 31-May-2011 |
rmind | sync with head
|
| 1.19.14.1 | 22-Feb-2008 |
keiichi | imported Mobile IPv6 code developed by the SHISA project
(http://www.mobileip.jp/).
|
| 1.21.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.21.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.24.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.24.2.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.25.14.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.25.14.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.25.14.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.27.4.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.27.2.1 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.28.14.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.28.12.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.28.12.1 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.3 | 01-Jan-2022 |
andvar | s/creting/creating/
|
| 1.2 | 10-Aug-2008 |
tls | branches: 1.2.2; 1.2.4; 1.2.6; 1.2.18;
Change copyright statement to NetBSD 2-clause with correct attribution.
|
| 1.1 | 04-Aug-2008 |
tls | Add accept filters, ported from FreeBSD by Coyote Point Systems. Add inetd
support for specifying an accept filter for a service (mostly as a usage
example, but it can be handy for other things). Manual pages to follow
in a day or so.
OK core@.
|
| 1.2.18.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.2.18.1 | 10-Aug-2008 |
yamt | file accept_filter.h was added on branch yamt-nfs-mp on 2009-05-04 08:14:17 +0000
|
| 1.2.6.2 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.2.6.1 | 10-Aug-2008 |
haad | file accept_filter.h was added on branch haad-dm on 2008-10-19 22:17:46 +0000
|
| 1.2.4.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.2.4.1 | 10-Aug-2008 |
mjf | file accept_filter.h was added on branch mjf-devfs2 on 2008-09-28 10:40:57 +0000
|
| 1.2.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.2.2.1 | 10-Aug-2008 |
wrstuden | file accept_filter.h was added on branch wrstuden-revivesa on 2008-09-18 04:37:00 +0000
|
| 1.8 | 07-Jul-2016 |
msaitoh | KNF. Remove extra spaces. No functional change.
|
| 1.7 | 20-Aug-2015 |
christos | include "ioconf.h" to get the 'void <driver>attach(int count);' prototype.
|
| 1.6 | 02-Sep-2009 |
tls | branches: 1.6.22; 1.6.40;
Add a direction argument to socket upcalls, so they can tell why they've
been called when, for example, they're waiting for space to write. From
Ritesh Agrawal at Coyote Point.
|
| 1.5 | 20-Nov-2008 |
ad | branches: 1.5.6;
Oops, make these build.
|
| 1.4 | 20-Nov-2008 |
ad | Rename the accept filter modules to make module name match filter name.
|
| 1.3 | 12-Nov-2008 |
ad | Remove LKMs and switch to the module framework, pass 1.
Proposed on tech-kern@.
|
| 1.2 | 14-Oct-2008 |
ad | branches: 1.2.2; 1.2.4;
Use designated initializers for struct accept_filter.
|
| 1.1 | 04-Aug-2008 |
tls | branches: 1.1.2; 1.1.4;
Add accept filters, ported from FreeBSD by Coyote Point Systems. Add inetd
support for specifying an accept filter for a service (mostly as a usage
example, but it can be handy for other things). Manual pages to follow
in a day or so.
OK core@.
|
| 1.1.4.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.1.4.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.1.4.1 | 04-Aug-2008 |
mjf | file accf_data.c was added on branch mjf-devfs2 on 2008-09-28 10:40:57 +0000
|
| 1.1.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.1.2.1 | 04-Aug-2008 |
wrstuden | file accf_data.c was added on branch wrstuden-revivesa on 2008-09-18 04:37:00 +0000
|
| 1.2.4.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.2.2.3 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.2.2.2 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.2.2.1 | 14-Oct-2008 |
haad | file accf_data.c was added on branch haad-dm on 2008-10-19 22:17:46 +0000
|
| 1.5.6.3 | 16-Sep-2009 |
yamt | sync with head
|
| 1.5.6.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.5.6.1 | 20-Nov-2008 |
yamt | file accf_data.c was added on branch yamt-nfs-mp on 2009-05-04 08:14:17 +0000
|
| 1.6.40.2 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.6.40.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.6.22.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.10 | 16-Mar-2020 |
pgoyette | Use the module subsystem's ability to process SYSCTL_SETUP() entries to
automate installation of sysctl nodes.
Note that there are still a number of device and pseudo-device modules
that create entries tied to individual device units, rather than to the
module itself. These are not changed.
|
| 1.9 | 20-Aug-2015 |
christos | branches: 1.9.18;
include "ioconf.h" to get the 'void <driver>attach(int count);' prototype.
|
| 1.8 | 25-Feb-2014 |
pooka | branches: 1.8.6;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.7 | 02-Sep-2009 |
tls | branches: 1.7.12; 1.7.22; 1.7.26;
Add a direction argument to socket upcalls, so they can tell why they've
been called when, for example, they're waiting for space to write. From
Ritesh Agrawal at Coyote Point.
|
| 1.6 | 21-Nov-2008 |
joerg | branches: 1.6.6;
Fix indentation.
|
| 1.5 | 20-Nov-2008 |
ad | Oops, make these build.
|
| 1.4 | 20-Nov-2008 |
ad | Rename the accept filter modules to make module name match filter name.
|
| 1.3 | 12-Nov-2008 |
ad | Remove LKMs and switch to the module framework, pass 1.
Proposed on tech-kern@.
|
| 1.2 | 14-Oct-2008 |
ad | branches: 1.2.2; 1.2.4;
Use designated initializers for struct accept_filter.
|
| 1.1 | 04-Aug-2008 |
tls | branches: 1.1.2; 1.1.4;
Add accept filters, ported from FreeBSD by Coyote Point Systems. Add inetd
support for specifying an accept filter for a service (mostly as a usage
example, but it can be handy for other things). Manual pages to follow
in a day or so.
OK core@.
|
| 1.1.4.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.1.4.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.1.4.1 | 04-Aug-2008 |
mjf | file accf_http.c was added on branch mjf-devfs2 on 2008-09-28 10:40:57 +0000
|
| 1.1.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.1.2.1 | 04-Aug-2008 |
wrstuden | file accf_http.c was added on branch wrstuden-revivesa on 2008-09-18 04:37:00 +0000
|
| 1.2.4.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.2.2.3 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.2.2.2 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.2.2.1 | 14-Oct-2008 |
haad | file accf_http.c was added on branch haad-dm on 2008-10-19 22:17:46 +0000
|
| 1.6.6.3 | 16-Sep-2009 |
yamt | sync with head
|
| 1.6.6.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.6.6.1 | 21-Nov-2008 |
yamt | file accf_http.c was added on branch yamt-nfs-mp on 2009-05-04 08:14:17 +0000
|
| 1.7.26.1 | 18-May-2014 |
rmind | sync with head
|
| 1.7.22.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.7.22.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.7.12.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.8.6.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.9.18.1 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.2 | 28-Aug-2018 |
rin | No need to update mlen also in the case of (mlen & 16) != 0.
|
| 1.1 | 25-Jan-2008 |
joerg | branches: 1.1.2; 1.1.4; 1.1.12; 1.1.102; 1.1.104;
Refactor in_cksum/in4_cksum/in6_cksum implementations:
- All three functions are included in the kernel by default.
They call a backend function cpu_in_cksum after possibly
computing the checksum of the pseudo header.
- cpu_in_cksum is the core to implement the one-complement sum.
The default implementation is moderate fast on most platforms
and provides a 32bit accumulator with 16bit addends for L32 platforms
and a 64bit accumulator with 32bit addends for L64 platforms.
It handles edge cases like very large mbuf chains (could happen with
native IPv6 in the future) and provides a good base for new native
implementations.
- Modify i386 and amd64 assembly to use the new interface.
This disables the MD implementations on !x86 until the conversion is
done. For Alpha, the portable version is faster.
|
| 1.1.104.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.1.102.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.1.12.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.1.12.1 | 25-Jan-2008 |
matt | file cpu_in_cksum.c was added on branch matt-armv6 on 2008-03-23 02:05:06 +0000
|
| 1.1.4.2 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.1.4.1 | 25-Jan-2008 |
mjf | file cpu_in_cksum.c was added on branch mjf-devfs on 2008-02-18 21:07:08 +0000
|
| 1.1.2.2 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.1.2.1 | 25-Jan-2008 |
yamt | file cpu_in_cksum.c was added on branch yamt-lazymbuf on 2008-02-04 09:24:39 +0000
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add DCCP protocol support from KAME.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 10-Feb-2015 |
jdolecek | file dccp.h was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp.h was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.3 | 26-Apr-2016 |
ozaki-r | branches: 1.3.16;
Sweep unnecessary route.h inclusions
|
| 1.2 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2;
Add DCCP protocol support from KAME.
|
| 1.1.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_cc_sw.c was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.3.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.3.16.1 | 26-Apr-2016 |
jdolecek | file dccp_cc_sw.c was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add DCCP protocol support from KAME.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 10-Feb-2015 |
jdolecek | file dccp_cc_sw.h was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_cc_sw.h was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.5 | 22-May-2022 |
andvar | fix various small typos, mainly in comments.
|
| 1.4 | 04-Jun-2019 |
msaitoh | Fix typo (s/recevie/receive/).
|
| 1.3 | 26-Apr-2016 |
ozaki-r | branches: 1.3.16; 1.3.20;
Sweep unnecessary route.h inclusions
|
| 1.2 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2;
Add DCCP protocol support from KAME.
|
| 1.1.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_tcplike.c was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.3.20.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.3.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.3.16.1 | 26-Apr-2016 |
jdolecek | file dccp_tcplike.c was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.3 | 04-Jun-2019 |
msaitoh | Fix typo (s/recevie/receive/).
|
| 1.2 | 07-Jul-2016 |
msaitoh | branches: 1.2.16; 1.2.20;
KNF. Remove extra spaces. No functional change.
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2;
Add DCCP protocol support from KAME.
|
| 1.1.2.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_tcplike.h was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.2.20.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.2.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.16.1 | 07-Jul-2016 |
jdolecek | file dccp_tcplike.h was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.12 | 12-May-2024 |
msaitoh | s/unitialized/uninitialized/
|
| 1.11 | 14-Aug-2023 |
mrg | avoid uninitialised variable use.
found by GCC 12.
|
| 1.10 | 07-Aug-2023 |
mrg | fix indentation issues.
found by GCC 12.
|
| 1.9 | 10-Dec-2021 |
andvar | branches: 1.9.4;
s/occured/occurred/ in comments, log messages and man pages.
|
| 1.8 | 05-Dec-2021 |
msaitoh | s/receieve/receive/
|
| 1.7 | 07-Sep-2021 |
andvar | s/aquire/acquire/ in comments, also one typo fix acqure->acquire.
|
| 1.6 | 27-Dec-2019 |
msaitoh | s/inital/initial/
|
| 1.5 | 04-Jun-2019 |
msaitoh | Fix typo (s/recevie/receive/).
|
| 1.4 | 07-Jul-2016 |
msaitoh | branches: 1.4.16; 1.4.20;
KNF. Remove extra spaces. No functional change.
|
| 1.3 | 26-Apr-2016 |
ozaki-r | Sweep unnecessary route.h inclusions
|
| 1.2 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2;
Add DCCP protocol support from KAME.
|
| 1.1.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_tfrc.c was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.4.20.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.4.20.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.4.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.4.16.1 | 07-Jul-2016 |
jdolecek | file dccp_tfrc.c was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.9.4.1 | 13-Sep-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #862):
sys/netinet/dccp_tfrc.c: revision 1.11
avoid uninitialised variable use.
found by GCC 12.
|
| 1.3 | 04-Jun-2019 |
msaitoh | Fix typo (s/recevie/receive/).
|
| 1.2 | 07-Jul-2016 |
msaitoh | branches: 1.2.16; 1.2.20;
KNF. Remove extra spaces. No functional change.
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2;
Add DCCP protocol support from KAME.
|
| 1.1.2.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_tfrc.h was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.2.20.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.2.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.16.1 | 07-Jul-2016 |
jdolecek | file dccp_tfrc.h was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add DCCP protocol support from KAME.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 10-Feb-2015 |
jdolecek | file dccp_tfrc_lookup.h was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_tfrc_lookup.h was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.27 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.26 | 04-Nov-2022 |
ozaki-r | branches: 1.26.8;
inpcb: rename functions to in6pcb_*
|
| 1.25 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.24 | 28-Oct-2022 |
ozaki-r | Adjust dccp and sctp for struct inpcb separation
|
| 1.23 | 28-Oct-2022 |
ozaki-r | Adjust pf, wg, dccp and sctp for struct inpcb integration
|
| 1.22 | 04-Dec-2021 |
andvar | fix typos in comments and log messages, mainly in establish(ed).
|
| 1.21 | 16-Dec-2018 |
christos | sbspace() does not return negative values anymore and that broke OOB data
sending. Instead of depending on negative values, account for the 1024
bytes sosend() adds so that it can use all the space here in a separate
function sbspace_oob(). Idea from mlelstv@
|
| 1.20 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.19 | 18-May-2018 |
maxv | branches: 1.19.2;
IP6_EXTHDR_GET -> M_REGION_GET, no functional change.
|
| 1.18 | 03-May-2018 |
maxv | Remove m_copy completely.
|
| 1.17 | 08-Feb-2018 |
dholland | branches: 1.17.2;
Typos.
|
| 1.16 | 07-May-2017 |
rjs | branches: 1.16.8;
Change bzero -> memset, bcopy -> memcpy.
|
| 1.15 | 07-May-2017 |
rjs | Change SPL around call to in_pcbbind().
|
| 1.14 | 07-May-2017 |
rjs | Remove some foreign conditional code, NFC intended.
|
| 1.13 | 03-Mar-2017 |
ozaki-r | branches: 1.13.4;
Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.12 | 16-Jan-2017 |
christos | ip6_sprintf -> IN6_PRINT so that we pass the size.
|
| 1.11 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.10 | 13-Dec-2016 |
ozaki-r | branches: 1.10.2;
Remove unnecessary inclusions of nd6.h
|
| 1.9 | 07-Jul-2016 |
msaitoh | branches: 1.9.2;
KNF. Remove extra spaces. No functional change.
|
| 1.8 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.7 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.6 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.5 | 27-Apr-2015 |
ozaki-r | Introduce in6_selecthlim_rt to consolidate an idiom for rt->rt_ifp
It consolidates a scattered routine:
(rt = rtcache_validate(&in6p->in6p_route)) != NULL ? rt->rt_ifp : NULL
|
| 1.4 | 26-Apr-2015 |
rtr | remove pr_generic from struct pr_usrreqs and all implementations of
pr_generic in protocols.
bump to 7.99.13
approved by rmind@
|
| 1.3 | 24-Apr-2015 |
rtr | make accept, getsockname and getpeername syscalls use sockaddr_big and modify
pr_{accept,sockname,peername} nam parameter type from mbuf * to sockaddr *.
* retained use of mbuftypes[MT_SONAME] for now.
* bump to netbsd version 7.99.12 for parameter type change.
patch posted to tech-net@ 2015/04/19
|
| 1.2 | 04-Apr-2015 |
rtr | branches: 1.2.2;
* update dccp_bind for struct mbuf * to struct sockaddr * parameter change
* pass NULL instead of casting 0 to a pointer when calling in_pcbbind()
|
| 1.1 | 10-Feb-2015 |
rjs | Add DCCP protocol support from KAME.
|
| 1.2.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.2.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.2.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.2.2.4 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.2.2.3 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.2.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.2.2.1 | 04-Apr-2015 |
skrll | file dccp_usrreq.c was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.9.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.9.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.10.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.13.4.1 | 11-May-2017 |
pgoyette | Sync with HEAD
|
| 1.16.8.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.16.8.1 | 07-May-2017 |
jdolecek | file dccp_usrreq.c was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.17.2.3 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.17.2.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.17.2.1 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.19.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.26.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.7 | 28-Oct-2022 |
ozaki-r | Adjust pf, wg, dccp and sctp for struct inpcb integration
|
| 1.6 | 07-Nov-2021 |
andvar | fix various typos, mainly s/prefered/preferred/
|
| 1.5 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.4 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.3 | 07-Jul-2016 |
msaitoh | branches: 1.3.16; 1.3.18; 1.3.20;
KNF. Remove extra spaces. No functional change.
|
| 1.2 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.1 | 10-Feb-2015 |
rjs | branches: 1.1.2;
Add DCCP protocol support from KAME.
|
| 1.1.2.4 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 10-Feb-2015 |
skrll | file dccp_var.h was added on branch nick-nhusb on 2015-04-06 15:18:22 +0000
|
| 1.3.20.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.3.18.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.3.18.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.3.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.3.16.1 | 07-Jul-2016 |
jdolecek | file dccp_var.h was added on branch tls-maxphys on 2017-12-03 11:39:03 +0000
|
| 1.71 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.70 | 06-Sep-2004 |
darrenr | Do not allow packets flagged with "out-of-window" (oow) to match "keep state"
rules and try to prevent such rules ("keep state with oow") from being loaded
into the kernel.
Pr: kern/26581
|
| 1.69 | 03-Sep-2004 |
smb | Don't try and add a state session if the packet has already been checked
and marked as out of window - trying to do the add will result in a failure
and the packet being blocked, incorrectly.
Committed By: darrenr
Tested By: smb
|
| 1.68 | 22-Aug-2004 |
chs | fix m_pulldown() usage, it's different from m_pullup().
fixes PRs 26666 and 26701.
|
| 1.67 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.66 | 16-Jun-2004 |
tron | Correct two errors in fr_check():
1.) Make sure that "pass" is always initialized.
2.) Make sure the code doesn't use a stale mbuf pointer after fr_makefrip()
has been called. This fixes PR kern/25868.
Analyzed and reviewed by Steve Woodford.
|
| 1.65 | 20-May-2004 |
christos | PR/25622: IPV6 return RST and through cloned interfaces was broken.
- checksum was computed incorrectly.
- ipv6 packet was not initialized properly.
- fixed code to be more similar to the v4 counterpart.
|
| 1.64 | 10-May-2004 |
christos | PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.63 | 04-May-2004 |
skd | Fix to update all references to mbuf. Fixes case where mbuf is freed twice.
|
| 1.62 | 01-Apr-2004 |
martin | A few more ioctl vs. copyin changes, spotted by Bill Studenmund.
|
| 1.61 | 28-Mar-2004 |
martti | branches: 1.61.2;
Sync with official IPFilter
|
| 1.60 | 28-Mar-2004 |
martti | Upgraded IPFilter to 4.1.1
|
| 1.59 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.58 | 19-Sep-2002 |
martti | branches: 1.58.6;
Resync with official IPF
|
| 1.57 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.56 | 09-Jun-2002 |
itojun | whitespace
|
| 1.55 | 02-May-2002 |
martti | branches: 1.55.2; 1.55.4;
Fix compilation problems
|
| 1.54 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.53 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.52 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.51 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.50 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.49 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.48 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.47 | 02-Jun-2001 |
thorpej | branches: 1.47.2; 1.47.6;
Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.46 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.45 | 05-Feb-2001 |
chs | branches: 1.45.2;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.44 | 18-Jan-2001 |
jdolecek | constify
|
| 1.43 | 12-Nov-2000 |
thorpej | Due to a quirk (err, bug?) in IP Filter (mbuf freed without setting *mp
to NULL), the NULL check is insufficient. Also make sure fr_check()
returned 0.
|
| 1.42 | 12-Nov-2000 |
thorpej | Oops, the mbuf may have been freed -- do a NULL check in the wrapper.
|
| 1.41 | 11-Nov-2000 |
thorpej | Restructure the PFIL_HOOKS mechanism a bit:
- All packets are passed to PFIL_HOOKS as they come off the wire, i.e.
fields in protocol headers in network order, etc.
- Allow for multiple hooks to be registered, using a "key" and a "dlt".
The "dlt" is a BPF data link type, indicating what type of header is
present.
- INET and INET6 register with key == AF_INET or AF_INET6, and
dlt == DLT_RAW.
- PFIL_HOOKS now take an argument for the filter hook, and mbuf **,
an ifnet *, and a direction (PFIL_IN or PFIL_OUT), thus making them
less IP (really, IP Filter) centric.
Maintain compatibility with IP Filter by adding wrapper functions for
IP Filter.
|
| 1.40 | 08-Oct-2000 |
itojun | ipfilter currently supports IPv4 only. do not try to touch non-IPv4
packets. PR 11082.
This is a short-term workaround. whenever new ipfilter comes out with
proper non-IPv4 support, we should migrate to the new ipfilter.
|
| 1.39 | 12-Aug-2000 |
enami | Put # endif directive after the right (i.e., matching) close brace
to prevent compilation error.
|
| 1.38 | 12-Aug-2000 |
veego | Protect a IPLLOG with ifdef IPFILTER_LOG. Patch from Darren Reed.
|
| 1.37 | 09-Aug-2000 |
veego | Resolve conflicts.
|
| 1.36 | 12-Jun-2000 |
veego | branches: 1.36.2;
Resolve conflicts.
|
| 1.35 | 23-May-2000 |
veego | branches: 1.35.2;
Resolve conflicts.
|
| 1.34 | 21-May-2000 |
veego | Resolve conflicts.
|
| 1.33 | 11-May-2000 |
veego | Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.32 | 10-May-2000 |
itojun | correct out-of-bound access when hlen == 1 and opt > 1.
reviewed by darren, darren committed to freebsd fil.c (1.12 -> 1.13)
so it should be correct enough.
|
| 1.31 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.30 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.29 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.28 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.27 | 02-Feb-1999 |
cjs | branches: 1.27.2; 1.27.6; 1.27.8; 1.27.14;
Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.26 | 23-Jan-1999 |
mycroft | Fix problems with fr_tcpsum() that prevented the FTP proxy from working.
|
| 1.25 | 26-Nov-1998 |
mrg | add a patch from darren reed, to make ipfilter use our cksum routine.
|
| 1.24 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.23 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.22 | 31-May-1998 |
cgd | Another demonstration that when you're converting variables from 'long's
to fixed 32-bit integers, you have to exercise care.
|
| 1.21 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.20 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.19 | 01-May-1998 |
thorpej | If packets are passed through IP Filter at all, don't allow fast-forward
flow entries to be created for them.
Eventually, IP Filter should be extended to allow IP src/dst pairs to
be specified as "fast forward OK".
|
| 1.18 | 17-Nov-1997 |
mrg | fix checksum problems (from marc boucher via darren reed).
|
| 1.17 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.16 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.15 | 21-Sep-1997 |
veego | branches: 1.15.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.14 | 08-Jul-1997 |
mrg | branches: 1.14.2;
put back IPFILTER_DEFAULT_BLOCK, as documented in options(4).
|
| 1.13 | 06-Jul-1997 |
thorpej | Restore original RCS IDs.
|
| 1.12 | 06-Jul-1997 |
thorpej | Fix a bug caught by gcc: add parenthesis to properly group a test.
|
| 1.11 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.10 | 16-Jun-1997 |
mrg | make it "options IPFILTER_DEFAULT_BLOCK".
|
| 1.9 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.8 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures again:
- garbage-collect unused variables, or #ifdef them as appropriate.
|
| 1.7 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.6 | 29-Mar-1997 |
darrenr | use IPLLOG instead of ipllog to easily mask parameters, fix up prototype
problems for compiling to user programs.
|
| 1.5 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.4 | 19-Feb-1997 |
scottr | Don't include ipfilter.h if building an LKM.
|
| 1.3 | 18-Feb-1997 |
mrg | pseudo-device ipfilter brings in PFIL_HOOKS.
|
| 1.2 | 05-Jan-1997 |
veego | branches: 1.2.4;
Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.27 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.26 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.25 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.24 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.23 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.22 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.21 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.20 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.19 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.18 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.17 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.16 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.15 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.14 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.2.4.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.14.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.15.2.6 | 29-Nov-1998 |
cgd | pull up rev 1.25 from trunk (mrg)
|
| 1.15.2.5 | 24-Nov-1998 |
cgd | pull up rev(s) 1.24 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.15.2.4 | 22-Sep-1998 |
cgd | Redo previous change (Pull up 1.20-1.23 (veego)). Previous commit actually
pulled up 1.21-1.23, and incorrectly omitted the changes in rev 1.20.
|
| 1.15.2.3 | 17-Sep-1998 |
mellon | Pull up 1.20-1.23 (veego)
|
| 1.15.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.15.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.27.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.27.8.4 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.27.8.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.27.8.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.27.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.27.6.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.27.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.27.2.2 | 13-May-2000 |
he | Pull up revision 1.32 (requested by darrenr):
Fix bug in dealing with large offsets inside very small options.
|
| 1.27.2.1 | 20-Dec-1999 |
he | Pull up revision 1.28 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.35.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.36.2.5 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.36.2.4 | 09-Feb-2002 |
he | Pull up revisions 1.41-1.52 (requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.36.2.3 | 08-Oct-2000 |
itojun | pullup 1.39 -> 1.40 (approved by releng-1-5)
ipfilter currently supports IPv4 only. do not try to touch non-IPv4
packets. PR 11082.
This is a short-term workaround. whenever new ipfilter comes out with
proper non-IPv4 support, we should migrate to the new ipfilter.
|
| 1.36.2.2 | 31-Aug-2000 |
veego | Pull up revisions 1.38-1.39 (requested by veego). Approved by releng-1-5.
>syssrc/sys/netinet/fil.c 1.38
>Committed By: veego
>Log Message:
>Protect a IPLLOG with ifdef IPFILTER_LOG. Patch from Darren Reed.
>syssrc/sys/netinet/fil.c 1.39
>Committed By: enami
>Log Message:
>Put # endif directive after the right (i.e., matching) close brace
>to prevent compilation error.
|
| 1.36.2.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.45.2.9 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.45.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.45.2.7 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.45.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.45.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.45.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.45.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.45.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.45.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.47.6.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.47.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.47.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.47.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.47.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.55.4.2 | 26-Oct-2005 |
riz | Apply patch (requested by darrenr in ticket #1780):
Don't pass a NULL pointer to ipfr_fastroute().
From Christoph Egger in PR#26141
|
| 1.55.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.55.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.58.6.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.58.6.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.58.6.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.58.6.3 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.58.6.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.58.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.61.2.10 | 10-Jan-2005 |
jmc | Pullup patch (requested by hubertf in ticket #1068)
Pull revs 1.2-1.7 from sys/dist/ipf/netinet/fil.c from trunk to fix panics
with ipf and IPv6. Fixes PR#28875 and #26839
|
| 1.61.2.9 | 12-Nov-2004 |
jmc | branches: 1.61.2.9.2;
Pullup patch (requested by darrenr in ticket #910)
Fix for previous revision was incorrect. it checks for FI_OOW regardless of what
type of data is stored in the rule (only a valid check for FR_T_IPF rules.)
|
| 1.61.2.8 | 04-Oct-2004 |
jmc | Pullup rev 1.69-1.70+patch (requested by jdolecek in ticket #888)
Do not allow packets flagged with "out-of-window" (oow) to match "keep state"
rules and try to prevent such rules ("keep state with oow") from being loaded
into the kernel. PR#26581
|
| 1.61.2.7 | 23-Aug-2004 |
tron | Pull up revision 1.68 (requested by chs in ticket #783):
fix m_pulldown() usage, it's different from m_pullup().
fixes PRs 26666 and 26701.
|
| 1.61.2.6 | 13-Aug-2004 |
jmc | Pullup rev 1.67 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.61.2.5 | 18-Jun-2004 |
grant | Pull up revision 1.66 (requested by tron in ticket #502):
Correct two errors in fr_check():
1.) Make sure that "pass" is always initialized.
2.) Make sure the code doesn't use a stale mbuf pointer after fr_makefrip()
has been called. This fixes PR kern/25868.
|
| 1.61.2.4 | 30-May-2004 |
tron | Pull up revision 1.65 (requested by christos in ticket #416):
PR/25622: IPV6 return RST and through cloned interfaces was broken.
- checksum was computed incorrectly.
- ipv6 packet was not initialized properly.
- fixed code to be more similar to the v4 counterpart.
|
| 1.61.2.3 | 30-May-2004 |
tron | Pull up revision 1.64 (requested by christos in ticket #416):
PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.61.2.2 | 30-May-2004 |
tron | Pull up revision 1.63 (requested by christos in ticket #416):
Fix to update all references to mbuf. Fixes case where mbuf is freed twice.
|
| 1.61.2.1 | 02-Apr-2004 |
tron | Pull up revision 1.62 (requested by martin in ticket #46):
A few more ioctl vs. copyin changes, spotted by Bill Studenmund.
|
| 1.61.2.9.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.71 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.16 | 04-Dec-2020 |
thorpej | Build ip_sync.c with -Wno-error to avoid failing due to excessive stack
usage.
|
| 1.15 | 23-Mar-2012 |
christos | branches: 1.15.52;
move back to 5.1.1 in the new place
|
| 1.14 | 15-Feb-2012 |
riz | Back out the recent import of IPFilter 5.1.1 for the upcoming branch,
which will now have IPFilter 4.1.34. IPFilter 5.1.1 will be restored
post-branch.
ok: core, releng.
|
| 1.13 | 30-Jan-2012 |
darrenr | New files required to build ipfilter into the kernel
|
| 1.12 | 02-Oct-2010 |
bad | branches: 1.12.8; 1.12.12;
Defopt the rest of the Ipfilter options and tunables.
Per discussion with darrenr@ a year ago.
|
| 1.11 | 17-Apr-2010 |
darrenr | fix spelling mistake: netient -> netinet
|
| 1.10 | 17-Apr-2010 |
darrenr | add IPFILTER_COMPAT to kernel config options recognised for IPFilter
|
| 1.9 | 24-Jan-2010 |
pooka | branches: 1.9.2; 1.9.4;
ipfilter depends on bpf_filter, not bpfilter (since the year 2000).
|
| 1.8 | 17-Sep-2006 |
yamt | branches: 1.8.54;
defflag IPFILTER_LOOKUP.
|
| 1.7 | 11-Dec-2005 |
christos | branches: 1.7.20;
merge ktrace-lwp.
|
| 1.6 | 26-Mar-2005 |
christos | branches: 1.6.2;
defopt IPFILTER_DEFAULT_BLOCK
|
| 1.5 | 01-Oct-2004 |
christos | branches: 1.5.4;
Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.4 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.3 | 28-Mar-2004 |
martti | branches: 1.3.2;
Sync with official IPFilter
|
| 1.2 | 11-Oct-2002 |
thorpej | branches: 1.2.2; 1.2.8;
Add missing "needs-flag".
|
| 1.1 | 10-Oct-2002 |
thorpej | Move netinet, netinet6, ipsec, and ipfilter config defns to
netinet/files.ipfilter, etinet/files.netinet, netinet6/files.netinet6,
and netinet6/files.netipsec.
XXX There are still a few stragglers in conf/files, which are entangled
with other network protocols.
|
| 1.2.8.3 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.2.8.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.2.8.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.2.2.2 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.1 | 11-Oct-2002 |
nathanw | file files.ipfilter was added on branch nathanw_sa on 2002-10-18 02:45:16 +0000
|
| 1.3.2.1 | 13-Aug-2004 |
jmc | branches: 1.3.2.1.2;
Pullup rev 1.4 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.3.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up patch (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.5.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.6.2.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.7.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.8.54.3 | 09-Oct-2010 |
yamt | sync with head
|
| 1.8.54.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.8.54.1 | 11-Mar-2010 |
yamt | sync with head
|
| 1.9.4.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.9.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.9.2.2 | 22-Oct-2010 |
uebayasi | Sync with HEAD (-D20101022).
|
| 1.9.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.12.12.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.12.12.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.12.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.15.52.1 | 14-Dec-2020 |
thorpej | Sync w/ HEAD.
|
| 1.30 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.29 | 08-Mar-2021 |
christos | remove now unused pseudo-random ip id code.
|
| 1.28 | 29-Jul-2017 |
maxv | branches: 1.28.16;
Remove TCP_COMPAT_42.
|
| 1.27 | 13-Oct-2015 |
rjs | Add core networking support for SCTP.
|
| 1.26 | 10-Feb-2015 |
rjs | Add DCCP protocol support from KAME.
|
| 1.25 | 02-Dec-2014 |
christos | add routines to print in_addr and sockaddr_in (in_print and sin_print)
|
| 1.24 | 25-Jun-2012 |
christos | branches: 1.24.2; 1.24.16;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.23 | 24-Sep-2011 |
christos | branches: 1.23.2;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.22 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.21 | 13-Jul-2010 |
rmind | branches: 1.21.2;
Split-off IPv4 re-assembly mechanism into a separate module. Abstract
into ip_reass_init(), ip_reass_lookup(), etc (note: abstraction is not
yet complete). No functional changes to the actual mechanism.
OK matt@
|
| 1.20 | 25-Jan-2008 |
joerg | branches: 1.20.10; 1.20.30; 1.20.32;
Refactor in_cksum/in4_cksum/in6_cksum implementations:
- All three functions are included in the kernel by default.
They call a backend function cpu_in_cksum after possibly
computing the checksum of the pseudo header.
- cpu_in_cksum is the core to implement the one-complement sum.
The default implementation is moderate fast on most platforms
and provides a 32bit accumulator with 16bit addends for L32 platforms
and a 64bit accumulator with 32bit addends for L64 platforms.
It handles edge cases like very large mbuf chains (could happen with
native IPv6 in the future) and provides a good base for new native
implementations.
- Modify i386 and amd64 assembly to use the new interface.
This disables the MD implementations on !x86 until the conversion is
done. For Alpha, the portable version is faster.
|
| 1.19 | 02-May-2007 |
dyoung | branches: 1.19.8; 1.19.14;
Remove obsolete files netinet/in_route.[ch].
|
| 1.18 | 02-May-2007 |
dyoung | Remove unused option.
|
| 1.17 | 09-Dec-2006 |
dyoung | branches: 1.17.2; 1.17.6; 1.17.8;
Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.16 | 25-Nov-2006 |
yamt | move tso-by-software code to their own files. no functional changes.
|
| 1.15 | 23-Nov-2006 |
tron | Backout accidental commit which broke kernel builds.
|
| 1.14 | 23-Nov-2006 |
rpaulo | New EtherIP driver based on tap(4) and gif(4) by Hans Rosenfeld.
Notable changes:
* Fixes PR 34268.
* Separates the code from gif(4) (which is more cleaner).
* Allows the usage of STP (Spanning Tree Protocol).
* Removed EtherIP implementation from gif(4)/tap(4).
Some input from Christos.
|
| 1.13 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.12 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.11 | 11-Dec-2005 |
christos | branches: 1.11.20; 1.11.22;
merge ktrace-lwp.
|
| 1.10 | 28-Feb-2005 |
jonathan | branches: 1.10.4;
Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.9 | 13-Jan-2005 |
drochner | branches: 1.9.2; 1.9.4;
compile tcp_debug.c only if the TCP_DEBUG option is set,
and remove the "#ifdef TCP_DEBUG" around everything
|
| 1.8 | 04-Sep-2004 |
manu | IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.7 | 01-May-2004 |
matt | defflag TCP_OUTPUT_COUNTERS and TCP_REASS_COUNTERS
|
| 1.6 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.5 | 26-Nov-2003 |
itojun | always compile ip_id.c
|
| 1.4 | 26-Nov-2003 |
itojun | define RANDOM_IP_ID by default (unifdef -DRANDOM_IP_ID).
one use remains in sys/netipsec, which is kept for freebsd source code compat.
|
| 1.3 | 17-Nov-2003 |
jonathan | Revert the (default) ip_id algorithm to the pre-randomid algorithm,
due to demonstrated low-period repeated IDs from the randomized IP_id
code. Consensus is that the low-period repetition (much less than
2^15) is not suitable for general-purpose use.
Allocators of new IPv4 IDs should now call the function ip_newid().
Randomized IP_ids is now a config-time option, "options RANDOM_IP_ID".
ip_newid() can use ip_random-id()_IP_ID if and only if configured
with RANDOM_IP_ID. A sysctl knob should be provided.
This API may be reworked in the near future to support linear ip_id
counters per (src,dst) IP-address pair.
|
| 1.2 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.1 | 10-Oct-2002 |
thorpej | branches: 1.1.2; 1.1.8;
Move netinet, netinet6, ipsec, and ipfilter config defns to
netinet/files.ipfilter, etinet/files.netinet, netinet6/files.netinet6,
and netinet6/files.netipsec.
XXX There are still a few stragglers in conf/files, which are entangled
with other network protocols.
|
| 1.1.8.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.1.8.4 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.1.8.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.8.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.8.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.1.2.1 | 10-Oct-2002 |
nathanw | file files.netinet was added on branch nathanw_sa on 2002-10-18 02:45:16 +0000
|
| 1.9.4.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.9.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.10.4.3 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.10.4.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.10.4.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.11.22.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.11.22.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.11.20.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.11.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.17.8.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.17.6.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.17.2.1 | 07-May-2007 |
yamt | sync with head.
|
| 1.19.14.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.19.8.1 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.20.32.2 | 31-May-2011 |
rmind | sync with head
|
| 1.20.32.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.20.30.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.20.10.1 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.21.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.23.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.24.16.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.24.16.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.24.16.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.24.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.28.16.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.61 | 06-Dec-2024 |
riastradh | netinet/icmp6.h: Nix trailing whitespace.
No functional change intended.
|
| 1.60 | 06-Dec-2024 |
riastradh | netinet/icmp6.h: Need sys/types.h and netinet/in.h.
- sys/types.h for u_intN_t
- netinet/in.h for struct in6_addr
|
| 1.59 | 29-Aug-2022 |
knakahara | branches: 1.59.10;
Add sysctl entry to control to send routing message for RTM_DYNAMIC.
Some routing daemons require such routing message to keep coherency.
If we want to let kernel send such message, set net.inet.icmp.dynamic_rt_msg=1
for IPv4, net.inet6.icmp6.dynamic_rt_msg=1 for IPv6.
Default(=0) is the same as before, that is, not send such routing message.
|
| 1.58 | 22-Aug-2022 |
knakahara | Add sysctl entry to enable/disable to use path MTU discovery for icmpv6 reflecting.
If we want to use path MTU discovery for icmp reflecting set
net.inet6.icmp6.reflect_pmtu=1. Default(=0) is the same as before, that is,
use IPV6_MINMTU.
|
| 1.57 | 27-Jul-2020 |
roy | icmp6: Remove __packed attribute from icmp6 structures
They should naturally align.
Add compile time assertations to icmp6.c to prove this.
|
| 1.56 | 15-Jun-2020 |
roy | icmp6.h: #define ND_RA_FLAG_PROXY
RFC 4389, experimental. Maybe someone will implement it one day.
|
| 1.55 | 15-Jun-2020 |
roy | icmp6.h: #define ND_OPT_PI_FLAG_ROUTER
We already define ND_RA_FLAG_HOME_AGENT and that kind of requires
ND_OPT_PI_FLAG_ROUTER.
|
| 1.54 | 12-Jun-2020 |
roy | Remove in-kernel handling of Router Advertisements
This is much better handled by a user-land tool.
Proposed on tech-net here:
https://mail-index.netbsd.org/tech-net/2020/04/22/msg007766.html
Note that the ioctl SIOCGIFINFO_IN6 no longer sets flags. That now
needs to be done using the pre-existing SIOCSIFINFO_FLAGS ioctl.
Compat is fully provided where it makes sense, but trying to turn on
RA handling will obviously throw an error as it no longer exists.
Note that if you use IPv6 temporary addresses, this now needs to be
turned on in dhcpcd.conf(5) rather than in sysctl.conf(5).
|
| 1.53 | 09-Mar-2020 |
roy | route: RTM_MISS now puts the message source address in RTA_AUTHOR
route(8) also reports this.
A userland app could use this to blacklist nodes who probe for machines
that doesn't exist on a subnet / prefix.
|
| 1.52 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.51 | 24-Apr-2018 |
maxv | branches: 1.51.2;
Add code 3 of paramprob, part of RFC7112: "IPv6 First Fragment has
incomplete IPv6 Header Chain". Handle this code in ping6.
|
| 1.50 | 06-Mar-2018 |
roy | nd6: add a nonce to DaD probes in-case they are looped back to us
This implements RFC 7527, based a similar change in FreeBSD.
|
| 1.49 | 23-Jan-2018 |
maxv | branches: 1.49.2;
Style, localify, remove XXX when there's no issue, and switch 'extra'
to int.
|
| 1.48 | 11-Dec-2016 |
ozaki-r | Remove unnecessary forward struct declarations
|
| 1.47 | 01-Jul-2013 |
christos | branches: 1.47.8; 1.47.12;
Add MLD_LISTENER_REDUCTION per RFC 3542. Requested by Lorenzo Colitti.
|
| 1.46 | 19-Jul-2012 |
spz | branches: 1.46.2; 1.46.4;
<grmbl>whitespace</grmbl>
|
| 1.45 | 19-Jul-2012 |
spz | this commit contains two sets of unrelated changes:
"while I was here" I checked other KAME implementations for their icmp6.h
version, and thus:
- added a define for MLDV2_LISTENER_REPORT from FreeBSD
- added defines for the missing ICMP6_DST_UNREACH codes
then on to what I actually wanted to do:
- adds strings for the types and codes (encapsulated by ICMP6_STRINGS)
for the use of npfctl and other tools that might want to parse
human-friendly names instead of the corresponding number for ipv6-icmp
types and codes.
The strings are ordered such that their index is (as far as is practical)
the number belonging to the name, which is why there are
icmp6_type_err (use directly) and icmp6_type_info (add 128)
|
| 1.44 | 10-Dec-2011 |
roy | branches: 1.44.2;
Add RDNSS and DNSSL support, RFC6106.
Replace custom lists with TAILQ lists.
Clean up plently of signed vs unsigned warnings and set WARNS=4.
Adapted from FreeBSD.
|
| 1.43 | 11-Nov-2011 |
gdt | branches: 1.43.4;
Move RTF_ANNOUNCE flag so that it no longer conflicts with RTF_PROTO2.
RTF_ANNOUNCE was defined as RTF_PROTO2. The flag is used to indicated
that host should act as a proxy for a link level arp or ndp request.
(If RTF_PROTO2 is used as an experimental flag (as advertised),
various problems can occur.)
This commit provides a first-class definition with its own bit for
RTF_ANNOUNCE, removes the old aliasing definitions, and adds support
for the new RTF_ANNOUNCE flag to netstat(8) and route(8).,
Also, remove unused RTF_ flags that collide with RTF_PROTO1:
netinet/icmp6.h defined RTF_PROBEMTU as RTF_PROTO1
netinet/if_inarp.h defined RTF_USETRAILERS as RTF_PROTO1
(Neither of these flags are used anywhere. Both have been removed
to reduce chances of collision with RTF_PROTO1.)
Figuring this out and the diff are the work of Beverly Schwartz of
BBN.
(Passed release build, boot in VM, with no apparently related atf
failures.)
Approved for Public Release, Distribution Unlimited
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073.
|
| 1.42 | 24-May-2011 |
spz | branches: 1.42.4;
RA flood mitigation via a limit on accepted routes:
- introduce a limit for the routes accepted via IPv6 Router Advertisement:
a common 2 interface client will have 6, the default limit is 100 and
can be adjusted via sysctl
- report the current number of routes installed via RA via sysctl
- count discarded route additions. Note that one RA message is two routes.
This is at present only across all interfaces even though per-interface
would be more useful, since the per-interface structure complies to RFC2466
- bump kernel version due to the previous change
- adjust netstat to use the new value (with netstat -p icmp6)
|
| 1.41 | 08-May-2011 |
spz | update (unused) ND option identifiers and corresponding comments
|
| 1.40 | 31-Oct-2009 |
christos | branches: 1.40.4; 1.40.6;
add enough info to let rtadvd compile with route-info.
|
| 1.39 | 11-Jul-2008 |
cyber | Add IANA allocation and header for RFC 5006 (RA RDNSS) IPv6 Router
Advertisement option.
|
| 1.38 | 15-Apr-2008 |
thorpej | branches: 1.38.4; 1.38.6; 1.38.8; 1.38.10;
Make ip6 and icmp6 stats per-cpu.
|
| 1.37 | 08-Apr-2008 |
thorpej | Change ICMP6 stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old icmp6stat structure; old netstat
binaries will continue to work properly.
|
| 1.36 | 25-Dec-2007 |
perry | branches: 1.36.2; 1.36.6;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.35 | 07-Mar-2006 |
wiz | branches: 1.35.36; 1.35.42; 1.35.46; 1.35.50;
'advertisment' -> 'advertisement', from leonardo chiquitto filho
via jmc@openbsd.
|
| 1.34 | 05-Mar-2006 |
rpaulo | branches: 1.34.2;
NDP-related improvements:
RFC4191
- supports host-side router-preference
RFC3542
- if DAD fails on a interface, disables IPv6 operation on the
interface
- don't advertise MLD report before DAD finishes
Others
- fixes integer overflow for valid and preferred lifetimes
- improves timer granularity for MLD, using callout-timer.
- reflects rtadvd's IPv6 host variable information into kernel
(router only)
- adds a sysctl option to enable/disable pMTUd for multicast
packets
- performs NUD on PPP/GRE interface by default
- Redirect works regardless of ip6_accept_rtadv
- removes RFC1885-related code
From the KAME project via SUZUKI Shinsuke.
Reviewed by core.
|
| 1.33 | 21-Jan-2006 |
rpaulo | branches: 1.33.2; 1.33.4; 1.33.6;
Better support of IPv6 scoped addresses.
- most of the kernel code will not care about the actual encoding of
scope zone IDs and won't touch "s6_addr16[1]" directly.
- similarly, most of the kernel code will not care about link-local
scoped addresses as a special case.
- scope boundary check will be stricter. For example, the current
*BSD code allows a packet with src=::1 and dst=(some global IPv6
address) to be sent outside of the node, if the application do:
s = socket(AF_INET6);
bind(s, "::1");
sendto(s, some_global_IPv6_addr);
This is clearly wrong, since ::1 is only meaningful within a single
node, but the current implementation of the *BSD kernel cannot
reject this attempt.
- and, while there, don't try to remove the ff02::/32 interface route
entry in in6_ifdetach() as it's already gone.
This also includes some level of support for the standard source
address selection algorithm defined in RFC3484, which will be
completed on in the future.
From the KAME project via JINMEI Tatuya.
Approved by core@.
|
| 1.32 | 10-Dec-2005 |
elad | branches: 1.32.2;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.31 | 29-May-2005 |
christos | branches: 1.31.2;
remove stupid hand-rolled loop and kernel conditional.
|
| 1.30 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.29 | 18-Apr-2004 |
matt | De __P()
|
| 1.28 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.27 | 06-Jun-2003 |
itojun | branches: 1.27.2;
separate RFC2292 decls for MLD; sync w/ kame
|
| 1.26 | 06-Jun-2003 |
itojun | - sync up MLD declaration with RFC3542 (s/MLD6/MLD/)
- routing header declaration with RFC3542
(note: sizeof(ip6_rthdr0) has changed!)
also, sync up with RFC2460 routing header definition (no "strict" source
routing mode any more)
part of advanced API update (RFC2292 -> 3542).
|
| 1.25 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.24 | 09-Jun-2002 |
itojun | whitespace
|
| 1.23 | 08-Jun-2002 |
itojun | sync with latest KAME in6_ifaddr/prefix/default router manipulation.
behavior changes:
- two iocts used by ndp(8) are now obsolete (backward compat provided).
use sysctl path instead.
- lo0 does not get ::1 automatically. it will get ::1 when lo0 comes up.
|
| 1.22 | 29-May-2002 |
itojun | move per-interface ip6/icmp6 stat to ifnet->if_afdata. sync w/kame
|
| 1.21 | 21-Dec-2001 |
itojun | branches: 1.21.8;
have packed attribute to protocol structs. sync with kame
|
| 1.20 | 07-Dec-2001 |
itojun | correct timing to increment icmp6 MIB variables. sync with kame
|
| 1.19 | 07-Feb-2001 |
itojun | branches: 1.19.2; 1.19.4;
during ip6/icmp6 inbound packet processing, do not call log() nor printf() in
normal operation (/var can get filled up by flodding bogus packets).
sysctl net.inet6.icmp6.nd6_debug will turn on diagnostic messages.
(#define ND6_DEBUG will turn it on by default)
improve stats in ND6 code.
lots of synchronziation with kame (including comments and cometic ones).
|
| 1.18 | 22-Jan-2001 |
itojun | fix RR result bit in little endian systems. sync with kame
|
| 1.17 | 21-Jan-2001 |
itojun | sync with latest kame.
- make icmp6.h spec conformant to 2292bis-02, regarding to router reumbering
flag bit.
- latest rtadvd.
|
| 1.16 | 09-Dec-2000 |
itojun | update icmp6 too big validation. the change is necessary since pmtud is
mandatory for IPv6 (so we can't just validate by using connected pcb - we need
to allow traffic from unconnected pcb to do pmtud).
- if the traffic is validated by xx_ctlinput, allow up to "hiwat" pmtud
route entries.
- if the traffic was not validated by xx_ctlinput, allow up to "lowat" pmtud
route entries (there's upper limit, so bad guys cannot blow up our routing
table).
sync with kame
XXX need to think again about default hiwat/lowat value.
XXX victim selection to help starvation case
|
| 1.15 | 18-Oct-2000 |
itojun | verify ICMPv6 too big messages based on TCP pcbs, and/or IPsec SA.
TODO: udp6, and sendto consideration. as pmtud is mandatory for IPv6,
it is rather important for us to support those cases.
TODO: more testing
TODO: kame sync
|
| 1.14 | 10-Oct-2000 |
itojun | sync with kame ($KAME$)
|
| 1.13 | 03-Aug-2000 |
itojun | clearifications in icmp6 node query support.
XXX previous commit included "supported qtypes" icmp6 node query support.
sorry commit message was mistaken.
|
| 1.12 | 03-Aug-2000 |
itojun | correct typo in #define. ICMP6_NI_SUCESS -> SUCCESS (notice missing C).
sync with kame.
|
| 1.11 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.10 | 06-Jul-2000 |
itojun | - do not use bitfield for router renumbering header.
- add protection mechanism against ND cache corruption due to bad NUD hints.
- more stats
- icmp6 pps limitation. TOOD: should implement ppsratecheck(9).
|
| 1.9 | 12-Jun-2000 |
itojun | branches: 1.9.2;
better conformance to draft-ietf-ipngwg-icmp-name-lookups-05.
the old code was chimera of 03 and 05 draft.
-n by default, since IPv6 reverse lookup takes too much time.
use -H to enable reverse name lookup.
|
| 1.8 | 09-Mar-2000 |
itojun | branches: 1.8.2;
change member name for icmp6_filter, to be conformant to RFC2292.
From: Francis Dupont
|
| 1.7 | 28-Feb-2000 |
itojun | support draft-ietf-ipngwg-icmp-name-lookups-05.txt, drop support for
draft-ietf-ipngwg-icmp-name-lookups-04.txt.
There are certain bitfield change in 04 draft to 05 draft, which makes
04 "ping6 -a" and 05 "ping6 -a" not interoperable. sigh.
|
| 1.6 | 28-Feb-2000 |
itojun | remove some of cross-BSD portability #ifdef.
remove xxCTL_VARS, which is BSDI specific.
|
| 1.5 | 26-Feb-2000 |
itojun | bring in recent KAME changes (only important and stable ones, as usual).
- remove net.inet6.ip6.nd6_proxyall. introduce proxy NDP code works
just like "arp -s".
- revise source address selection.
be more careful about use of yet-to-be-valid addresses as source.
- as router, transmit ICMP6_DST_UNREACH_BEYONDSCOPE against out-of-scope
packet forwarding attempt.
- path MTU discovery takes care of routing header properly.
- be more strict about mbuf chain parsing.
|
| 1.4 | 06-Feb-2000 |
itojun | to be more rfc2292 complient, move ip6.h and icmp6.h into netinet.
(netinet6/{ip6,icmp6}.h is non-standard path - these files should go away)
it was not possible to use cvsmove in this case.
when you try to look at history, chase it toward netinet6/{ip6,icmp6}.h.
|
| 1.3 | 03-Jul-1999 |
thorpej | branches: 1.3.2;
RCS ID police.
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 28-Jun-1999 |
itojun | branches: 1.1.2;
file icmp6.h was initially added on branch kame.
|
| 1.1.2.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 01-Jul-1999 |
thorpej | file icmp6.h was added on branch chs-ubc2 on 1999-07-01 23:47:00 +0000
|
| 1.3.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.3.2.2 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.3.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.8.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.9.2.4 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.9.2.3 | 04-Aug-2000 |
itojun | pullup (approved by releng-1-5)
sys/netinet6/icmp6.h 1.11 -> 1.13
sys/netinet6/icmp6.c 1.39 -> 1.41
cvs rdiff -r1.11 -r1.12 syssrc/sys/netinet/icmp6.h
cvs rdiff -r1.39 -r1.40 syssrc/sys/netinet6/icmp6.c
correct typo in #define. ICMP6_NI_SUCESS -> SUCCESS (notice missing C).
sync with kame.
cvs rdiff -r1.12 -r1.13 syssrc/sys/netinet/icmp6.h
cvs rdiff -r1.40 -r1.41 syssrc/sys/netinet6/icmp6.c
clearifications in icmp6 node query support.
XXX previous commit included "supported qtypes" icmp6 node query support.
sorry commit message was mistaken.
|
| 1.9.2.2 | 20-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
- do not use bitfield for router renumbering header.
part of sys/netinet/icmp6.h 1.9 -> 1.10
|
| 1.9.2.1 | 20-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
- add protection mechanism against ND cache corruption due to bad NUD hints.
this is part of:
sys/netinet/icmp6.h 1.9 -> 1.10
sys/netinet/tcp_input.c 1.111 -> 1.112
sys/netinet6/icmp6.c 1.34 -> 1.35
sys/netinet6/nd6.c 1.30 -> 1.31
sys/netinet6/nd6.h 1.14 -> 1.15
|
| 1.19.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.19.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.19.2.3 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.19.2.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.19.2.1 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.21.8.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.21.8.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.27.2.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.27.2.4 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.27.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.27.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.27.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.31.2.2 | 21-Jan-2008 |
yamt | sync with head
|
| 1.31.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.32.2.1 | 01-Feb-2006 |
yamt | sync with head.
|
| 1.33.6.1 | 13-Mar-2006 |
yamt | sync with head.
|
| 1.33.4.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.33.2.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.34.2.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.35.50.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.35.46.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.35.42.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.35.36.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.36.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.36.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.36.2.1 | 22-Feb-2008 |
keiichi | imported Mobile IPv6 code developed by the SHISA project
(http://www.mobileip.jp/).
|
| 1.38.10.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.38.8.1 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.38.6.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.38.4.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.38.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.40.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.40.4.1 | 31-May-2011 |
rmind | sync with head
|
| 1.42.4.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.42.4.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.42.4.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.43.4.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.44.2.1 | 25-Jul-2012 |
jdc | Pull up revisions:
src/sys/netinet/icmp6.h revisions 1.45,1.46
(requested by rmind in ticket #434).
this commit contains two sets of unrelated changes:
"while I was here" I checked other KAME implementations for their icmp6.h
version, and thus:
- added a define for MLDV2_LISTENER_REPORT from FreeBSD
- added defines for the missing ICMP6_DST_UNREACH codes
then on to what I actually wanted to do:
- adds strings for the types and codes (encapsulated by ICMP6_STRINGS)
for the use of npfctl and other tools that might want to parse
human-friendly names instead of the corresponding number for ipv6-icmp
types and codes.
The strings are ordered such that their index is (as far as is practical)
the number belonging to the name, which is why there are
icmp6_type_err (use directly) and icmp6_type_info (add 128)
<grmbl>whitespace</grmbl>
|
| 1.46.4.1 | 28-Aug-2013 |
rmind | sync with head
|
| 1.46.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.46.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.47.12.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.47.8.1 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.49.2.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.49.2.2 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.49.2.1 | 15-Mar-2018 |
pgoyette | Synch with HEAD
|
| 1.51.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.51.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.59.10.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.5 | 17-Feb-2021 |
christos | - pass the alignment instead of the mask (as Roy asked and to match the
other macro)
- use alignof to determine that alignment and CTASSERT what we expect
- remove unused macros
|
| 1.4 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.3 | 28-Apr-2008 |
martin | branches: 1.3.4; 1.3.102;
Remove clause 3 and 4 from TNF licenses
|
| 1.2 | 23-Apr-2008 |
thorpej | branches: 1.2.2;
Use <net/net_stats.h> / netstat_sysctl().
|
| 1.1 | 12-Apr-2008 |
thorpej | branches: 1.1.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.1.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.2.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.3.102.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.3.4.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.3.4.1 | 28-Apr-2008 |
mjf | file icmp_private.h was added on branch mjf-devfs2 on 2008-06-02 13:24:23 +0000
|
| 1.32 | 29-Aug-2022 |
knakahara | Add sysctl entry to control to send routing message for RTM_DYNAMIC.
Some routing daemons require such routing message to keep coherency.
If we want to let kernel send such message, set net.inet.icmp.dynamic_rt_msg=1
for IPv4, net.inet6.icmp6.dynamic_rt_msg=1 for IPv6.
Default(=0) is the same as before, that is, not send such routing message.
|
| 1.31 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.30 | 18-Feb-2015 |
christos | branches: 1.30.16; 1.30.18;
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
XXX: pullup-7
|
| 1.29 | 24-Dec-2011 |
christos | branches: 1.29.2; 1.29.6; 1.29.8; 1.29.16; 1.29.22; 1.29.24;
put the histograms last and make them autosize (breaks compat with netstat).
|
| 1.28 | 07-Dec-2009 |
christos | branches: 1.28.12; 1.28.16;
PR/42243: Yasuoka Masahiko: Add "net.inet.icmp.bmcastecho" sysctl support,
to disable icmp replies to the broadcast address.
|
| 1.27 | 12-Apr-2008 |
thorpej | branches: 1.27.4;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.26 | 06-Apr-2008 |
thorpej | Change ICMP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old icmpstat structure; old netstat
binaries will continue to work properly.
|
| 1.25 | 10-Dec-2005 |
elad | branches: 1.25.70;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.24 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.23 | 03-Aug-2004 |
cube | branches: 1.23.12;
Remove a common (icmpstat).
|
| 1.22 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.21 | 30-Jun-2002 |
thorpej | branches: 1.21.6;
Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.20 | 09-Jun-2002 |
itojun | whitespace
|
| 1.19 | 30-Oct-2001 |
kml | branches: 1.19.8;
Add in support for timing out IPv4 routes added due to redirects,
as discussed in tech-net several weeks ago. It turned out that
KAME had already added this functionality to the IPv6 stack, so
I followed their example in adding the sysctl variables
net.inet.icmp.rediraccept and net.inet.icmp.redirtimeout.
|
| 1.18 | 18-Oct-2000 |
itojun | branches: 1.18.2; 1.18.4; 1.18.8;
count successful path MTU changes. good for debugging.
(there could be some discussion on when to increase the counter...)
|
| 1.17 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.16 | 10-Jul-2000 |
itojun | implement net.inet.icmp.errppslimit.
make default value for net.inet.icmp.erratelimit to 0, as < 10ms value
does not do the right thing.
|
| 1.15 | 10-Jun-2000 |
darrenr | branches: 1.15.2;
add icmpreturndatabytes kernel variable (default 8) which specifies the
number of extra data bytes to return in ICMP error messages. This is
also available via sysctl as net.icmp.returndatabytes and is limited to
[8,512].
|
| 1.14 | 15-Feb-2000 |
thorpej | branches: 1.14.2;
Add ICMP error rate limiting, based on the same for ICMP6.
Note, we're reusing the previously unused slot for "MTU discovery" (which
was moved to the "net.inet.ip" branch of the sysctl tree quite some time
ago).
|
| 1.13 | 19-Nov-1999 |
bouyer | Update protocoles and interfaces stats counters to 64bit.
RTM_IFINFO is now 0xf, 0xe is RTM_OIFINFO which returns the old (if_msghdr14)
struct with 32bit counters (binary compat, conditioned on COMPAT_14).
Same for sysctl: node 3 is renamed NET_RT_OIFLIST, NET_RT_IFLIST is now node 4.
Change rt_msg1() to add an mbuf to the mbuf chain instead of just panic()
when the message is larger than MHLEN.
|
| 1.12 | 10-Feb-1998 |
perry | branches: 1.12.14; 1.12.20;
add/cleanup multiple inclusion protection.
|
| 1.11 | 18-Oct-1997 |
kml | remove extraneous icmp_do_mtudisc
|
| 1.10 | 18-Oct-1997 |
kml | change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.9 | 17-Oct-1997 |
kml | Path MTU Discovery support. This is turned off by default.
Use sysctl -w net.inet.icmp.mtudisc=1 to turn on.
Still to come: path removal after some period, black hole detection
|
| 1.8 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.7 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.6 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.5 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.4 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.12.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.12.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.14.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.15.2.1 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.18.8.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.18.4.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.18.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.18.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.18.2.3 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.18.2.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.18.2.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.19.8.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.19.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.21.6.6 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.21.6.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.21.6.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.21.6.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.21.6.2 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.21.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.23.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.25.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.27.4.1 | 11-Mar-2010 |
yamt | sync with head
|
| 1.28.16.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.28.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.29.24.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.29.22.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #537):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
XXX: pullup-7
|
| 1.29.16.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #1258):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
|
| 1.29.8.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #1258):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
|
| 1.29.6.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.29.2.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #1258):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
|
| 1.30.18.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.30.16.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.7 | 16-Mar-1997 |
is | move if_arc.h to sys/net
|
| 1.6 | 15-Mar-1997 |
is | New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.5 | 07-Jun-1995 |
cgd | branches: 1.5.8;
update from Ignatios Souvatzis
|
| 1.4 | 14-Apr-1995 |
chopps | update arc_input() proto to match reality.
|
| 1.3 | 29-Mar-1995 |
briggs | KERNEL -> _KERNEL
|
| 1.2 | 02-Mar-1995 |
chopps | add prototypes
|
| 1.1 | 23-Feb-1995 |
glass | preliminary arcnet support. uses lame but RFC address resolution
|
| 1.5.8.2 | 09-Mar-1997 |
is | netinet/if_ether.h -> netinet/if_inarp.h
|
| 1.5.8.1 | 08-Feb-1997 |
is | Extinguish the link level address from struct arccom, too.
XXX Todo: change this in the hardware driver.
|
| 1.317 | 13-Nov-2024 |
roy | ARP/ND6: Revert prior
Turns out some people actually use this behaviour and strictly speaking
it is allowed by RFC5227 2.4 where it says:
At any time, if a host receives
an ARP packet (Request *or* Reply) where the 'sender IP address' is
(one of) the host's own IP address(es) configured on that interface,
but the 'sender hardware address' does not match any of the host's
own interface addresses, then this is a conflicting ARP packet
The key part is "any of the host's own interface addreses".
|
| 1.316 | 04-Oct-2024 |
roy | ARP: only ignore messages from the receving interface
This means that you can no longer share the same IP addresses
on different interfaces connected to the same subnet.
This was a very non standard solution and you can now
get the same functionality by using lagg(4) instead.
While here, flip log movements sysctl back to 1.
This reverts r1.253 and r1.286.
|
| 1.315 | 09-Sep-2024 |
ozaki-r | arp: allow to send packets without an ARP resolution just after receiving an ARP request
On receiving an ARP request, the current implemention creates an ARP
cache entry but with ND_LLINFO_NOSTATE. Such an entry still needs
an ARP resolution to send back a packet to the requester. The original
behavior before introducing the common ND framework didn't need the
resolution. IPv6 doesn't as well. To restore the original behavior,
make a new ARP cache entry with ND_LLINFO_STALE like IPv6 does.
|
| 1.314 | 20-Aug-2024 |
ozaki-r | arp: fix the behavior on detecting an address duplication without IPv4 DAD
On receiving an ARP request that has the same source protocol address as
the own address, i.e., address duplication, the original behavior of
a kernel prior to supporing IPv4 DAD is to send an ARP reply. It is
the same with a latest kernel with DAD enabled. However, a latest
kernel without DAD sends back an GARP packet. Restore the original
behavior.
|
| 1.313 | 29-Jun-2024 |
riastradh | branches: 1.313.2;
netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.312 | 24-Feb-2024 |
mlelstv | Attribute debug message.
Fixes PR 57959
|
| 1.311 | 15-Nov-2022 |
roy | branches: 1.311.2;
arp: Validate ARP source hardware address matches Ethernet source
RFC 5227 section 1.1 states that for a DaD ARP probe the sender hardware
address must match the hardware address of the interface sending the
packet.
We can now verify this by checking the mbuf tag PACKET_TAG_ETHERNET_SRC.
This fixes an obsure issue where an old router was sending out bogus
ARP probes.
Thanks to Ryo Shimizu <ryo@nerv.org> for the re-implementation.
|
| 1.310 | 15-Nov-2022 |
roy | Revert prior.
|
| 1.309 | 14-Nov-2022 |
roy | arp: Validate L2 sender hardware address matches ARP probe
RFC 5227 section 1.1 states that for a DaD ARP probe the sender hardware
address must match the hardware address of the interface sending the
packet.
We can now verify this by checking the mbuf packet header.
This fixes an obsure issue where an old router was sending out bogus
ARP probes.
|
| 1.308 | 03-Sep-2022 |
thorpej | Convert ARP from a legacy netisr to pktqueue.
|
| 1.307 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.306 | 16-Feb-2021 |
martin | One more time: backout arp header alignment, now that the alignment
asserted has been aligned to reality.
Also remove unused ARP_HDR_ALIGNED_P macro. Pointed out by roy.
|
| 1.305 | 16-Feb-2021 |
martin | Undo previous backout: alignment is needed here.
The reason for the previous backout was a misunderstanding (POINTER_ALIGNED_P
was broken, but the assertion fired even after it got fixed).
|
| 1.304 | 15-Feb-2021 |
christos | Undo previous; POINTER_ALIGNED_P was broken.
|
| 1.303 | 15-Feb-2021 |
christos | put back alignment (reported by martin@)
|
| 1.302 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.301 | 14-Feb-2021 |
roy | if_arp: Just KASSERT that arphrd is aligned
While here improve readability of checking ARP IEEE1394 matches interface.
|
| 1.300 | 13-Feb-2021 |
roy | Prior alignment fixes should not use an offset
|
| 1.299 | 13-Feb-2021 |
roy | if_arp: Ensure that arphdr is aligned
|
| 1.298 | 02-Feb-2021 |
yamt | arp: Plug an mbuf leak
|
| 1.297 | 15-Sep-2020 |
roy | branches: 1.297.2;
Implement RFC 7048, making Neighbor Unreachability Detection less impatient
RFC 7048 Section 3 says in the UNREACHABLE state packets continue to be
sent to the link-layer address and then backoff exponentially.
We adjust this slightly and move to the INCOMPLETE state after
`nd_mmaxtries` probes and then start backing off.
This results in simpler code whilst providing a more robust model which
doubles the time to failure over what we did before.
We don't want to be back to the old ARP model where no unreachability
errors are returned because very few applications would look at
unreachability hints provided such as ND_LLINFO_UNREACHABLE or RTM_MISS.
|
| 1.296 | 14-Sep-2020 |
roy | nd: Name l3addr union of llentry and use in-place of nd_addr.
Probably makes more sense and makes nd.h less messy.
|
| 1.295 | 11-Sep-2020 |
roy | ARP: Use ND rather than our own.
This brings the benefit of Neighbour Unreachability Detection which is
something ARP sorely lacks.
The new timings mirror those of IPv6 and are adjustable via sysctl(8).
Unlike IPv6 ND, these are global and not per interface.
|
| 1.294 | 09-Mar-2020 |
roy | route: RTM_MISS now puts the message source address in RTA_AUTHOR
route(8) also reports this.
A userland app could use this to blacklist nodes who probe for machines
that doesn't exist on a subnet / prefix.
|
| 1.293 | 09-Mar-2020 |
roy | arp: report RTM_MISS when removing an unresolved entry in the arp table
Otherwise we only get it when renewing and we've sent too many requests.
This mirrors INET6 behaviour.
|
| 1.292 | 23-Jan-2020 |
roy | arp: find source address then target address when processing input
This fixes the case where another host having a duplicate ip address
starts using it right away without probing for it's availability.
While here, prefer ifatoia over a strict cast.
|
| 1.291 | 20-Jan-2020 |
thorpej | Remove FDDI support.
|
| 1.290 | 19-Jan-2020 |
thorpej | Remove Token Ring support.
|
| 1.289 | 11-Oct-2019 |
roy | branches: 1.289.2;
ARP: Don't defend ARP probes.
We should let the nature of ARP takes it's course here when our address
is neither tentative nor duplicated.
This allows the host to work with ARP ping, which was broken in r1.279.
|
| 1.288 | 25-Sep-2019 |
ozaki-r | Initialize DAD components properly
The original code initialized each component in non-init functions such as
arp_dad_start and nd6_dad_find, conditionally based on a global flag for each.
However, it was racy because the flag and the code around it were not
protected by a lock and could cause a kernel panic at worst.
Fix the issue by initializing the components in bootup as usual.
|
| 1.287 | 01-Sep-2019 |
roy | inet: Send RTM_MISS when we fail to resolve an address.
Takes the same approach as when adding a new address - we no longer
announce the new lladdr right away but we announce the result.
This will either be RTM_ADD or RTM_MISS.
RTM_DELETE is only sent if we have a lladdr assigned OR gc'ed.
This tells us when a new lladdr has been added (RTM_ADD),
changed (RTM_CHANGE), deleted (RTM_DELETED) or has failed to been
resolved (RTM_MISS). The latter case can be interpreted as unreachable.
|
| 1.286 | 30-Aug-2019 |
roy | ARP: change default sysctl entry log_movements to 0
IP address sharing is a thing and shouldn't cause needless diagnostics
by default.
|
| 1.285 | 30-Aug-2019 |
roy | ARP: remove unused sysctl entry log_unknown_network
|
| 1.284 | 22-Aug-2019 |
roy | rtsock: rework rt_clonedmsg to take a message type and lladdr
We will use this in a future patch to notify userland of lladdr
changes.
XXX pullup -8 -9
|
| 1.283 | 19-Aug-2019 |
ozaki-r | Initialize dom_mowner for MBUFTRACE
|
| 1.282 | 29-Apr-2019 |
roy | branches: 1.282.2;
Introduce rt_addrmsg_src which adds RTA_AUTHOR to the message.
Use this when we notify userland of a duplicate address
and set RTA_AUTHOR to the hardware address of the sender.
While here, match the logging diagnostic of INET6 to the simpler one
of INET so it's consistent.
|
| 1.281 | 29-Apr-2019 |
roy | Move lla_snprintf from if_arp.c to dl_print.c
|
| 1.280 | 29-Apr-2019 |
roy | rtsock: Route address message simplification
Rename rt_newaddrmsg to rt_addrmsg_rt.
Add rt_addrmsg which drops the error and route arguments which are only
needed by one caller.
|
| 1.279 | 24-Apr-2019 |
roy | ARP: Enable Address Defence again.
Revert the tentative/duplicated check and test for if it's been broadcast
or not. This reverts r1.245.
|
| 1.278 | 22-Dec-2018 |
maxv | Replace M_ALIGN and MH_ALIGN by m_align.
|
| 1.277 | 29-Nov-2018 |
ozaki-r | Introduce and use ip_dad_enabled() and ip6_dad_enabled() functions
|
| 1.276 | 30-Oct-2018 |
ozaki-r | Avoid double rt_replace_ifa on rtrequest1(RTM_ADD)
Some callers of rtrequest1(RTM_ADD) adjust rt_ifa of an rtentry created by
rtrequest1 that may change rt_ifa (in ifa_rtrequest) with another ifa that is
different from requested one. It's wasteful and even worse introduces a race
condition. rtrequest1 should just use a passed ifa as is if a caller hopes so.
|
| 1.275 | 11-May-2018 |
maxv | branches: 1.275.2;
static
|
| 1.274 | 01-May-2018 |
maxv | Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.273 | 11-Apr-2018 |
maxv | Add XXX.
|
| 1.272 | 10-Apr-2018 |
maxv | Remove unused mbuf argument from arpcreate() and arplookup().
|
| 1.271 | 08-Mar-2018 |
ozaki-r | Fix a race condition on DAD destructions (again)
The previous fix to DAD timers was wrong; it avoided a use-after-free but
instead introduced a memory leak. The destruction method had delegated
a destruction of a DAD timer to the timer itself and told that by setting NULL
to dp->dad_ifa. However, the previous fix made DAD timers do nothing on
the sign.
Fixing the issue with using callout_stop isn't easy. One approach is to have
a refcount on dp but it introduces extra complexity that we want to avoid.
The new fix falls back to using callout_halt, which was abandoned because of
softnet_lock. Fortunately now the network stack is protected by KERNEL_LOCK
so we can remove softnet_lock from DAD timers (callout) and use callout_halt
safely.
|
| 1.270 | 06-Mar-2018 |
ozaki-r | Fix reference leaks of llentry
callout_reset and callout_halt can cancel a pending callout without telling us.
Detect a cancel and remove a reference by using callout_pending and
callout_stop (it's a bit tricy though, we can detect it).
While here, we can remove remaining abuses of mutex_owned for softnet_lock.
|
| 1.269 | 06-Mar-2018 |
ozaki-r | Tweak sanity checks
Scheduling a timer of static entries is wrong.
|
| 1.268 | 01-Mar-2018 |
roy | branches: 1.268.2;
Sprinkle some const.
|
| 1.267 | 24-Feb-2018 |
ozaki-r | Avoid a race condition of DAD timer destructions
When we see dp->dad_ifa == NULL, it means that the ifa is being deleted and also
the callout is scheduled again by someone. We shouldn't rely on a result of
callout_pending to know if the callout is scheduled because it returns false if
the subsequent callout handler is already on the fly.
We have to always delegate the destruction of dp to the subsequent handler
unconditionally if dp->dad_ifa == NULL. Otherwise, the first handler destroys
the dp and the second handler tries to handle destroyed dp.
|
| 1.266 | 14-Feb-2018 |
maxv | Remove IFF_STATICARP, we don't support this, and the code is useless in its
current form.
ok ozaki-r@
|
| 1.265 | 13-Feb-2018 |
maxv | Mmh. Add a missing check: if ARP was disabled on the interface, don't
process ARP packets. Otherwise the kernel will add ARP entries even if
ifconfig wm0 -arp
was entered.
|
| 1.264 | 13-Feb-2018 |
maxv | Be tougher:
* In arpintr(), don't allow IEEE1394 packets on non-IEEE1394 interfaces.
* In revarpinput(), kick IEEE1394 packets right away. They are not
supported.
|
| 1.263 | 13-Feb-2018 |
maxv | Same change as rev1.258, but this time in revarpinput: use m_pullup.
|
| 1.262 | 13-Feb-2018 |
maxv | Minor stylistic changes, and use C99 types.
|
| 1.261 | 13-Feb-2018 |
maxv | Replace dead code by KASSERT.
|
| 1.260 | 13-Feb-2018 |
maxv | Don't force ARPHRD_IEEE1394 on IEEE1394 interfaces. If it's not there, then
kick the packet. And do this earlier.
|
| 1.259 | 13-Feb-2018 |
maxv | Use only one label, clearer.
|
| 1.258 | 13-Feb-2018 |
maxv | Fix three things in arpintr():
* mtod can't return NULL.
* It is wrong to kick the packet if m->m_len < arplen. While this check
always returns false for native Ethernet interfaces, it may not if the
frame is encapsulated in EtherIP/L2TP. Use m_pullup instead.
* Remove XXX, it is fine. Reduce the indentation level afterwards.
|
| 1.257 | 13-Feb-2018 |
maxv | Style, no functional change.
|
| 1.256 | 16-Jan-2018 |
ozaki-r | Make DAD destructions (MP-)safe with callout_stop
arp_dad_stoptimer and nd6_dad_stoptimer can be called with or without
softnet_lock held and unfortunately we have no easy way to statically know which.
So it is hard to use callout_halt there.
To address the situation, we use callout_stop to make the code safe. The new
approach copes with the issue by delegating the destruction of a callout to
callout itself, which allows us to not wait the callout to finish. This can be
done thanks to that DAD objects are separated from other data such as ifa.
The approach is suggested by riastradh@
Proposed on tech-kern@ and tech-net@
|
| 1.255 | 17-Nov-2017 |
ozaki-r | Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
|
| 1.254 | 10-Nov-2017 |
ozaki-r | Fix a deadlock between a route update and lltable
It happens because rtalloc1 is called from lltable with holding
IF_AFDATA_WLOCK.
If a route update is in action, rtalloc1 would wait for its completion with
holding IF_AFDATA_WLOCK. At the same moment, a softint (e.g., arpintr) may try
to take IF_AFDATA_WLOCK and get stuck on it. Unfortunately the stuck softint
prevents the route update from progressing because the route update calls
psref_target_destroy that needs the softint to complete.
A resource allocation graph of the senario looks like this:
route update =(psref_target_destroy)=> softint => IF_AFDATA_WLOCK
=(rt_update_wait)=> route update
Fix the deadlock by pulling rtalloc1 out of the lltable codes inside
IF_AFDATA_WLOCK.
Note that the deadlock happens only if NET_MPSAFE is enabled.
|
| 1.253 | 27-Jun-2017 |
roy | Use if_get_bylla() instead of just looking at the lla of the interface
the address belongs to.
This allows any ARP message we receieved from another interface to
be correctly dropped.
While here, move the protocol length check higher up the food chain.
|
| 1.252 | 21-Jun-2017 |
ozaki-r | Don't create a permanent L2 cache entry on adding an address to an interface
It was created to copy FreeBSD, however actually the cache isn't
necessary. Remove it to simplify the code and reduce the cost to
maintain it (e.g., keep a consistency with a corresponding local
route).
|
| 1.251 | 16-Jun-2017 |
ozaki-r | Sending a routing message (RTM_ADD) on adding an llentry
A message used to be sent on adding a cloned route. Restore the
behavior for backward compatibility.
Requested by ryo@
|
| 1.250 | 18-May-2017 |
ozaki-r | branches: 1.250.2;
Lookup caches from a CARP interface if the packet is routed to the interface
This fixes CARP setups without carpdev (physical interface) having any IPs.
|
| 1.249 | 12-May-2017 |
ryo | replace in_fmtaddr() by IN_PRINT(), and delete function in_fmtaddr()
|
| 1.248 | 04-Apr-2017 |
ozaki-r | branches: 1.248.4;
Get rid of unused macros
|
| 1.247 | 17-Mar-2017 |
roy | If we're not doing DAD, don't set IN_IFF_TENTATIVE.
|
| 1.246 | 10-Mar-2017 |
roy | If an ARP packet is received to the null host (0.0.0.0) then look for
an address matching the sender IP address on the interface.
This allows DAD to fail during the probe phase when a reverse ARP
proxy is present.
|
| 1.245 | 09-Mar-2017 |
roy | Only check target address collision if the sender address is the null
address (ie a DAD probe) or our matching address is either TENTATIVE
or DUPLICATED.
|
| 1.244 | 24-Feb-2017 |
roy | Only do DaD if the interface actually has the address.
|
| 1.243 | 21-Feb-2017 |
ozaki-r | Replace malloc for DAD with kmem and move them out of the lock for DAD
|
| 1.242 | 11-Feb-2017 |
roy | Allow Unicast Poll from RFC 1122 to bypass DaD checking.
|
| 1.241 | 07-Feb-2017 |
ozaki-r | Add missing NULL checks for m_get_rcvif
|
| 1.240 | 24-Jan-2017 |
ozaki-r | Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.239 | 21-Jan-2017 |
maxv | Add some checks, mostly same as in_arpinput.
|
| 1.238 | 20-Jan-2017 |
maxv | Make sure the protocol address length equals that of IPv4. Also, make sure
the hardware address length equals that of the interface we received the
packet on. Otherwise a packet could easily set them both to zero and make
the kernel read beyond the allocated mbuf, which is terrible.
Note: for the latter we drop the packet instead of replying, since it is
malformed.
Note: I also added an ugly hack in CARP, since it apparently expects at
least six bytes.
|
| 1.237 | 20-Jan-2017 |
maxv | Style
|
| 1.236 | 20-Jan-2017 |
maxv | Reput a nullcheck that was mistakenly removed in rev1.204. ar_hrd is
packet-controlled.
|
| 1.235 | 16-Jan-2017 |
christos | rename arplog -> ARPLOG to make it clear that it is a macro and tuck-in the
buffer used for address formatting.
|
| 1.234 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.233 | 12-Dec-2016 |
ozaki-r | branches: 1.233.2;
Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.232 | 05-Nov-2016 |
roy | Don't handle ARP duplication for the unspecified address.
|
| 1.231 | 18-Oct-2016 |
ozaki-r | Don't hold global locks if NET_MPSAFE is enabled
If NET_MPSAFE is enabled, don't hold KERNEL_LOCK and softnet_lock in
part of the network stack such as IP forwarding paths. The aim of the
change is to make it easy to test the network stack without the locks
and reduce our local diffs.
By default (i.e., if NET_MPSAFE isn't enabled), the locks are held
as they used to be.
Reviewed by knakahara@
|
| 1.230 | 11-Oct-2016 |
roy | Implement RFC 5227 2.4 Ongoing Conflict Detection and Address Defence.
If ip_dad_count is 0, then the conflict is just logged and the address
is not marked as duplicated.
|
| 1.229 | 11-Oct-2016 |
roy | Mark arprequest static and introduce arpannounce so that gratuitous
ARP requests are only send from valid addresses.
|
| 1.228 | 03-Oct-2016 |
ozaki-r | Fix race condition on ifqueue used by traditional netisr
If a underlying network device driver supports MSI/MSI-X, RX interrupts
can be delivered to arbitrary CPUs. This means that Layer 2 subroutines
such as ether_input (softint) and subsequent Layer 3 subroutines (softint)
which are called via traditional netisr can be dispatched on an arbitrary
CPU. Layer 2 subroutines now run without any locks (expected) and so a
Layer 2 subroutine and a Layer 3 subroutine can run in parallel.
There is a shared data between a Layer 2 routine and a Layer 3 routine,
that is ifqueue and IF_ENQUEUE (from L2) and IF_DEQUEUE (from L3) on it
are racy now.
To fix the race condition, use ifqueue#ifq_lock to protect ifqueue
instead of splnet that is meaningless now.
The same race condition exists in route_intr. Fix it as well.
Reviewed by knakahara@
|
| 1.227 | 18-Sep-2016 |
christos | Dealing with arplog is a bit more complicated...
|
| 1.226 | 16-Sep-2016 |
roy | Clear IN_IFF_TENTATIVE when stopping DaD here.
|
| 1.225 | 16-Sep-2016 |
roy | Don't setup DaD for INADDR_ANY
|
| 1.224 | 15-Sep-2016 |
roy | Allow arplog to be used outside of if_arp.c
|
| 1.223 | 07-Sep-2016 |
roy | Refine arplog to be like nd6log.
|
| 1.222 | 08-Aug-2016 |
ozaki-r | Restore ARP_STAT_DFRTOTAL deleted unexpectedly
|
| 1.221 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.220 | 28-Jul-2016 |
ozaki-r | Fix panic on adding/deleting IP addresses under network load
Adding and deleting IP addresses aren't serialized with other network
opeartions, e.g., forwarding packets. So if we add or delete an IP
address under network load, a kernel panic may happen on manipulating
network-related shared objects such as rtentry and rtcache.
To avoid such panicks, we still need to hold softnet_lock in in_control
and in6_control that are called via ioctl and do network-related operations
including IP address additions/deletions.
Fix PR kern/51356
|
| 1.219 | 25-Jul-2016 |
ozaki-r | Make DAD of ARP/NDP MP-safe with coarse-grained locks
The change also prevents arp_dad_timer/nd6_dad_timer from running if
arp_dad_stop/nd6_dad_stop is called, which makes sure that callout_reset
won't be called during callout_halt.
|
| 1.218 | 25-Jul-2016 |
ozaki-r | Use KASSERT for checking non-NULL of ifa->ifa_ifp
ifa->ifa_ifp should be always non-NULL, so doing the check only if
DIAGNOSTIC is ok.
|
| 1.217 | 08-Jul-2016 |
ozaki-r | branches: 1.217.2;
Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.216 | 06-Jul-2016 |
ozaki-r | Tweak indentation
|
| 1.215 | 01-Jul-2016 |
ozaki-r | Make sure to free all interface addresses in if_detach
Addresses of an interface (struct ifaddr) have a (reverse) pointer of an
interface object (ifa->ifa_ifp). If the addresses are surely freed when
their interface is destroyed, the pointer is always valid and we don't
need a tweak of replacing the pointer to if_index like mbuf.
In order to make sure the assumption, the following changes are required:
- Deactivate the interface at the firstish of if_detach. This prevents
in6_unlink_ifa from saving multicast addresses (wrongly)
- Invalidate rtcache(s) and clear a rtentry referencing an address on
RTM_DELETE. rtcache(s) may delay freeing an address
- Replace callout_stop with callout_halt of DAD timers to ensure stopping
such timers in if_detach
|
| 1.214 | 30-Jun-2016 |
ozaki-r | Make sure that ifaddr is published after its initialization finished
Basically we should insert an item to a collection (say a list) after
item's initialization has been completed to avoid accessing an item
that is initialized halfway. ifaddr (in{,6}_ifaddr) isn't processed
like so and needs to be fixed.
In order to do so, we need to tweak {arp,nd6}_rtrequest that depend
on that an ifaddr is inserted during its initialization; they explore
interface's address list to determine that rt_getkey(rt) of a given
rtentry is in the list to know whether the route's interface should
be a loopback, which doesn't work after the change. To make it work,
first check RTF_LOCAL flag that is set in rt_ifa_addlocal that calls
{arp,nd6}_rtrequest eventually. Note that we still need the original
code for the case to remove and re-add a local interface route.
|
| 1.213 | 28-Jun-2016 |
ozaki-r | Add missing NULL checks for m_get_rcvif_psref
|
| 1.212 | 20-Jun-2016 |
knakahara | apply if_output_lock() to L3 callers which call ifp->if_output() of L2(or L3 tunneling).
|
| 1.211 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.210 | 17-May-2016 |
ozaki-r | Get rid of unnecessary assignment
|
| 1.209 | 25-Apr-2016 |
ozaki-r | Check error of rt_setgate and rt_settag
|
| 1.208 | 19-Apr-2016 |
ozaki-r | Constify rtentry of arpresolve
We don't need to (rather shouldn't) modify rtentry in there.
|
| 1.207 | 18-Apr-2016 |
ozaki-r | Fix panic on receiving an ARP request
The panic happened if an ARP request has a spa (i.e., IP address) whose
ARP entry already exists in the table as a static ARP entry.
|
| 1.206 | 13-Apr-2016 |
ozaki-r | ddb: rename show arptab to show routes
show arptab command of ddb is now inappropriate because it actually dumps
routes but arp entries aren't routes anymore. So rename it to show routes
and move the code from if_arp.c to route.c.
ok christos@
|
| 1.205 | 07-Apr-2016 |
christos | - tidy up error messages
- add a length argument to arpresolve()
- add KASSERT for overflow
|
| 1.204 | 04-Apr-2016 |
ozaki-r | Separate nexthop caches from the routing table
By this change, nexthop caches (IP-MAC address pair) are not stored
in the routing table anymore. Instead nexthop caches are stored in
each network interface; we already have lltable/llentry data structure
for this purpose. This change also obsoletes the concept of cloning/cloned
routes. Cloned routes no longer exist while cloning routes still exist
with renamed to connected routes.
Noticeable changes are:
- Nexthop caches aren't listed in route show/netstat -r
- sysctl(NET_RT_DUMP) doesn't return them
- If RTF_LLDATA is specified, it returns nexthop caches
- Several definitions of routing flags and messages are removed
- RTF_CLONING, RTF_XRESOLVE, RTF_LLINFO, RTF_CLONED and RTM_RESOLVE
- RTF_CONNECTED is added
- It has the same value of RTF_CLONING for backward compatibility
- route's -xresolve, -[no]cloned and -llinfo options are removed
- -[no]cloning remains because it seems there are users
- -[no]connected is introduced and recommended
to be used instead of -[no]cloning
- route show/netstat -r drops some flags
- 'L' and 'c' are not seen anymore
- 'C' now indicates a connected route
- Gateway value of a route of an interface address is now not
a L2 address but "link#N" like a connected (cloning) route
- Proxy ARP: "arp -s ... pub" doesn't create a route
You can know details of behavior changes by seeing diffs under tests/.
Proposed on tech-net and tech-kern:
http://mail-index.netbsd.org/tech-net/2016/03/11/msg005701.html
|
| 1.203 | 25-Jan-2016 |
ozaki-r | Remove unnecessary LLE_REMREF
The code around it was copied from arptimer, but LLE_REMREF
is unnecessary because it is needed only for arptimer that
is called after LLE_ADDREF.
This is a possible fix for PR#50548, PR#50702 and PR#50704.
|
| 1.202 | 21-Jan-2016 |
riastradh | Revert previous: ran cvs commit when I meant cvs diff. Sorry!
Hit up-arrow one too few times.
|
| 1.201 | 21-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.200 | 20-Jan-2016 |
riastradh | Eliminate struct protosw::pr_output.
You can't use this unless you know what it is a priori: the formal
prototype is variadic, and the different instances (e.g., ip_output,
route_output) have different real prototypes.
Convert the only user of it, raw_send in net/raw_cb.c, to take an
explicit callback argument. Convert the only instances of it,
route_output and key_output, to such explicit callbacks for raw_send.
Use assertions to make sure the conversion to explicit callbacks is
warranted.
Discussed on tech-net with no objections:
https://mail-index.netbsd.org/tech-net/2016/01/16/msg005484.html
|
| 1.199 | 05-Jan-2016 |
ozaki-r | Make revarprequest static
|
| 1.198 | 17-Dec-2015 |
ozaki-r | Fix memory leak of llentry#la_opaque
llentry#la_opaque which is for token ring is allocated in arp.c
and freed in arp.c when freeing llentry. However, llentry can be
freed from other places, e.g., lltable_free. In such cases,
la_opaque is never freed.
To fix that, add a new callback (lle_ll_free) to llentry and
register a destruction function of la_opque to it. On freeing a
llentry, we can surely free la_opque via the callback.
|
| 1.197 | 16-Dec-2015 |
ozaki-r | Fix token_rif extractions from llentry
|
| 1.196 | 11-Dec-2015 |
ozaki-r | Introduce arp_settimer
No functional change.
|
| 1.195 | 30-Nov-2015 |
ozaki-r | Get rid of a big block in in_arpinput
No functional change.
|
| 1.194 | 19-Nov-2015 |
ozaki-r | Restore softnet_lock and KERNEL_LOCK for rtrequest and rtfree
We still need them for rt operations.
|
| 1.193 | 06-Nov-2015 |
ozaki-r | Fix db_print_llinfo
rt_llinfo is now struct llentry.
|
| 1.192 | 06-Nov-2015 |
ozaki-r | Fix inappropriate rt_flags check
It depended on either RTF_CLONED or RTF_CLONING must be set, however,
the assumption didn't meet for userland problems that create a route
via RTM_ADD.
This fixes an issue that running rarpd causes the following kernel panic
reported by nonaka@:
panic: kernel diagnostic assertion "(la->la_flags & LLE_STATIC) == 0"
failed: file "/usr/src/sys/netinet/if_arp.c", line 1339
|
| 1.191 | 20-Oct-2015 |
ozaki-r | Stop callout in arp_rtrequest(RTM_DELETE)
This change fixes arptimer panic after removing an interface
(say by drvctl -d), which is reported by Takahiro Hayashi.
This change also fixes llentry's reference counting; we have
to take into account rtentry#rt_llinfo as well as arptimer.
|
| 1.190 | 20-Oct-2015 |
ozaki-r | Stop using softnet_lock (fix possible deadlock)
Using softnet_lock for mutual exclusion between lltable_free and
arptimer was wrong and had an issue causing a deadlock between
them; lltable_free waits arptimer completion by calling
callout_halt with softnet_lock that is held in arptimer, however
lltable_free also holds llentry's lock that is also held in
arptimer so arptimer never obtain the lock and both never go
forward eventually. We have to pass llentry's lock to
callout_halt instead.
|
| 1.189 | 14-Oct-2015 |
roy | In the event of an error within arpresolve(), delete the cloned route
otherwise it would never be deleted.
|
| 1.188 | 14-Oct-2015 |
roy | Save and clear the la route while we have a write lock
|
| 1.187 | 13-Oct-2015 |
roy | arpresolve() now returns 0 on success otherwise an error code.
Callers of arpresolve() now pass the error code back to their caller,
masking out EWOULDBLOCK.
This allows applications such as ping(8) to display a suitable error
condition.
|
| 1.186 | 13-Oct-2015 |
roy | Move the NOARP check up a bit so that it works when an la is created
but hasn't been resolved yet.
Fixes PR kern/17611.
|
| 1.185 | 13-Oct-2015 |
roy | Simplify la handling in arpresolve() by asking arplookup() not to create
a la. If a la is needed arpresolve() will then create it or mark the
current la as writable.
|
| 1.184 | 08-Oct-2015 |
roy | Create a temporary define involving IFF_STATICARP if we have it
instead of just testing for __FreeBSD__.
No functional change.
ok: ozaki-r@
|
| 1.183 | 07-Oct-2015 |
ozaki-r | Create an llentry after fixing an interface to store
In case of RTF_LOCAL routes, we change an output interface
of a route from original one to lo0ifp. An llentry also
has to be stored to lo0ifp in such cases.
Problem reported by roy@
|
| 1.182 | 05-Oct-2015 |
ozaki-r | Fix arplookup logic
It should first lookup and then create an entry if not found (and if
creation is requested).
|
| 1.181 | 11-Sep-2015 |
roy | If, for whatever reason, a local interface route is removed and then
re-added, mark it as a local route.
While here, if changing the route to go via the loopback interface
remove any inherited MTU value.
|
| 1.180 | 09-Sep-2015 |
ozaki-r | Remove wrong KASSERT in arptfree
la_rt can be NULL because arptimer that calls arptfree doesn't always
free llentry so llentry can remain with la_rt == NULL. So we instead
check whether la_rt is NULL or not and do arptfree if not.
This fixes PR kern/50184 (confirmed by martin@) and
PR kern/50186 (maybe).
|
| 1.179 | 09-Sep-2015 |
ozaki-r | Revert v1.176 for further proper fix
|
| 1.178 | 07-Sep-2015 |
ozaki-r | CID 1322880: remove unnecessary m != NULL checks
|
| 1.177 | 07-Sep-2015 |
ozaki-r | CID 1322878: simplify log output flow
|
| 1.176 | 02-Sep-2015 |
christos | XXX: Disable KASSERT for now since locking is broken for interface removals.
|
| 1.175 | 31-Aug-2015 |
ozaki-r | Remove obsolete global variables and sysctl MIBs
|
| 1.174 | 31-Aug-2015 |
ozaki-r | Replace ARP cache (llinfo) with lltable/llentry
Highlights of the change are:
- Use llentry instead of llinfo to manage ARP caches
- ARP specific data are stored in the hashed list
of an interface instead of the global list (llinfo_arp)
- Fine-grain locking on llentry
- arptimer (callout) per ARP cache
- the global timer callout with the big locks can be
removed (though softnet_lock is still required for now)
- net.inet.arp.prune is now obsoleted
- it was the interval of the global timer callout
- net.inet.arp.refresh is now obsoleted
- it was a parameter that prevents expiration of active caches
- Removed to simplify the timer logic, but we may be able to
restore the feature if really needed
Proposed on tech-kern and tech-net.
|
| 1.173 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.172 | 12-Aug-2015 |
ozaki-r | Move insane goto label
|
| 1.171 | 07-Aug-2015 |
ozaki-r | Use time_uptime instead of time_second to avoid time leaps
Some codes in sys/net* use time_second to manage time periods such as
cache expirations. However, time_second doesn't increase monotonically
and can leap by say settimeofday(2) according to time_second(9). We
should use time_uptime instead of it to avoid such time leaps.
This change replaces time_second with time_uptime. Additionally it
converts a time based on time_uptime to a time based on time_second
when the kernel passes the time to userland programs that expect
the latter, and vice versa.
Note that we shouldn't leak time_uptime to other hosts over the
netowrk. My investigation shows there is no such leak:
http://mail-index.netbsd.org/tech-net/2015/08/06/msg005332.html
Discussed on tech-kern and tech-net.
|
| 1.170 | 15-Jul-2015 |
ozaki-r | Make global variables static
|
| 1.169 | 22-May-2015 |
ozaki-r | Replace NARC with NARCNET to follow renaming at 2007
Hmm, is anyone using this?
|
| 1.168 | 21-May-2015 |
ozaki-r | Use LIST_FOREACH{,_SAFE}
The first loop doesn't remove any items in it, so we can use
LIST_FOREACH instead of LIST_FOREACH_SAFE.
|
| 1.167 | 21-May-2015 |
ozaki-r | Use NULL instead of 0 for pointers
|
| 1.166 | 21-May-2015 |
ozaki-r | Make arp_init, in_revarpinput and revarprequest static
|
| 1.165 | 16-May-2015 |
roy | Separate ARP handling DAD from inet.
This is done by signalling the intent to try tentative addresses
and then clearing the intent once the address is setup.
When the ARP handler is installed (arp_ifinit) then it adds
dad start and stop functions to the address which are used instead
of calling ARP directly.
|
| 1.164 | 03-May-2015 |
justin | Rename delay variable as it shadows a global on arm.
|
| 1.163 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.162 | 23-Mar-2015 |
roy | Add RTF_BROADCAST to mark routes used for the broadcast address when
they are created on the fly. This makes it clear what the route is for
and allows an optimisation in ip_output() by avoiding a call to
in_broadcast() because most of the time we do talk to a host.
It also avoids a needless allocation for the storage of llinfo_arp and
thus vanishes from arp(8) - it showed as incomplete anyway so this
is a nice side effect.
Guard against this and routes marked with RTF_BLACKHOLE in
ip_fastforward().
While here, guard against routes marked with RTF_BLACKHOLE in
ip6_fastforward().
RTF_BROADCAST is IPv4 only, so don't bother checking that here.
|
| 1.161 | 26-Feb-2015 |
roy | Introduce the routing flag RTF_LOCAL to track local address routes.
Add functions rt_ifa_addlocal() and rt_ifa_remlocal() to add and remove
local routes for the address and announce the new address and route
to the routing socket.
Add in_ifaddlocal() and in_ifremlocal() to use these functions.
Rename in6_if{add,rem}loop() to in6_if{add,rem}local() and use these
functions.
rtinit() no longer announces the address, just the network route for the
address. As such, calls to rt_newaddrmsg() have been removed from
in_addprefix() and in_scrubprefix().
This solves the problem of potentially more than one announcement, or no
announcement at all for the address in certain situations.
|
| 1.160 | 13-Nov-2014 |
christos | branches: 1.160.2;
Add sysctl to selectively log arp packets from unknown network. (Adrien URBAN).
|
| 1.159 | 05-Sep-2014 |
matt | Deanonymize structure for llinfo_arp.
|
| 1.158 | 03-Jun-2014 |
ozaki-r | branches: 1.158.2; 1.158.4;
Call ifp->if_output in revarprequest with KERNEL_LOCK held
Otherwise, it hits KASSERT(KERNEL_LOCKED_P()) in ether_output
when nfs_boot fails and tries RARP.
|
| 1.157 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.156 | 12-Apr-2014 |
gdt | revarprequest: Avoid leaking mbuf.
In revarprequest, an mbuf could perhaps be leaked in an error path.
My reading of the code is that this is not possible, because ar_pro is
set to ETHERNET_IP, and ar_tha can only be null in the 1394 case.
But, better to have the free call anyway; ar_tha does not have a
documented interface contract :-)
Pointed out by Maxime Villard.
|
| 1.155 | 25-Feb-2014 |
pooka | branches: 1.155.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.154 | 02-Jan-2012 |
liamjfoy | branches: 1.154.2; 1.154.6; 1.154.8; 1.154.10; 1.154.16;
Remove dead variable
|
| 1.153 | 31-Dec-2011 |
christos | - fix offsetof usage, and redundant defines
- kill pointer casts to 0
|
| 1.152 | 27-Aug-2011 |
christos | branches: 1.152.2; 1.152.6;
Add 3 logging sysctls for arp from freebsd:
1. log_movements: do you want to log the arp overwritten message or not?
2. log_wrong_iface: do you want to log when an arp arrives at the wrong
interface?
3. log_permanent_modify: do you want to log when an arp message attempts
to overwrite a static entry?
I did not call the sysctls log_arp like FreeBSD does, because we already
have an arp sysctl level. The default is on for all three of them.
|
| 1.151 | 03-May-2011 |
dyoung | arp_drain() may be called with locks held, so instead of doing any work
in arp_drain(), set a drain-needed flag. Do the work in the fasttimo
handler.
Contributed by Coyote Point Systems, Inc.
|
| 1.150 | 01-Feb-2011 |
matt | Add a new AF/PF_ROUTE which is 64-bit clean which makes the routing socket
interface (and its associated sysctls) act identically for both 32 and 64 bit
programs. The old unclean one remains for backward compatibility.
|
| 1.149 | 20-Nov-2009 |
christos | branches: 1.149.4; 1.149.6; 1.149.8;
ar_tha() can return NULL; treat this as an error.
|
| 1.148 | 03-Nov-2009 |
christos | Handle RFC 5227 ARP probes properly, don't drop 0.0.0.0 source packets
silently. (Patrik Lahti <plahti at qnx dot com>)
|
| 1.147 | 16-Sep-2009 |
pooka | Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.146 | 12-Aug-2009 |
dyoung | Don't require the gateway address to have room for both an interface
name and address. Room for an address will do. This should fix
a regression in 'arp -s ...' on interfaces such as xennet0 with
unusually long names.
I will request a pull-up to netbsd-5.
|
| 1.145 | 11-Jan-2009 |
christos | merge christos-time_t
|
| 1.144 | 07-Nov-2008 |
dyoung | *** Summary ***
When a link-layer address changes (e.g., ifconfig ex0 link
02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor
Advertisement to update the network-/link-layer address bindings
on our LAN peers.
Refuse a change of ethernet address to the address 00:00:00:00:00:00
or to any multicast/broadcast address. (Thanks matt@.)
Reorder ifnet ioctl operations so that driver ioctls may inherit
the functions of their "class"---ether_ioctl(), fddi_ioctl(), et
cetera---and the class ioctls may inherit from the generic ioctl,
ifioctl_common(), but both driver- and class-ioctls may override
the generic behavior. Make network drivers share more code.
Distinguish a "factory" link-layer address from others for the
purposes of both protecting that address from deletion and computing
EUI64.
Return consistent, appropriate error codes from network drivers.
Improve readability. KNF.
*** Details ***
In if_attach(), always initialize the interface ioctl routine,
ifnet->if_ioctl, if the driver has not already initialized it.
Delete if_ioctl == NULL tests everywhere else, because it cannot
happen.
In the ioctl routines of network interfaces, inherit common ioctl
behaviors by calling either ifioctl_common() or whichever ioctl
routine is appropriate for the class of interface---e.g., ether_ioctl()
for ethernets.
Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In
the user->kernel interface, SIOCSIFADDR's argument was an ifreq,
but on the protocol->ifnet interface, SIOCSIFADDR's argument was
an ifaddr. That was confusing, and it would work against me as I
make it possible for a network interface to overload most ioctls.
On the protocol->ifnet interface, replace SIOCSIFADDR with
SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to
invoke SIOCINITIFADDR.
In ifioctl(), give the interface the first shot at handling most
interface ioctls, and give the protocol the second shot, instead
of the other way around. Finally, let compatibility code (COMPAT_OSOCK)
take a shot.
Pull device initialization out of switch statements under
SIOCINITIFADDR. For example, pull ..._init() out of any switch
statement that looks like this:
switch (...->sa_family) {
case ...:
..._init();
...
break;
...
default:
..._init();
...
break;
}
Rewrite many if-else clauses that handle all permutations of IFF_UP
and IFF_RUNNING to use a switch statement,
switch (x & (IFF_UP|IFF_RUNNING)) {
case 0:
...
break;
case IFF_RUNNING:
...
break;
case IFF_UP:
...
break;
case IFF_UP|IFF_RUNNING:
...
break;
}
unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and
#ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4).
In ipw(4), remove an if_set_sadl() call that is out of place.
In nfe(4), reuse the jumbo MTU logic in ether_ioctl().
Let ethernets register a callback for setting h/w state such as
promiscuous mode and the multicast filter in accord with a change
in the if_flags: ether_set_ifflags_cb() registers a callback that
returns ENETRESET if the caller should reset the ethernet by calling
if_init(), 0 on success, != 0 on failure. Pull common code from
ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(),
and register if_flags callbacks for those drivers.
Return ENOTTY instead of EINVAL for inappropriate ioctls. In
zyd(4), use ENXIO instead of ENOTTY to indicate that the device is
not any longer attached.
Add to if_set_sadl() a boolean 'factory' argument that indicates
whether a link-layer address was assigned by the factory or some
other source. In a comment, recommend using the factory address
for generating an EUI64, and update in6_get_hw_ifid() to prefer a
factory address to any other link-layer address.
Add a routing message, RTM_LLINFO_UPD, that tells protocols to
update the binding of network-layer addresses to link-layer addresses.
Implement this message in IPv4 and IPv6 by sending a gratuitous
ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD
messages on a change of an interface's link-layer address.
In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address
that is broadcast/multicast or equal to 00:00:00:00:00:00.
Make ether_ioctl() call ifioctl_common() to handle ioctls that it
does not understand.
In gif(4), initialize if_softc and use it, instead of assuming that
the gif_softc and ifp overlap.
Let ifioctl_common() handle SIOCGIFADDR.
Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels
that certain invariants on a struct route are satisfied.
In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit
about the ioctls that we do not allow on an agr(4) member interface.
bzero -> memset. Delete unnecessary casts to void *. Use
sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with
NULL instead of "testing truth". Replace some instances of (type
*)0 with NULL. Change some K&R prototypes to ANSI C, and join
lines.
|
| 1.143 | 24-Oct-2008 |
dyoung | branches: 1.143.2; 1.143.4; 1.143.6; 1.143.8;
Constify the rt_addrinfo argument to the ifa_rtrequest member
function of struct ifaddr.
|
| 1.142 | 21-Oct-2008 |
ad | arp_drain: no reason to complain if arp_lock is already held.
|
| 1.141 | 28-Aug-2008 |
uebayasi | Missing "\n" in log(9) messages.
|
| 1.140 | 13-May-2008 |
dyoung | branches: 1.140.4;
bzero -> memset, bcopy -> memcpy.
|
| 1.139 | 13-May-2008 |
dyoung | Cosmetic: use __arraycount(). s/0/NULL/ where appropriate. Pass
"null" instead of 0 to printf %s. Remove superfluous parentheses
in return statements. Compare pointers with NULL instead of "testing
truth."
|
| 1.138 | 11-May-2008 |
dyoung | Use memset() instead of Bzero().
In arplookup1(), put the static sockaddr_inarp onto the stack, and
zero it before use.
|
| 1.137 | 04-May-2008 |
thorpej | branches: 1.137.2;
Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.136 | 02-May-2008 |
ad | PR kern/38497 Out of memory allocating ksiginfo
Work around: don't acquire softnet_lock in protocol drain routines.
|
| 1.135 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.134 | 24-Apr-2008 |
ad | branches: 1.134.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.133 | 23-Apr-2008 |
thorpej | Use <net/net_stats.h> / netstat_sysctl().
|
| 1.132 | 15-Apr-2008 |
thorpej | branches: 1.132.2;
Make ARP stats per-cpu.
|
| 1.131 | 20-Jan-2008 |
joerg | branches: 1.131.6; 1.131.8;
Now that __HAVE_TIMECOUNTER and __HAVE_GENERIC_TODR are invariants,
remove the conditionals and the code associated with the undef case.
|
| 1.130 | 07-Dec-2007 |
elad | branches: 1.130.4;
Use struct initializers. No functional change.
|
| 1.129 | 14-Nov-2007 |
cube | branches: 1.129.2;
Follow up on arc -> arcnet renaming. Pointed out by joerg@.
|
| 1.128 | 02-Sep-2007 |
dyoung | branches: 1.128.4; 1.128.6;
We cannot sleep in a software interrupt, so do not sockaddr_dl_alloc(...,
M_WAITOK). Instead, sockaddr_dl_init() a sockaddr_dl on the stack.
|
| 1.127 | 30-Aug-2007 |
dyoung | Use malloc(9) for sockaddrs instead of pool(9), and remove dom_sa_pool
and dom_sa_len members from struct domain. Pools of fixed-size
objects are too rigid for sockaddr_dls, whose size can vary over
a wide range.
Return sockaddr_dl to its "historical" size. Now that I'm using
malloc(9) instead of pool(9) to allocate sockaddr_dl, I can create
a sockaddr_dl of any size in the kernel, so expanding sockaddr_dl
is useless.
Avoid using sizeof(struct sockaddr_dl) in the kernel.
Introduce sockaddr_dl_alloc() for allocating & initializing an
arbitrary sockaddr_dl on the heap.
Add an argument, the sockaddr length, to sockaddr_alloc(),
sockaddr_copy(), and sockaddr_dl_setaddr().
Constify: LLADDR() -> CLLADDR().
Where the kernel overwrites LLADDR(), use sockaddr_dl_setaddr(),
instead. Used properly, sockaddr_dl_setaddr() will not overrun
the end of the sockaddr.
|
| 1.126 | 27-Aug-2007 |
dyoung | branches: 1.126.2;
Reorganize and extract arplookup1() for code-sharing. Share
null_sdl. Introduce arp_setgate() for initializing a link-layer
nexthop, and use it to fulfill RTM_SETGATE requests.
|
| 1.125 | 19-Jul-2007 |
dyoung | branches: 1.125.4; 1.125.6;
Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.124 | 09-Jul-2007 |
ad | branches: 1.124.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.123 | 12-Jun-2007 |
dyoung | Complete removal of radix_node knowledge.
|
| 1.122 | 09-Jun-2007 |
dyoung | Get rid of radix_node_head.rnh_walktree, because it is only ever
set to rn_walktree.
Introduce rt_walktree(), which applies a subroutine to every route
in a particular address family. Use it instead of rn_walktree()
virtually everywhere. This helps to hide the routing table
implementation.
|
| 1.121 | 04-Mar-2007 |
christos | branches: 1.121.2; 1.121.4; 1.121.6;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.120 | 22-Feb-2007 |
matt | Fix lossage from boolean_t -> bool and updated x86 bus_dma.
|
| 1.119 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.118 | 17-Feb-2007 |
dyoung | branches: 1.118.2;
bcopy -> memcpy
Use NULL instead of (struct rtentry *)0.
|
| 1.117 | 24-Nov-2006 |
christos | branches: 1.117.2; 1.117.4; 1.117.8;
fix spelling of accidentally; from Zapher
|
| 1.116 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.115 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.114 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.113 | 30-Aug-2006 |
christos | branches: 1.113.2; 1.113.4;
Fix initializers
|
| 1.112 | 07-Jun-2006 |
kardel | merge FreeBSD timecounters from branch simonb-timecounters
- struct timeval time is gone
time.tv_sec -> time_second
- struct timeval mono_time is gone
mono_time.tv_sec -> time_uptime
- access to time via
{get,}{micro,nano,bin}time()
get* versions are fast but less precise
- support NTP nanokernel implementation (NTP API 4)
- further reading:
Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf
NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
|
| 1.111 | 25-May-2006 |
bouyer | Make sure the mbuf is writable before trying to write to it.
|
| 1.110 | 18-May-2006 |
liamjfoy | branches: 1.110.2;
Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.109 | 12-May-2006 |
mrg | since ar_tha() can return NULL, don't pass it directly to functions
that expect real addresses. explicitly KASSERT() that it is not
NULL in the kernel and just avoid using it userland.
(the kernel could be more defensive about this, but, until now it
would have just crashed anyway.)
|
| 1.108 | 24-Dec-2005 |
perry | branches: 1.108.4; 1.108.6; 1.108.8; 1.108.12;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.107 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.106 | 20-Jun-2005 |
atatat | branches: 1.106.2;
Change the rest of the sysctl subsystem to use const consistently.
The __UNCONST macro is now used only where necessary and the RW macros
are gone. Most of the changes here are consumers of the
sysctl_createv(9) interface that now takes a pair of const pointers
which used not to be.
|
| 1.105 | 01-Jun-2005 |
drochner | need a "const"
|
| 1.104 | 29-May-2005 |
christos | - remove local copy of hexdigits.
- sprinkle const
- use mem*() instead of b*()
|
| 1.103 | 26-Feb-2005 |
perry | branches: 1.103.2; 1.103.4; 1.103.6;
nuke trailing whitespace
|
| 1.102 | 02-Feb-2005 |
perry | de-__P, do some ANSIfication.
|
| 1.101 | 23-Jan-2005 |
matt | branches: 1.101.2;
Change initialzie of domains to use link sets. Switch to using STAILQ.
Add a convenience macro DOMAIN_FOREACH to interate through the domain.
|
| 1.100 | 04-Dec-2004 |
peter | branches: 1.100.4;
Convert lo(4) to a clonable device.
This also removes the loif array and changes all code to use the new
lo0ifp pointer which points to the lo0 ifnet structure.
Approved by christos.
|
| 1.99 | 29-Sep-2004 |
christos | PR/21902: Sean Boudreau: arplookup() incrementing arpstat.as_allocfail
erroneously.
|
| 1.98 | 25-May-2004 |
atatat | Sysctl descriptions under net subtree (net.key not done)
|
| 1.97 | 28-Apr-2004 |
ragge | Send an arp request before the arp entry times out if the entry is active,
to avoid deleting active entries.
Add sysctl support to tune the default arp timeout values.
|
| 1.96 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.95 | 21-Apr-2004 |
matt | Constify if.c radix.c and route.c (and fix related fallout).
|
| 1.94 | 24-Sep-2003 |
itojun | on arplookup() failure, nuke cloned route - otherwise outsider could use massive
number of bogus ARPs for DoS attack. FreeBSD-SA-03:14.arp
|
| 1.93 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.92 | 26-Feb-2003 |
matt | branches: 1.92.2;
Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.91 | 20-Nov-2002 |
dyoung | Squash a panic: do not try to print the name of a NULL interface.
|
| 1.90 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.89 | 25-Jun-2002 |
enami | If we need to fix up ar_hrd field, we must do it before using ar_tpa/tha.
|
| 1.88 | 25-Jun-2002 |
itojun | in arprequest(), fill ar_hrd only for IEEE1394. for other cases,
ifp->if_output will fill it for us.
|
| 1.87 | 25-Jun-2002 |
enami | No need to include same file twice.
|
| 1.86 | 25-Jun-2002 |
enami | Use if_addrlen macro rather than if_data.ifi_addrlen.
|
| 1.85 | 24-Jun-2002 |
enami | The ieee1394 arp reply should be broadcast.
|
| 1.84 | 24-Jun-2002 |
enami | Don't use a pointer before it is initialized.
|
| 1.83 | 24-Jun-2002 |
itojun | set ar_hrd for RFC-defined cases
|
| 1.82 | 24-Jun-2002 |
itojun | integrate IEEE1394 ARP into generic ARP logic.
XXX there's no check at all in ar_hrd, and we don't set ar_hrd on outgoing.
it seems like a bad thing.
|
| 1.81 | 09-Jun-2002 |
itojun | whitespace
|
| 1.80 | 09-Jun-2002 |
itojun | enforce IPv4 link MTU for FDDI and ARCNET even in RTF_GATEWAY case.
PR 17151.
|
| 1.79 | 13-Nov-2001 |
lukem | branches: 1.79.8; 1.79.10;
add RCSIDs
|
| 1.78 | 20-Aug-2001 |
itojun | if I'm bridging and got a packet to interface address on if A from if B,
advertise MAC address for if A with ARP reply.
|
| 1.77 | 17-Aug-2001 |
thorpej | Permit weaker interface matches for incoming ARP packets if the packet was
received on an interface that is part of a bridge and we find an ifaddr on
an interface that is part of the same bridge.
|
| 1.76 | 04-Jul-2001 |
itojun | branches: 1.76.2;
better support for multiple IPv4 addresses on a single interface.
- consider non-primary (2nd and beyond) IPv4 address as "local", and prevent
outgoing ARP.
- for routing entries generated by ARP, make sure to set rt->rt_ifa equal to
rt_key, to help IPv4 source address selection for traffic to myself.
PR 13311.
caveats/TODOs:
- interface routes ("connected routes" in cisco terminlogy) is tied with the
primary (1st) IPv4 address on the interface. should be fixed with updates
to rt_ifinit().
- source address selection for offlink locations. 1st address tend to be used
with the current code
(you can configure it right by setting rt->rt_ifa accordingly).
|
| 1.75 | 11-Jun-2001 |
tron | Make arplookup error messages more informative. Patch supplied by
Andrew Brown in PR kern/13162.
|
| 1.74 | 14-May-2001 |
matt | Use the LIST_NEXT & LIST_FIRST macros instead of refering to
le_next & lh_first.
|
| 1.73 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.72 | 26-Jan-2001 |
is | branches: 1.72.2;
Make diagnostic actually useful - needed to debug other ARP PRs.
Suggested by Geoff C. Wing in PR 10815.
|
| 1.71 | 17-Jan-2001 |
itojun | pull post-4.4BSD change to sys/net/route.c from BSD/OS 4.2 (UCB copyrighted).
have sys/net/route.c:rtrequest1(), which takes rt_addrinfo * as the argument.
pass rt_addrinfo all the way down to rtrequest, and ifa->ifa_rtrequest.
3rd arg of ifa->ifa_rtrequest is now rt_addrinfo * instead of sockaddr *
(almost noone is using it anyways).
benefit: the follwoing command now works. previously we need two route(8)
invocations, "add" then "change".
# route add -inet6 default ::1 -ifp gif0
remove unsafe typecast in rtrequest(), from rtentry * to sockaddr *. it was
introduced by 4.3BSD-reno and never corrected.
XXX is eon_rtrequest() change correct regarding to 3rd arg?
eon_rtrequest() and rtrequest() were incorrect since 4.3BSD-reno,
so i do not have correct answer in the source code.
someone with more clue about netiso-over-ip, please help.
|
| 1.70 | 15-Aug-2000 |
jhawk | Add kernel counters for arp events, displayable with netstat -s -f arp
|
| 1.69 | 20-May-2000 |
jhawk | branches: 1.69.4;
Install "show arptab" (db_show_arptab) in the ddb command tree.
Move prototype from netinet/if_inarp.h to ddb/db_interface.h.
Change function to have standard ddb parameters (though they're
ignored).
|
| 1.68 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.67 | 23-Mar-2000 |
thorpej | New callout mechanism with two major improvements over the old
timeout()/untimeout() API:
- Clients supply callout handle storage, thus eliminating problems of
resource allocation.
- Insertion and removal of callouts is constant time, important as
this facility is used quite a lot in the kernel.
The old timeout()/untimeout() API has been removed from the kernel.
|
| 1.66 | 25-Sep-1999 |
is | branches: 1.66.2;
Decouple IP mtu for ARCnet devices from interface MTU.
This is important, because for most protocols, link level fragmentation is
used, but with different default effective MTUs. (e.g.: IPv4 default MTU
is 1500 octets, IPv6 default MTU is 9072 octets).
|
| 1.65 | 21-Aug-1999 |
matt | Cleanup a little kludge in mtu handling in route.c. Bring down FDDI
mtu to legal IP max but don't affect other protocols.
|
| 1.64 | 05-Aug-1999 |
sommerfeld | Delete debug printfs from arp_drain()
|
| 1.63 | 05-Aug-1999 |
sommerfeld | Implement arp_drain(), which frees packets tied up in the arp cache if
mbufs are in short supply.
Create a (trivial) protocol domain for arp so that the drain routine will
be called from m_reclaim()
|
| 1.62 | 18-Jun-1999 |
thorpej | When sending an ARP reply, make sure to set the length of the outgoing
packet.
Slightly modified from PR #7809, Zdenek Salvet <salvet@ics.muni.cz>.
|
| 1.61 | 30-May-1999 |
bad | Fix thinko of mine in previous. The source route info is not at m->m_data
after various m_adj()s have been done. Kludge around this with a cheesy
macro that knows where the drivers put the mac header in the first mbuf.
XXX There should be a better way to do this.
|
| 1.60 | 29-May-1999 |
bad | Don't assume the Token-Ring source route is in the m_pktdat. Use
m_data instead. This isn't a problem with ARP packets but is correct
way to this.
Noticed by pmara@cactus.org (Shashi Mara).
|
| 1.59 | 23-May-1999 |
ad | For completeness sake, allow this to compile with no loopback interfaces
configured.
|
| 1.58 | 04-May-1999 |
is | Fixes PR 7489 by Olaf Seibert. Fix by Zdenek Salvet (PR 7497).
|
| 1.57 | 04-May-1999 |
is | Fix for PR 7490 by Olaf Seibert, fix mostly from PR 7497 bei Zdenek Salvet,
but with more verbose error messages.
|
| 1.56 | 22-Mar-1999 |
bad | branches: 1.56.2;
Add support for Token-Ring source routes in the ARP cache.
By Onno van der Linden.
|
| 1.55 | 21-Feb-1999 |
drochner | -always do an RARP if revarpwhoarewe() is called, it might be for another
interface or the server's configuration has changed
-g/c revarpwhoami()
|
| 1.54 | 19-Dec-1998 |
thorpej | Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.53 | 01-Oct-1998 |
drochner | branches: 1.53.4;
print reason for arplookup() failure (ala FreeBSD)
|
| 1.52 | 30-Sep-1998 |
tls | Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.51 | 05-Jul-1998 |
jonathan | defopt INET, NETATALK.
|
| 1.50 | 04-Jul-1998 |
jonathan | defopt DDB.
|
| 1.49 | 02-Jul-1998 |
is | Thinko in last fix: we have to actually check each address for a copy on
our ifp, else we might fail for some strange configurations.
|
| 1.48 | 02-Jul-1998 |
is | The rewrite of if_arp.c to work with the hashed interface address lists
(1.44) missed a test for the right interface, making some machines answer
to some bogus arp requests (like for WHO-HAS 127.0.0.1).
The quick patch in 1.46-1.47 does not work for so-called "unnumbered"
interfaces, that is, (point-to-point) interfaces that share their local
address with another (e.g., the Ethernet) interface.
We add a macro to in_var.h, to step (in the current implementation) through
the hash chain and fine more entries with the same address, and use that
in if_arp.c to find one which belongs to our interface.
|
| 1.47 | 25-Jun-1998 |
tls | Fix buglet where we might respond to arp on wrong interface.
|
| 1.46 | 29-May-1998 |
matt | Change arp so its console log messages print out IP addresses in
dotted quad format instead of hex.
|
| 1.45 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.44 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.43 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.42 | 31-Oct-1997 |
gwr | Get rid of the messages: "arp: zero IP addr from ..."
If one really wants to know about those confused PCs
trying to use IP address zero, they can use tcpdump.
|
| 1.41 | 02-Oct-1997 |
is | branches: 1.41.2;
Reimplement a test for broadcast addresses advertized, which was left out
when rewriting the ARP system.
|
| 1.40 | 29-Aug-1997 |
gwr | Tweaks to allow operation with an interface address of 0.0.0.0
(needed for NFS mountroot using BOOTP to get boot parameters)
|
| 1.39 | 04-Aug-1997 |
lukem | struct loif is an array of NLOOP (from "loop.h") elements
|
| 1.38 | 27-May-1997 |
gwr | branches: 1.38.4;
Allow revarpwhoami() to be called multiple times.
(Just return the answer if we already have it.)
Without this, the RB_ASKNAME loop fails on every
call to nfs_mountroot after the first call.
|
| 1.37 | 07-Apr-1997 |
jtk | add newlines at end of debugging log messages which were missing them
|
| 1.36 | 23-Mar-1997 |
is | Fix several bugs related to the new ARP code, and ARCnet ARP support.
Among other, add ARPHRD_ARCNET definition, make sure the hardware type is
set on outgoing ARP packets, make sure we dont send out replies as broadcasts.
|
| 1.35 | 15-Mar-1997 |
is | New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.34 | 13-Oct-1996 |
christos | branches: 1.34.4;
backout previous kprintf changes
|
| 1.33 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.32 | 09-Oct-1996 |
thorpej | Merge netbsd-1-2 branch back into mainline.
|
| 1.31 | 11-May-1996 |
mycroft | branches: 1.31.4;
When sending an ARP request, use the interface address for the route, rather
than the first address assigned. This gives slightly different behaviour in
the presence of aliases. From Bill Fenner, via Pete Bentley.
|
| 1.30 | 07-May-1996 |
thorpej | Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
|
| 1.29 | 30-Mar-1996 |
christos | Fix db_printf formats
|
| 1.28 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.27 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.26 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.25 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.24 | 15-May-1995 |
cgd | spacing fixups and KNF. #define ether address size, so it's not
hardcoded as '6' all over.
|
| 1.23 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.22 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.21 | 11-Apr-1995 |
mycroft | Remove some explicit references to loif.
|
| 1.20 | 07-Apr-1995 |
mycroft | Add a common function to initialize ARP-related variables. `Insired'
by Garrett Wollman.
|
| 1.19 | 06-Mar-1995 |
glass | remove references to arptnew. fix spelling error
|
| 1.18 | 27-Jul-1994 |
mycroft | Fix byte-order bug in printf() statement.
|
| 1.17 | 24-Jul-1994 |
cgd | kill conflicting externs
|
| 1.16 | 29-Jun-1994 |
cgd | branches: 1.16.2;
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.15 | 21-Jun-1994 |
chopps | config.new hack for if_ether.c for lack of an `and' in the grammer
and protect some ether specific code in in.c
|
| 1.14 | 04-Jun-1994 |
gwr | Back out some of my changes which Keith Sklower convinced me are
unnecessary. Leaving in just the essentials of the fix.
|
| 1.13 | 03-Jun-1994 |
gwr | Avoid accidentaly creating permanent entries at time==0
Routes created with RTM_ADD (i.e. manually added) are
permanent so leave their expiration time set to zero.
|
| 1.12 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.11 | 29-Apr-1994 |
cgd | change timeout/untimeout/wakeup/sleep/tsleep args to void *
|
| 1.10 | 18-Apr-1994 |
mycroft | Dummy arpintr() for now.
|
| 1.9 | 18-Apr-1994 |
glass | revised nfs diskless support. uses bootp+rpc to gather parameters
|
| 1.8 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.7 | 23-Jan-1994 |
deraadt | ether_output() & ether_input() take ether_type as a net-short.
AF_UNSPEC does not swap byte order of ether_type.
NOTE: this requires driver changes
|
| 1.6 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 12-Dec-1993 |
hpeyerl | >From cmaeda@cs.washington.edu; part of the multicast patches derived
from the Multicast patches for BSDI.
Thanx to Brad Parker for making me realize i'd forgotten to commit
this patch..(color me dopey)
|
| 1.3 | 27-Jun-1993 |
andrew | branches: 1.3.4;
ANSIfications - removed all implicit function return types and argument
definitions. Ensured that all files include "systm.h" to gain access to
general prototypes. Casts where necessary.
|
| 1.2 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.3.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.16.2.2 | 28-Jul-1994 |
cgd | from trunk.
|
| 1.16.2.1 | 24-Jul-1994 |
cgd | from trunk.
|
| 1.31.4.2 | 31-Aug-1996 |
thorpej | Delete these from the NetBSD 1.2 release branch; they were tagged in error.
|
| 1.31.4.1 | 14-Jun-1996 |
is | Not needed for the netbsd-1-2 release
|
| 1.34.4.1 | 09-Mar-1997 |
is | netinet/if_ether.h -> netinet/if_inarp.h
|
| 1.38.4.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.38.4.2 | 01-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.38.4.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.41.2.2 | 01-Oct-1998 |
cgd | pull up revisions 1.43-1.45, 1.47-1.49, 1.52 from trunk. (tls)
|
| 1.41.2.1 | 31-Oct-1997 |
mellon | Pull rev 1.42 up from trunk (gwr)
|
| 1.53.4.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.56.2.4 | 02-Jul-2000 |
he | Apply patch (partially revision 1.63, requested by sommerfeld):
Protect arp table handling with splnet() to avoid interrupt races
when ip_flow is in use. Fixes PR#10351.
|
| 1.56.2.3 | 20-Jun-1999 |
perry | pullup 1.61->1.62 (thorpej)
|
| 1.56.2.2 | 04-May-1999 |
perry | branches: 1.56.2.2.2; 1.56.2.2.4;
pullup 1.57->1.58 (is)
|
| 1.56.2.1 | 04-May-1999 |
perry | pullup 1.56->1.57 (is)
|
| 1.56.2.2.4.1 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.56.2.2.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.66.2.4 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.66.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.66.2.2 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.66.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.69.4.1 | 01-Oct-2003 |
msaitoh | Pull up revision 1.94 via patch (requested by itojun in ticket #86):
On arplookup() failure, nuke cloned route - otherwise outsider could use
massive number of bogus ARPs for DoS attack. FreeBSD-SA-03:14.arp
|
| 1.72.2.7 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.72.2.6 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.72.2.5 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.72.2.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.72.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.72.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.72.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.76.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.76.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.76.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.76.2.1 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.79.10.1 | 01-Oct-2003 |
tron | Pull up revision 1.94 via patch (requested by itojun in ticket #1482):
on arplookup() failure, nuke cloned route - otherwise outsider could use massive
number of bogus ARPs for DoS attack. FreeBSD-SA-03:14.arp
|
| 1.79.8.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.79.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.92.2.9 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.92.2.8 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.92.2.7 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.92.2.6 | 24-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.92.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.92.2.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.92.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.92.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.92.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.100.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.101.2.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.101.2.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.103.6.1 | 29-Aug-2008 |
bouyer | Pull up following revision(s) (requested by uebayasi in ticket #1956):
sys/netinet/if_arp.c: revision 1.141
Missing "\n" in log(9) messages.
|
| 1.103.4.1 | 29-Aug-2008 |
bouyer | Pull up following revision(s) (requested by uebayasi in ticket #1956):
sys/netinet/if_arp.c: revision 1.141
Missing "\n" in log(9) messages.
|
| 1.103.2.1 | 29-Aug-2008 |
bouyer | Pull up following revision(s) (requested by uebayasi in ticket #1956):
sys/netinet/if_arp.c: revision 1.141
Missing "\n" in log(9) messages.
|
| 1.106.2.6 | 21-Jan-2008 |
yamt | sync with head
|
| 1.106.2.5 | 15-Nov-2007 |
yamt | sync with head.
|
| 1.106.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.106.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.106.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.106.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.108.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.108.8.3 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.108.8.2 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.108.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.108.6.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.108.6.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time(), use bintime's for timeout
calculations and use "time_second" instead of "time.tv_sec".
|
| 1.108.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.110.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.113.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.113.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.113.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.113.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.117.8.1 | 29-Aug-2008 |
bouyer | Pull up following revision(s) (requested by uebayasi in ticket #1191):
sys/netinet/if_arp.c: revision 1.141
Missing "\n" in log(9) messages.
|
| 1.117.4.1 | 04-Sep-2008 |
skrll | Sync with netbsd-4.
|
| 1.117.2.1 | 29-Aug-2008 |
bouyer | Pull up following revision(s) (requested by uebayasi in ticket #1191):
sys/netinet/if_arp.c: revision 1.141
Missing "\n" in log(9) messages.
|
| 1.118.2.3 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.118.2.2 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.118.2.1 | 17-Feb-2007 |
yamt | file if_arp.c was added on branch yamt-idlelwp on 2007-02-27 16:54:52 +0000
|
| 1.121.6.1 | 09-Dec-2007 |
reinoud | Pullup to HEAD
|
| 1.121.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.121.2.4 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.121.2.3 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.121.2.2 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.121.2.1 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.124.2.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.124.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.125.6.2 | 19-Jul-2007 |
dyoung | Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.125.6.1 | 19-Jul-2007 |
dyoung | file if_arp.c was added on branch matt-mips64 on 2007-07-19 20:48:54 +0000
|
| 1.125.4.3 | 09-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.125.4.2 | 14-Nov-2007 |
joerg | Sync with HEAD.
|
| 1.125.4.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.126.2.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.126.2.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.126.2.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.128.6.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.128.6.2 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.128.6.1 | 19-Nov-2007 |
mjf | Sync with HEAD.
|
| 1.128.4.1 | 18-Nov-2007 |
bouyer | Sync with HEAD
|
| 1.129.2.1 | 08-Dec-2007 |
ad | Sync with head.
|
| 1.130.4.1 | 23-Jan-2008 |
bouyer | Sync with HEAD.
|
| 1.131.8.4 | 04-Jan-2009 |
christos | fix time_t format
|
| 1.131.8.3 | 09-Nov-2008 |
christos | merge with head.
|
| 1.131.8.2 | 01-Nov-2008 |
christos | Sync with head.
|
| 1.131.8.1 | 29-Mar-2008 |
christos | Welcome to the time_t=long long dev_t=uint64_t branch.
|
| 1.131.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.131.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.131.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.132.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.134.2.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.134.2.3 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.134.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.134.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.137.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.137.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.140.4.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.140.4.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.143.8.2 | 11-May-2010 |
matt | Fix printf for u_quad_t route changes.
|
| 1.143.8.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.143.6.1 | 21-Nov-2009 |
snj | Pull up following revision(s) (requested by christos in ticket #1156):
sys/net/if_arcsubr.c: revision 1.61
sys/net/if_ethersubr.c: revision 1.173
sys/net/if_fddisubr.c: revision 1.78
sys/net/if_tokensubr.c: revision 1.58 via patch
sys/netinet/if_arp.c: revision 1.149
ar_tha() can return NULL; treat this as an error.
|
| 1.143.4.2 | 21-Nov-2009 |
snj | Pull up following revision(s) (requested by christos in ticket #1156):
sys/net/if_arcsubr.c: revision 1.61
sys/net/if_ethersubr.c: revision 1.173
sys/net/if_fddisubr.c: revision 1.78
sys/net/if_tokensubr.c: revision 1.58 via patch
sys/netinet/if_arp.c: revision 1.149
ar_tha() can return NULL; treat this as an error.
|
| 1.143.4.1 | 05-Sep-2009 |
bouyer | Pull up following revision(s) (requested by dyoung in ticket #911):
sys/netinet/if_arp.c: revision 1.146
Don't require the gateway address to have room for both an interface
name and address. Room for an address will do. This should fix
a regression in 'arp -s ...' on interfaces such as xennet0 with
unusually long names. Fix PR #41878.
|
| 1.143.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.149.8.1 | 08-Feb-2011 |
bouyer | Sync with HEAD
|
| 1.149.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.149.4.2 | 31-May-2011 |
rmind | sync with head
|
| 1.149.4.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.152.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.152.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.152.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.154.16.2 | 05-Feb-2017 |
snj | Pull up following revision(s) (requested by maxv in ticket #1432):
sys/netinet/if_arp.c: 1.238, 1.239 via patch
Make sure the protocol address length equals that of IPv4. Also, make sure
the hardware address length equals that of the interface we received the
packet on. Otherwise a packet could easily set them both to zero and make
the kernel read beyond the allocated mbuf, which is terrible.
Note: for the latter we drop the packet instead of replying, since it is
malformed.
Note: I also added an ugly hack in CARP, since it apparently expects at
least six bytes.
--
Add some checks, mostly same as in_arpinput.
|
| 1.154.16.1 | 18-Jun-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1067):
sys/dist/ipf/netinet/ip_fil_netbsd.c 1.9 via patch
sys/net/if_ethersubr.c 1.197 via patch
sys/net/if_loop.c 1.77 via patch
sys/net/if_vlan.c 1.70 via patch
sys/netinet/if_arp.c 1.158
sys/netinet/ip_carp.c 1.54 via patch
sys/netinet6/ip6_flow.c 1.23 via patch
sys/netinet6/nd6.c 1.150 via patch
sys/rump/librump/rumpkern/klock.c 1.4
Make sure *(if_output)() is called with KERNEL_LOCK held to avoid mbuf leak.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details. For netinet6, the problem report, fix and test were done
by njoly@ on current-users@
|
| 1.154.10.2 | 18-May-2014 |
rmind | sync with head
|
| 1.154.10.1 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.154.8.2 | 05-Feb-2017 |
snj | Pull up following revision(s) (requested by maxv in ticket #1432):
sys/netinet/if_arp.c: 1.238, 1.239 via patch
Make sure the protocol address length equals that of IPv4. Also, make sure
the hardware address length equals that of the interface we received the
packet on. Otherwise a packet could easily set them both to zero and make
the kernel read beyond the allocated mbuf, which is terrible.
Note: for the latter we drop the packet instead of replying, since it is
malformed.
Note: I also added an ugly hack in CARP, since it apparently expects at
least six bytes.
--
Add some checks, mostly same as in_arpinput.
|
| 1.154.8.1 | 18-Jun-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1067):
sys/dist/ipf/netinet/ip_fil_netbsd.c 1.9 via patch
sys/net/if_ethersubr.c 1.197 via patch
sys/net/if_loop.c 1.77 via patch
sys/net/if_vlan.c 1.70 via patch
sys/netinet/if_arp.c 1.158
sys/netinet/ip_carp.c 1.54 via patch
sys/netinet6/ip6_flow.c 1.23 via patch
sys/netinet6/nd6.c 1.150 via patch
sys/rump/librump/rumpkern/klock.c 1.4
Make sure *(if_output)() is called with KERNEL_LOCK held to avoid mbuf leak.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details. For netinet6, the problem report, fix and test were done
by njoly@ on current-users@
|
| 1.154.6.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.154.6.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.154.2.3 | 05-Feb-2017 |
snj | Pull up following revision(s) (requested by maxv in ticket #1432):
sys/netinet/if_arp.c: 1.238, 1.239 via patch
Make sure the protocol address length equals that of IPv4. Also, make sure
the hardware address length equals that of the interface we received the
packet on. Otherwise a packet could easily set them both to zero and make
the kernel read beyond the allocated mbuf, which is terrible.
Note: for the latter we drop the packet instead of replying, since it is
malformed.
Note: I also added an ugly hack in CARP, since it apparently expects at
least six bytes.
--
Add some checks, mostly same as in_arpinput.
|
| 1.154.2.2 | 15-Nov-2015 |
bouyer | Pull up following revision(s) (requested by ozaki-r in ticket #1328):
sys/netinet/if_arp.c: revision 1.160
Add sysctl to selectively log arp packets from unknown network. (Adrien URBAN).
|
| 1.154.2.1 | 03-Jun-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1067):
sys/dist/ipf/netinet/ip_fil_netbsd.c 1.9 via patch
sys/net/if_ethersubr.c 1.197 via patch
sys/net/if_loop.c 1.77 via patch
sys/net/if_vlan.c 1.70 via patch
sys/netinet/if_arp.c 1.158
sys/netinet/ip_carp.c 1.54 via patch
sys/netinet6/ip6_flow.c 1.23 via patch
sys/netinet6/nd6.c 1.150 via patch
sys/rump/librump/rumpkern/klock.c 1.4
Make sure *(if_output)() is called with KERNEL_LOCK held to avoid mbuf leak.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details. For netinet6, the problem report, fix and test were done
by njoly@ on current-users@
|
| 1.155.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.158.4.1 | 05-Feb-2017 |
snj | Pull up following revision(s) (requested by maxv in ticket #1356):
sys/netinet/if_arp.c: revision 1.238, 1.239 via patch
Make sure the protocol address length equals that of IPv4. Also, make sure
the hardware address length equals that of the interface we received the
packet on. Otherwise a packet could easily set them both to zero and make
the kernel read beyond the allocated mbuf, which is terrible.
Note: for the latter we drop the packet instead of replying, since it is
malformed.
Note: I also added an ugly hack in CARP, since it apparently expects at
least six bytes.
--
Add some checks, mostly same as in_arpinput.
|
| 1.158.2.2 | 05-Feb-2017 |
snj | Pull up following revision(s) (requested by maxv in ticket #1356):
sys/netinet/if_arp.c: revision 1.238, 1.239 via patch
Make sure the protocol address length equals that of IPv4. Also, make sure
the hardware address length equals that of the interface we received the
packet on. Otherwise a packet could easily set them both to zero and make
the kernel read beyond the allocated mbuf, which is terrible.
Note: for the latter we drop the packet instead of replying, since it is
malformed.
Note: I also added an ugly hack in CARP, since it apparently expects at
least six bytes.
--
Add some checks, mostly same as in_arpinput.
|
| 1.158.2.1 | 06-Nov-2015 |
riz | branches: 1.158.2.1.2;
Pull up following revision(s) (requested by ozaki-r in ticket #985):
sys/netinet/if_arp.c: revision 1.160
Add sysctl to selectively log arp packets from unknown network. (Adrien URBAN).
|
| 1.158.2.1.2.1 | 13-Mar-2017 |
skrll | Sync with netbsd-7-1-RELEASE
|
| 1.160.2.12 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.160.2.11 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.160.2.10 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.160.2.9 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.160.2.8 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.160.2.7 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.160.2.6 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.160.2.5 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.160.2.4 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.160.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.160.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.160.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.217.2.6 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.217.2.5 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.217.2.4 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.217.2.3 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.217.2.2 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.217.2.1 | 26-Jul-2016 |
pgoyette | Sync with HEAD
|
| 1.233.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.248.4.1 | 19-May-2017 |
pgoyette | Resolve conflicts from previous merge (all resulting from $NetBSD
keywork expansion)
|
| 1.250.2.10 | 30-Sep-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1396):
sys/netinet6/nd6.h: revision 1.88
sys/netinet6/nd6_nbr.c: revision 1.174
sys/netinet6/nd6.c: revision 1.264
sys/netinet/if_arp.c: revision 1.288 (patch)
Initialize DAD components properly
The original code initialized each component in non-init functions such as
arp_dad_start and nd6_dad_find, conditionally based on a global flag for each.
However, it was racy because the flag and the code around it were not
protected by a lock and could cause a kernel panic at worst.
Fix the issue by initializing the components in bootup as usual.
|
| 1.250.2.9 | 06-Nov-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1080):
sys/netinet6/nd6.c: revision 1.251
sys/netinet/if_arp.c: revision 1.276
sys/net/if.c: revision 1.438
sys/net/if.c: revision 1.439
sys/net/route.c: revision 1.214
sys/net/route.c: revision 1.215
sys/net/route.c: revision 1.216
sys/netinet6/in6.c: revision 1.270
sys/net/route.h: revision 1.120
sys/net/if.c: revision 1.440
Remove a wrong assertion in ifaref
-
Doing ifref on an ifa with IFA_DESTROYING is not a problem; the reference should
be dropped during the destruction of the ifa.
-
Use atomic operations for ifa_refcnt
-
Avoid a dangling pointer during rt_replace_ifa
-
Avoid double rt_replace_ifa on rtrequest1(RTM_ADD)
Some callers of rtrequest1(RTM_ADD) adjust rt_ifa of an rtentry created by
rtrequest1 that may change rt_ifa (in ifa_rtrequest) with another ifa that is
different from requested one. It's wasteful and even worse introduces a race
condition. rtrequest1 should just use a passed ifa as is if a caller hopes so.
-
Use rt_update framework on updating a rtentry
|
| 1.250.2.8 | 02-Apr-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #686):
sys/netinet/if_arp.c: revision 1.271
sys/netinet6/nd6_nbr.c: revision 1.151,1.152
Avoid passing NULL to nd6_dad_duplicated
Fix PR kern/53075
Fix a race condition on DAD destructions (again)
The previous fix to DAD timers was wrong; it avoided a use-after-free but
instead introduced a memory leak. The destruction method had delegated
a destruction of a DAD timer to the timer itself and told that by setting NULL
to dp->dad_ifa. However, the previous fix made DAD timers do nothing on
the sign.
Fixing the issue with using callout_stop isn't easy. One approach is to have
a refcount on dp but it introduces extra complexity that we want to avoid.
The new fix falls back to using callout_halt, which was abandoned because of
softnet_lock. Fortunately now the network stack is protected by KERNEL_LOCK
so we can remove softnet_lock from DAD timers (callout) and use callout_halt
safely.
|
| 1.250.2.7 | 13-Mar-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #622):
sys/netinet/if_arp.c: revision 1.270
sys/net/if_llatbl.c: revision 1.24 (patch)
sys/net/if_llatbl.c: revision 1.25
sys/net/if_llatbl.c: revision 1.26
sys/net/route.c: revision 1.204
sys/netinet6/in6.c: revision 1.261
sys/netinet6/in6.c: revision 1.262 (patch)
sys/netinet6/in6.c: revision 1.263
sys/netinet/in.c: revision 1.216
sys/netinet6/in6.c: revision 1.264
sys/netinet6/nd6.c: revision 1.246 (patch)
sys/netinet/if_arp.c: revision 1.269
sys/net/if_llatbl.h: revision 1.14
sys/netinet6/in6.c: revision 1.259
sys/netinet/in.c: revision 1.220
sys/netinet/in.c: revision 1.221 (patch)
sys/netinet/in.c: revision 1.222
sys/netinet/in.c: revision 1.223
Suppress noisy debugging outputs
Even if DEBUG they are too noisy under load.
Tweak sanity checks
Scheduling a timer of static entries is wrong.
Add assertions
We must not destroy llentries holding mbufs.
Fix reference leaks of llentry
callout_reset and callout_halt can cancel a pending callout without telling us.
Detect a cancel and remove a reference by using callout_pending and
callout_stop (it's a bit tricy though, we can detect it).
While here, we can remove remaining abuses of mutex_owned for softnet_lock.
Fix memory leaks on arp -d and ndp -d for static entries
We have to delete entries on in_lltable_delete and in6_lltable_delete
unconditionally. Note that we don't need to worry about LLE_IFADDR because
there is no such entries now.
Use pool(9) for llentry allocations
llentry is easy to be leaked and pool suits for it because pool is usable to
detect leaks.
Also sweep unnecessary wrappers for llentry, in_llentry and in6_llentry.
|
| 1.250.2.6 | 26-Feb-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #589):
sys/netinet/if_arp.c: revision 1.267
sys/netinet6/nd6_nbr.c: revision 1.146-1.148
Use KASSERT for checking a programming error
Simplify; pass dp to nd6_dad_duplicated instead of looking it up again in it
Avoid a race condition of DAD timer destructions
When we see dp->dad_ifa == NULL, it means that the ifa is being deleted and also
the callout is scheduled again by someone. We shouldn't rely on a result of
callout_pending to know if the callout is scheduled because it returns false if
the subsequent callout handler is already on the fly.
We have to always delegate the destruction of dp to the subsequent handler
unconditionally if dp->dad_ifa == NULL. Otherwise, the first handler destroys
the dp and the second handler tries to handle destroyed dp.
|
| 1.250.2.5 | 26-Jan-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #511):
sys/kern/kern_timeout.c: revision 1.54
sys/netinet6/nd6_nbr.c: revision 1.141
sys/netinet6/nd6_nbr.c: revision 1.144
sys/netinet/if_arp.c: revision 1.256
Fix a deadlock on callout_halt of nd6_dad_timer
We must not call callout_halt of nd6_dad_timer with holding nd6_dad_lock because
the lock is taken in nd6_dad_timer. Once softnet_lock goes away, we can pass the
lock to callout_halt, but for now we cannot.
Make DAD destructions (MP-)safe with callout_stop
arp_dad_stoptimer and nd6_dad_stoptimer can be called with or without
softnet_lock held and unfortunately we have no easy way to statically know which.
So it is hard to use callout_halt there.
To address the situation, we use callout_stop to make the code safe. The new
approach copes with the issue by delegating the destruction of a callout to
callout itself, which allows us to not wait the callout to finish. This can be
done thanks to that DAD objects are separated from other data such as ifa.
The approach is suggested by riastradh@
Proposed on tech-kern@ and tech-net@
Sanity-check if interlock is held when it's passed
|
| 1.250.2.4 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.250.2.3 | 17-Nov-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #353):
sys/net/if_llatbl.c: 1.22
sys/net/if_llatbl.h: 1.13
sys/netinet/if_arp.c: 1.254
sys/netinet/in.c: 1.208-1.209
sys/netinet6/in6.c: 1.249-1.250
sys/netinet6/nd6.c: 1.237
Remove redundant KASSERTMSG
The function is static, has just one caller and the caller does the same check.
--
Fix a deadlock between a route update and lltable
It happens because rtalloc1 is called from lltable with holding
IF_AFDATA_WLOCK.
If a route update is in action, rtalloc1 would wait for its completion with
holding IF_AFDATA_WLOCK. At the same moment, a softint (e.g., arpintr) may try
to take IF_AFDATA_WLOCK and get stuck on it. Unfortunately the stuck softint
prevents the route update from progressing because the route update calls
psref_target_destroy that needs the softint to complete.
A resource allocation graph of the senario looks like this:
route update =(psref_target_destroy)=> softint => IF_AFDATA_WLOCK
=(rt_update_wait)=> route update
Fix the deadlock by pulling rtalloc1 out of the lltable codes inside
IF_AFDATA_WLOCK.
Note that the deadlock happens only if NET_MPSAFE is enabled.
|
| 1.250.2.2 | 07-Jul-2017 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #107):
usr.sbin/arp/arp.c: revision 1.56
sys/net/rtsock.c: revision 1.218
sys/net/if_llatbl.c: revision 1.20
usr.sbin/arp/arp.c: revision 1.57
sys/net/rtsock.c: revision 1.219
sys/net/if_llatbl.c: revision 1.21
usr.sbin/arp/arp.c: revision 1.58
tests/net/net_common.sh: revision 1.19
sys/netinet6/nd6.h: revision 1.84
sys/netinet6/nd6.h: revision 1.85
tests/net/arp/t_arp.sh: revision 1.23
sys/netinet6/in6.c: revision 1.246
tests/net/arp/t_arp.sh: revision 1.24
sys/netinet6/in6.c: revision 1.247
tests/net/arp/t_arp.sh: revision 1.25
sys/netinet6/in6.c: revision 1.248
tests/net/arp/t_arp.sh: revision 1.26
usr.sbin/ndp/ndp.c: revision 1.49
tests/net/arp/t_arp.sh: revision 1.27
tests/net/ndp/t_ndp.sh: revision 1.20
tests/net/arp/t_arp.sh: revision 1.28
tests/net/ndp/t_ndp.sh: revision 1.21
tests/net/arp/t_arp.sh: revision 1.29
tests/net/ndp/t_ndp.sh: revision 1.22
tests/net/ndp/t_ndp.sh: revision 1.23
tests/net/route/t_flags6.sh: revision 1.13
tests/net/ndp/t_ndp.sh: revision 1.24
tests/net/route/t_flags6.sh: revision 1.14
tests/net/ndp/t_ndp.sh: revision 1.25
tests/net/route/t_flags6.sh: revision 1.15
tests/net/ndp/t_ndp.sh: revision 1.26
sbin/route/rtutil.c: revision 1.9
tests/net/ndp/t_ndp.sh: revision 1.27
tests/net/ndp/t_ndp.sh: revision 1.28
tests/net/net/t_ipv6address.sh: revision 1.14
tests/net/ndp/t_ra.sh: revision 1.28
tests/net/ndp/t_ndp.sh: revision 1.29
sys/net/route.h: revision 1.113
tests/net/ndp/t_ra.sh: revision 1.29
sys/net/rtsock.c: revision 1.220
sys/net/rtsock.c: revision 1.221
sys/net/rtsock.c: revision 1.222
sys/net/rtsock.c: revision 1.223
tests/net/route/t_route.sh: revision 1.13
sys/net/rtsock.c: revision 1.224
sys/net/route.c: revision 1.196
sys/net/if_llatbl.c: revision 1.19
sys/net/route.c: revision 1.197
sbin/route/route.c: revision 1.156
tests/net/route/t_flags.sh: revision 1.16
tests/net/route/t_flags.sh: revision 1.17
usr.sbin/ndp/ndp.c: revision 1.50
tests/net/route/t_flags.sh: revision 1.18
sys/netinet/in.c: revision 1.204
tests/net/route/t_flags.sh: revision 1.19
sys/netinet/in.c: revision 1.205
tests/net/arp/t_arp.sh: revision 1.30
tests/net/arp/t_arp.sh: revision 1.31
sys/net/if_llatbl.h: revision 1.11
tests/net/arp/t_arp.sh: revision 1.32
sys/net/if_llatbl.h: revision 1.12
tests/net/arp/t_arp.sh: revision 1.33
sys/netinet6/nd6.c: revision 1.233
sys/netinet6/nd6.c: revision 1.234
sys/netinet/if_arp.c: revision 1.251
sys/netinet6/nd6.c: revision 1.235
sys/netinet/if_arp.c: revision 1.252
sbin/route/route.8: revision 1.57
sys/net/rtsock.c: revision 1.214
sys/net/rtsock.c: revision 1.215
sys/net/rtsock.c: revision 1.216
sys/net/rtsock.c: revision 1.217
whitespace police
Simplify
We can assume that rt_ifp is always non-NULL.
Sending a routing message (RTM_ADD) on adding an llentry
A message used to be sent on adding a cloned route. Restore the
behavior for backward compatibility.
Requested by ryo@
Drop RTF_CONNECTED from a result of RTM_GET for ARP/NDP entries
ARP/NDP entries aren't connected routes.
Reported by ryo@
Support -c <count> option for route monitor
route command exits if it receives <count> routing messages where
<count> is a value specified by -c.
The option is useful to get only particular message(s) in a test script.
Test routing messages emitted on operations of ARP/NDP entries
Do netstat -a for an appropriate protocol
Add missing declarations for cleanup
Set net.inet.arp.keep only if it's required
Don't create a permanent L2 cache entry on adding an address to an interface
It was created to copy FreeBSD, however actually the cache isn't
necessary. Remove it to simplify the code and reduce the cost to
maintain it (e.g., keep a consistency with a corresponding local
route).
Fix typo
Fix in_lltable_match_prefix
The function has not been used but will be used soon.
Remove unused function (nd6_rem_ifa_lle)
Allow in6_lltable_free_entry to be called without holding the afdata lock of ifp as well as in_lltable_free_entry
This behavior is a bit odd and should be fixed in the future...
Purge ARP/NDP entries on an interface when the interface is down
Fix PR kern/51179
Purge all related L2 caches on removing a route
The change addresses situations similar to PR 51179.
Purge L2 caches on changing an interface of a route
The change addresses situations similar to PR 51179.
Test implicit removals of ARP/NDP entries
One test case reproudces PR 51179.
Fix build of kernels without both INET and INET6
Tweak lltable_sysctl_dumparp
- Rename lltable_sysctl_dumparp to lltable_sysctl_dump
because it's not only for ARP
- Enable it not only for INET but also for INET6
Fix usage of routing messages on arp -d and ndp -d
It didn't work as we expected; we should set RTA_GATEWAY not
RTA_IFP on RTM_GET to return an if_index and the kernel should
use it on RTM_DELETE.
Improve backward compatibility of (fake) routing messages on adding an ARP/NDP entry
A message originally included only DST and GATEWAY. Restore it.
Fix ifdef; care about a case w/ INET6 and w/o INET
Drop RTF_UP from a routing message of a deleted ARP/NDP entry
Check existence of ARP/NDP entries
Checking ARP/NDP entries is valid rather than checking routes.
Fix wrong comment
Drop RTF_LLINFO flag (now it's RTF_LLDATA) from local routes
They don't have llinfo anymore. And also the change fixes unexpected
behavior of ARP proxy.
Restore ARP/NDP entries to route show and netstat -r
Requested by dyoung@ some time ago
Enable to remove multiple ARP/NDP entries for one destination
The kernel can have multiple ARP/NDP entries which have an indentical
destination on different interfaces. This is normal and can be
reproduce easily by ping -I or ping6 -S. We should be able to remove
such entries.
arp -d <ip> and ndp -d <ip> are changed to fetch all ARP/NDP entries
and remove matched entries. So we can remove multiple entries
described above. This fetch all and selective removal behavior is
the same as arp <ip> and ndp <ip>; they also do fetch all entries
and show only matched entries.
Related to PR 51179
Check if ARP/NDP entries are purged when a related route is deleted
|
| 1.250.2.1 | 01-Jul-2017 |
snj | Pull up following revision(s) (requested by roy in ticket #77):
sys/net/if.h: revision 1.240
sys/netinet/if_arp.c: revision 1.253
sys/net/if.c: revision 1.395
Introduce if_get_bylla to find an interface with the active
local link address.
--
Use if_get_bylla() instead of just looking at the lla of the interface
the address belongs to.
This allows any ARP message we receieved from another interface to
be correctly dropped.
While here, move the protocol length check higher up the food chain.
|
| 1.268.2.6 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.268.2.5 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.268.2.4 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.268.2.3 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.268.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.268.2.1 | 15-Mar-2018 |
pgoyette | Synch with HEAD
|
| 1.275.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.275.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.282.2.7 | 24-Aug-2024 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1883):
tests/net/arp/t_dad.sh: revision 1.16
sys/netinet/in.c: revision 1.248
tests/net/arp/t_arp.sh: revision 1.46
sys/netinet/if_arp.c: revision 1.314
arp: fix the behavior on detecting an address duplication without IPv4 DAD
On receiving an ARP request that has the same source protocol address as
the own address, i.e., address duplication, the original behavior of
a kernel prior to supporing IPv4 DAD is to send an ARP reply. It is
the same with a latest kernel with DAD enabled. However, a latest
kernel without DAD sends back an GARP packet. Restore the original
behavior.
inet: send GARP on link up if DAD is disabled
This behavior was accidentally removed at rev 1.233.
tests, arp: add tests of address duplications without DAD
tests, arp: add tests for GARP on link up
|
| 1.282.2.6 | 24-Jan-2020 |
martin | Pull up following revision(s) (requested by roy in ticket #645):
sys/netinet/if_arp.c: revision 1.292
arp: find source address then target address when processing input
This fixes the case where another host having a duplicate ip address
starts using it right away without probing for it's availability.
While here, prefer ifatoia over a strict cast.
|
| 1.282.2.5 | 11-Oct-2019 |
martin | Pull up following revision(s) (requested by roy in ticket #300):
sys/netinet/if_arp.c: revision 1.289
ARP: Don't defend ARP probes.
We should let the nature of ARP takes it's course here when our address
is neither tentative nor duplicated.
This allows the host to work with ARP ping, which was broken in r1.279.
|
| 1.282.2.4 | 30-Sep-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #269):
sys/netinet6/nd6.h: revision 1.88
sys/net/rtsock_shared.c: revision 1.10
sys/netinet6/nd6_nbr.c: revision 1.174
sys/netinet6/nd6.c: revision 1.264
sys/netinet/if_arp.c: revision 1.283
sys/netinet/if_arp.c: revision 1.288
Initialize DAD components properly
The original code initialized each component in non-init functions such as
arp_dad_start and nd6_dad_find, conditionally based on a global flag for each.
However, it was racy because the flag and the code around it were not
protected by a lock and could cause a kernel panic at worst.
Fix the issue by initializing the components in bootup as usual.
-
Initialize dom_mowner for MBUFTRACE
|
| 1.282.2.3 | 05-Sep-2019 |
martin | Pull up following revision(s) (requested by roy in ticket #170):
sys/netinet/if_arp.c: revision 1.287
inet: Send RTM_MISS when we fail to resolve an address.
Takes the same approach as when adding a new address - we no longer
announce the new lladdr right away but we announce the result.
This will either be RTM_ADD or RTM_MISS.
RTM_DELETE is only sent if we have a lladdr assigned OR gc'ed.
This tells us when a new lladdr has been added (RTM_ADD),
changed (RTM_CHANGE), deleted (RTM_DELETED) or has failed to been
resolved (RTM_MISS). The latter case can be interpreted as unreachable.
|
| 1.282.2.2 | 01-Sep-2019 |
martin | Pull up following revision(s) (requested by roy in ticket #147):
sys/netinet/if_arp.c: revision 1.285
sys/netinet/if_arp.c: revision 1.286
ARP: remove unused sysctl entry log_unknown_network
ARP: change default sysctl entry log_movements to 0
IP address sharing is a thing and shouldn't cause needless diagnostics
by default.
|
| 1.282.2.1 | 26-Aug-2019 |
martin | Pull up following revision(s) (requested by roy in ticket #109):
sys/net/route.h: revision 1.124
sys/netinet6/nd6.c: revision 1.258
sys/netinet6/nd6.c: revision 1.259
sys/net/rtsock.c: revision 1.251
sys/netinet/if_arp.c: revision 1.284
sys/netinet6/nd6_nbr.c: revision 1.167
rtsock: rework rt_clonedmsg to take a message type and lladdr
We will use this in a future patch to notify userland of lladdr
changes.
XXX pullup -8 -9
-
nd6: notify userland of neighbour lla updates once more
XXX pullup -8 -9
|
| 1.289.2.1 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.297.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.311.2.3 | 13-Sep-2024 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #859):
tests/net/arp/t_arp.sh: revision 1.47
tests/net/arp/t_arp.sh: revision 1.48
sys/netinet/if_arp.c: revision 1.315
arp: allow to send packets without an ARP resolution just after
receiving an ARP request
On receiving an ARP request, the current implemention creates an ARP
cache entry but with ND_LLINFO_NOSTATE. Such an entry still needs
an ARP resolution to send back a packet to the requester. The original
behavior before introducing the common ND framework didn't need the
resolution. IPv6 doesn't as well. To restore the original behavior,
make a new ARP cache entry with ND_LLINFO_STALE like IPv6 does.
tests: dedup t_arp.sh like others (NFC)
tests: add tests for ARP cache entry creations
|
| 1.311.2.2 | 24-Aug-2024 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #812):
tests/net/arp/t_dad.sh: revision 1.16
sys/netinet/in.c: revision 1.248
tests/net/arp/t_arp.sh: revision 1.46
sys/netinet/if_arp.c: revision 1.314
arp: fix the behavior on detecting an address duplication without IPv4 DAD
On receiving an ARP request that has the same source protocol address as
the own address, i.e., address duplication, the original behavior of
a kernel prior to supporing IPv4 DAD is to send an ARP reply. It is
the same with a latest kernel with DAD enabled. However, a latest
kernel without DAD sends back an GARP packet. Restore the original
behavior.
inet: send GARP on link up if DAD is disabled
This behavior was accidentally removed at rev 1.233.
tests, arp: add tests of address duplications without DAD
tests, arp: add tests for GARP on link up
|
| 1.311.2.1 | 10-Mar-2024 |
martin | Pull up following revision(s) (requested by riastradh in ticket #618):
sys/netinet/if_arp.c: revision 1.312
Attribute debug message.
Fixes PR 57959
|
| 1.313.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.25 | 15-Mar-1997 |
is | New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.24 | 15-Jan-1997 |
gwr | branches: 1.24.2;
sync with if_ether.h
|
| 1.23 | 09-Oct-1996 |
thorpej | branches: 1.23.2;
Merge netbsd-1-2 branch back into mainline.
|
| 1.22 | 11-May-1996 |
mycroft | branches: 1.22.4;
When sending an ARP request, use the interface address for the route, rather
than the first address assigned. This gives slightly different behaviour in
the presence of aliases. From Bill Fenner, via Pete Bentley.
|
| 1.21 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.20 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.19 | 16-May-1995 |
cgd | parenthesize macro arg usage
|
| 1.18 | 15-May-1995 |
cgd | spacing fixups and KNF. #define ether address size, so it's not
hardcoded as '6' all over.
|
| 1.17 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.16 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.15 | 10-Apr-1995 |
mycroft | Remove now unneeded #ifdef. Prototype new function.
|
| 1.14 | 29-Mar-1995 |
briggs | KERNEL -> _KERNEL
|
| 1.13 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.12 | 06-Mar-1995 |
glass | remove references to arptnew. fix spelling error
|
| 1.11 | 27-Feb-1995 |
glass | fix some typos. from frank@fwi.uva.nl (Frank van der Linden)
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.7 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.6 | 30-Dec-1993 |
deraadt | "struct ether_addr" for ethers(3) functions.
|
| 1.5 | 13-Dec-1993 |
hpeyerl | >From cmaeda@cs.washington.edu; part of the multicast patches derived
from the Multicast patches for BSDI.
(I am a "big dopey bear" for having forgotten this. Thanx Havard.)
|
| 1.4 | 05-Sep-1993 |
cassidy | Add definition for reverse address resolution protocol.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.22.4.2 | 31-Aug-1996 |
thorpej | Delete these from the NetBSD 1.2 release branch; they were tagged in error.
|
| 1.22.4.1 | 14-Jun-1996 |
is | Not needed for the netbsd-1-2 release
|
| 1.23.2.1 | 18-Jan-1997 |
thorpej | Update from trunk.
|
| 1.24.2.1 | 09-Mar-1997 |
is | netinet/if_ether.h -> netinet/if_inarp.h
|
| 1.40 | 06-Sep-2018 |
maxv | Remove the network ATM code.
|
| 1.39 | 12-Dec-2016 |
ozaki-r | branches: 1.39.14; 1.39.16;
Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.38 | 28-Apr-2016 |
ozaki-r | branches: 1.38.2;
Constify rtentry of if_output
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
|
| 1.37 | 04-Apr-2016 |
ozaki-r | Separate nexthop caches from the routing table
By this change, nexthop caches (IP-MAC address pair) are not stored
in the routing table anymore. Instead nexthop caches are stored in
each network interface; we already have lltable/llentry data structure
for this purpose. This change also obsoletes the concept of cloning/cloned
routes. Cloned routes no longer exist while cloning routes still exist
with renamed to connected routes.
Noticeable changes are:
- Nexthop caches aren't listed in route show/netstat -r
- sysctl(NET_RT_DUMP) doesn't return them
- If RTF_LLDATA is specified, it returns nexthop caches
- Several definitions of routing flags and messages are removed
- RTF_CLONING, RTF_XRESOLVE, RTF_LLINFO, RTF_CLONED and RTM_RESOLVE
- RTF_CONNECTED is added
- It has the same value of RTF_CLONING for backward compatibility
- route's -xresolve, -[no]cloned and -llinfo options are removed
- -[no]cloning remains because it seems there are users
- -[no]connected is introduced and recommended
to be used instead of -[no]cloning
- route show/netstat -r drops some flags
- 'L' and 'c' are not seen anymore
- 'C' now indicates a connected route
- Gateway value of a route of an interface address is now not
a L2 address but "link#N" like a connected (cloning) route
- Proxy ARP: "arp -s ... pub" doesn't create a route
You can know details of behavior changes by seeing diffs under tests/.
Proposed on tech-net and tech-kern:
http://mail-index.netbsd.org/tech-net/2016/03/11/msg005701.html
|
| 1.36 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.35 | 17-Jul-2015 |
ozaki-r | Reform use of rt_refcnt
rt_refcnt of rtentry was used in bad manners, for example, direct rt_refcnt++
and rt_refcnt-- outside route.c, "rt->rt_refcnt++; rtfree(rt);" idiom, and
touching rt after rt->rt_refcnt--.
These abuses seem to be needed because rt_refcnt manages only references
between rtentry and doesn't take care of references during packet processing
(IOW references from local variables). In order to reduce the above abuses,
the latter cases should be counted by rt_refcnt as well as the former cases.
This change improves consistency of use of rt_refcnt:
- rtentry is always accessed with rt_refcnt incremented
- rtentry's rt_refcnt is decremented after use (rtfree is always used instead
of rt_refcnt--)
- functions returning rtentry increment its rt_refcnt (and caller rtfree it)
Note that rt_refcnt prevents rtentry from being freed but doesn't prevent
rtentry from being updated. Toward MP-safe, we need to provide another
protection for rtentry, e.g., locks. (Or introduce a better data structure
allowing concurrent readers during updates.)
|
| 1.34 | 10-Nov-2014 |
maxv | branches: 1.34.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.33 | 24-Sep-2012 |
msaitoh | branches: 1.33.12;
Add missing "\n" in log(9)
|
| 1.32 | 01-Feb-2011 |
chuck | branches: 1.32.4; 1.32.10; 1.32.14;
udpate license clauses on my code to match the new-style BSD licenses.
based on diff that rmind@ sent me.
no functional change with this commit.
|
| 1.31 | 18-Apr-2009 |
tsutsui | branches: 1.31.4; 1.31.6; 1.31.8;
Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.30 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.29 | 24-Oct-2008 |
dyoung | branches: 1.29.2; 1.29.8;
Constify the rt_addrinfo argument to the ifa_rtrequest member
function of struct ifaddr.
|
| 1.28 | 22-May-2008 |
dyoung | branches: 1.28.4;
Don't cast to void * unnecessarily.
|
| 1.27 | 05-Sep-2007 |
dyoung | branches: 1.27.20; 1.27.22; 1.27.24; 1.27.26;
We cannot sleep in a software interrupt, so do not sockaddr_dl_alloc(...,
M_WAITOK). Instead, sockaddr_dl_init() a sockaddr_dl on the stack.
|
| 1.26 | 30-Aug-2007 |
dyoung | Use malloc(9) for sockaddrs instead of pool(9), and remove dom_sa_pool
and dom_sa_len members from struct domain. Pools of fixed-size
objects are too rigid for sockaddr_dls, whose size can vary over
a wide range.
Return sockaddr_dl to its "historical" size. Now that I'm using
malloc(9) instead of pool(9) to allocate sockaddr_dl, I can create
a sockaddr_dl of any size in the kernel, so expanding sockaddr_dl
is useless.
Avoid using sizeof(struct sockaddr_dl) in the kernel.
Introduce sockaddr_dl_alloc() for allocating & initializing an
arbitrary sockaddr_dl on the heap.
Add an argument, the sockaddr length, to sockaddr_alloc(),
sockaddr_copy(), and sockaddr_dl_setaddr().
Constify: LLADDR() -> CLLADDR().
Where the kernel overwrites LLADDR(), use sockaddr_dl_setaddr(),
instead. Used properly, sockaddr_dl_setaddr() will not overrun
the end of the sockaddr.
|
| 1.25 | 07-Aug-2007 |
dyoung | branches: 1.25.2; 1.25.4;
Use satocsdl() et cetera instead of SDL(). Constify.
|
| 1.24 | 19-Jul-2007 |
dyoung | branches: 1.24.4;
Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.23 | 04-Mar-2007 |
christos | branches: 1.23.2; 1.23.10;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.22 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.21 | 16-Nov-2006 |
christos | branches: 1.21.4;
__unused removal on arguments; approved by core.
|
| 1.20 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.19 | 30-Aug-2006 |
christos | branches: 1.19.2; 1.19.4;
fix initializers and add const.
|
| 1.18 | 11-Dec-2005 |
christos | branches: 1.18.4; 1.18.8;
merge ktrace-lwp.
|
| 1.17 | 02-Feb-2005 |
perry | branches: 1.17.6;
de-__P, do some ANSIfication.
|
| 1.16 | 11-Sep-2002 |
itojun | branches: 1.16.6; 1.16.14; 1.16.16;
KNF - return is not a function. sync w/kame.
|
| 1.15 | 09-Jun-2002 |
itojun | whitespace
|
| 1.14 | 13-Nov-2001 |
lukem | branches: 1.14.8;
add RCSIDs
|
| 1.13 | 17-Jan-2001 |
itojun | branches: 1.13.2; 1.13.4;
pull post-4.4BSD change to sys/net/route.c from BSD/OS 4.2 (UCB copyrighted).
have sys/net/route.c:rtrequest1(), which takes rt_addrinfo * as the argument.
pass rt_addrinfo all the way down to rtrequest, and ifa->ifa_rtrequest.
3rd arg of ifa->ifa_rtrequest is now rt_addrinfo * instead of sockaddr *
(almost noone is using it anyways).
benefit: the follwoing command now works. previously we need two route(8)
invocations, "add" then "change".
# route add -inet6 default ::1 -ifp gif0
remove unsafe typecast in rtrequest(), from rtentry * to sockaddr *. it was
introduced by 4.3BSD-reno and never corrected.
XXX is eon_rtrequest() change correct regarding to 3rd arg?
eon_rtrequest() and rtrequest() were incorrect since 4.3BSD-reno,
so i do not have correct answer in the source code.
someone with more clue about netiso-over-ip, please help.
|
| 1.12 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.11 | 01-Jul-1999 |
itojun | branches: 1.11.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.10 | 13-Sep-1998 |
christos | branches: 1.10.8; 1.10.10;
Fix copyright spacing and 'Van' -> 'van' for consistency.
|
| 1.9 | 05-Jul-1998 |
jonathan | defopt NATM.
|
| 1.8 | 05-Jul-1998 |
jonathan | defopt INET, NETATALK.
|
| 1.7 | 29-Apr-1998 |
thorpej | Fix some whitespace.
|
| 1.6 | 13-Oct-1996 |
christos | backout previous kprintf changes
|
| 1.5 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.4 | 03-Jul-1996 |
chuck | ported ATM to FreeBSD 2.2-960612-SNAP
|
| 1.3 | 29-Jun-1996 |
chuck | change:
- change asock to rxhand and adjust all for this [esp atm_input]
|
| 1.2 | 26-Jun-1996 |
chuck | fixes/new stuff:
[1] if user tries to enter in a bogus PVC don't leave it in the routing
table ... remove it
[2] change ioctl arg to include rxso for lower layer
[3] add hooks (inside "NATM" ifdef) for native mode atm sockets so that
they don't clash with IP PVCs. [i am still debugging the native
mode atm socket protosw code]
|
| 1.1 | 22-Jun-1996 |
chuck | network support for ATM networks (ATM == Async Transfer Mode, not
Automatic Teller Machine).
Currently supports PVCs only (no ATM ARP either).
|
| 1.10.10.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.10.10.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.10.8.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.11.2.2 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.11.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.13.4.3 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.13.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.13.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.13.2.3 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.13.2.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.13.2.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.14.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.16.16.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.16.14.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.16.6.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.17.6.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.17.6.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.17.6.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.17.6.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.18.8.1 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.18.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.19.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.19.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.19.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.21.4.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.21.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.23.10.3 | 10-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.23.10.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.23.10.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.23.2.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.23.2.1 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.24.4.3 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.24.4.2 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.24.4.1 | 09-Aug-2007 |
jmcneill | Sync with HEAD.
|
| 1.25.4.2 | 07-Aug-2007 |
dyoung | Use satocsdl() et cetera instead of SDL(). Constify.
|
| 1.25.4.1 | 07-Aug-2007 |
dyoung | file if_atm.c was added on branch matt-mips64 on 2007-08-07 04:37:05 +0000
|
| 1.25.2.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.27.26.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.27.24.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.27.22.1 | 04-Jun-2008 |
yamt | sync with head
|
| 1.27.20.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.27.20.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.28.4.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.29.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.29.2.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.31.8.1 | 08-Feb-2011 |
bouyer | Sync with HEAD
|
| 1.31.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.31.4.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.32.14.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.32.14.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.32.10.1 | 23-Oct-2012 |
riz | Pull up following revision(s) (requested by msaitoh in ticket #616):
sys/netinet/if_atm.c: revision 1.33
sys/net/if_arcsubr.c: revision 1.64
sys/netinet/ip_mroute.c: revision 1.126
Add missing "\n" in log(9)
|
| 1.32.4.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.33.12.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.34.2.4 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.34.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.34.2.2 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.34.2.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.38.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.39.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.39.14.1 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.14 | 06-Sep-2018 |
maxv | Remove the network ATM code.
|
| 1.13 | 28-Apr-2016 |
ozaki-r | branches: 1.13.16; 1.13.18;
Constify rtentry of if_output
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
|
| 1.12 | 01-Feb-2011 |
chuck | branches: 1.12.14; 1.12.32;
udpate license clauses on my code to match the new-style BSD licenses.
based on diff that rmind@ sent me.
no functional change with this commit.
|
| 1.11 | 24-Oct-2008 |
dyoung | branches: 1.11.16; 1.11.22; 1.11.24;
Constify the rt_addrinfo argument to the ifa_rtrequest member
function of struct ifaddr.
|
| 1.10 | 17-Feb-2007 |
dyoung | branches: 1.10.38; 1.10.42; 1.10.48;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.9 | 10-Dec-2005 |
elad | branches: 1.9.26;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.8 | 21-Apr-2004 |
itojun | branches: 1.8.12;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.7 | 18-Apr-2004 |
matt | De __P()
|
| 1.6 | 09-Jun-2002 |
itojun | branches: 1.6.6;
whitespace
|
| 1.5 | 17-Jan-2001 |
itojun | branches: 1.5.2; 1.5.4; 1.5.16;
pull post-4.4BSD change to sys/net/route.c from BSD/OS 4.2 (UCB copyrighted).
have sys/net/route.c:rtrequest1(), which takes rt_addrinfo * as the argument.
pass rt_addrinfo all the way down to rtrequest, and ifa->ifa_rtrequest.
3rd arg of ifa->ifa_rtrequest is now rt_addrinfo * instead of sockaddr *
(almost noone is using it anyways).
benefit: the follwoing command now works. previously we need two route(8)
invocations, "add" then "change".
# route add -inet6 default ::1 -ifp gif0
remove unsafe typecast in rtrequest(), from rtentry * to sockaddr *. it was
introduced by 4.3BSD-reno and never corrected.
XXX is eon_rtrequest() change correct regarding to 3rd arg?
eon_rtrequest() and rtrequest() were incorrect since 4.3BSD-reno,
so i do not have correct answer in the source code.
someone with more clue about netiso-over-ip, please help.
|
| 1.4 | 13-Sep-1998 |
christos | branches: 1.4.12;
Fix copyright spacing and 'Van' -> 'van' for consistency.
|
| 1.3 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.2 | 03-Jul-1996 |
chuck | ported ATM to FreeBSD 2.2-960612-SNAP
|
| 1.1 | 22-Jun-1996 |
chuck | network support for ATM networks (ATM == Async Transfer Mode, not
Automatic Teller Machine).
Currently supports PVCs only (no ATM ARP either).
|
| 1.4.12.1 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.5.16.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.5.4.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.5.2.1 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.6.6.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.6.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.6.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.6.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.8.12.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.8.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.9.26.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.10.48.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.10.42.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.10.38.1 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.11.24.1 | 08-Feb-2011 |
bouyer | Sync with HEAD
|
| 1.11.22.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.11.16.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.12.32.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.12.14.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.13.18.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.13.16.1 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.35 | 15-Mar-1997 |
is | New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.34 | 13-Oct-1996 |
christos | branches: 1.34.4;
backout previous kprintf changes
|
| 1.33 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.32 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.31 | 11-May-1996 |
mycroft | When sending an ARP request, use the interface address for the route, rather
than the first address assigned. This gives slightly different behaviour in
the presence of aliases. From Bill Fenner, via Pete Bentley.
|
| 1.30 | 07-May-1996 |
thorpej | Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
|
| 1.29 | 30-Mar-1996 |
christos | Fix db_printf formats
|
| 1.28 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.27 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.26 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.25 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.24 | 15-May-1995 |
cgd | spacing fixups and KNF. #define ether address size, so it's not
hardcoded as '6' all over.
|
| 1.23 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.22 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.21 | 11-Apr-1995 |
mycroft | Remove some explicit references to loif.
|
| 1.20 | 07-Apr-1995 |
mycroft | Add a common function to initialize ARP-related variables. `Insired'
by Garrett Wollman.
|
| 1.19 | 06-Mar-1995 |
glass | remove references to arptnew. fix spelling error
|
| 1.18 | 27-Jul-1994 |
mycroft | Fix byte-order bug in printf() statement.
|
| 1.17 | 24-Jul-1994 |
cgd | kill conflicting externs
|
| 1.16 | 29-Jun-1994 |
cgd | branches: 1.16.2;
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.15 | 21-Jun-1994 |
chopps | config.new hack for if_ether.c for lack of an `and' in the grammer
and protect some ether specific code in in.c
|
| 1.14 | 04-Jun-1994 |
gwr | Back out some of my changes which Keith Sklower convinced me are
unnecessary. Leaving in just the essentials of the fix.
|
| 1.13 | 03-Jun-1994 |
gwr | Avoid accidentaly creating permanent entries at time==0
Routes created with RTM_ADD (i.e. manually added) are
permanent so leave their expiration time set to zero.
|
| 1.12 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.11 | 29-Apr-1994 |
cgd | change timeout/untimeout/wakeup/sleep/tsleep args to void *
|
| 1.10 | 18-Apr-1994 |
mycroft | Dummy arpintr() for now.
|
| 1.9 | 18-Apr-1994 |
glass | revised nfs diskless support. uses bootp+rpc to gather parameters
|
| 1.8 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.7 | 23-Jan-1994 |
deraadt | ether_output() & ether_input() take ether_type as a net-short.
AF_UNSPEC does not swap byte order of ether_type.
NOTE: this requires driver changes
|
| 1.6 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 12-Dec-1993 |
hpeyerl | >From cmaeda@cs.washington.edu; part of the multicast patches derived
from the Multicast patches for BSDI.
Thanx to Brad Parker for making me realize i'd forgotten to commit
this patch..(color me dopey)
|
| 1.3 | 27-Jun-1993 |
andrew | branches: 1.3.4;
ANSIfications - removed all implicit function return types and argument
definitions. Ensured that all files include "systm.h" to gain access to
general prototypes. Casts where necessary.
|
| 1.2 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.3.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.16.2.2 | 28-Jul-1994 |
cgd | from trunk.
|
| 1.16.2.1 | 24-Jul-1994 |
cgd | from trunk.
|
| 1.34.4.11 | 09-Mar-1997 |
is | netinet/if_ether.h -> netinet/if_inarp.h
|
| 1.34.4.10 | 06-Mar-1997 |
is | Use <net/ethertypes.h> instead of including all the Ethernet declarations.
|
| 1.34.4.9 | 19-Feb-1997 |
is | In arpresolve(), only copy up to ifp->if_data.ifi_addrlen bytes to the edst[]
array.
Earlier, excessively long address ARP entries would overwrite part of the
stack after edst[]s end.
|
| 1.34.4.8 | 18-Feb-1997 |
is | Don't use data gotten over the network as the length for yet another bcmp.
We don't want to be vulnerable to crashes due to bad guys sending us strange
packets.
|
| 1.34.4.7 | 18-Feb-1997 |
is | Add sanity check for ah->ar_hln in in_arpinput(). We don't want to crash
when people send us bogus ARP packets. At the moment, works only for
link level protocols with constant address length.
|
| 1.34.4.6 | 18-Feb-1997 |
is | Having converted everything, remove the struct ether_arp definition completely.
Some small cleanup.
STILLTODO: some sanity checks of the (now) variable link level address length
in incoming packets..
|
| 1.34.4.5 | 17-Feb-1997 |
is | Converted revarprequest() to the new world order.
|
| 1.34.4.4 | 17-Feb-1997 |
is | Changed in_revarpinput() to new ARP.
|
| 1.34.4.3 | 13-Feb-1997 |
is | Convert the arp receive/reply code to the new world order.
|
| 1.34.4.2 | 12-Feb-1997 |
is | Changed arprequest() to use AF_ARP sockaddr and NOT build its own Ethernet
header. Added some missing pieces in ether_output() to support this.
|
| 1.34.4.1 | 07-Feb-1997 |
is | Snapshot of new ARP code.
Our old ARP code was hardwired for 6-byte length medium
addresses, while the protocol is designed for any size.
This snapshot contains a first hack at getting rid of
Ethernet specific data structures. The ep driver is updated
(and tested on the PCI bus), the iy and fpa drivers have been
updated, but not real life tested yet.
If you want to test this with other drivers, you have to update
them first yourself, and probably tag the relevant directories.
Better contact me if you want to do this.
|
| 1.37 | 03-Feb-2021 |
roy | Guard CTASSERT
|
| 1.36 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.35 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.34 | 25-Dec-2007 |
perry | branches: 1.34.110;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.33 | 25-Sep-2006 |
sketch | branches: 1.33.24; 1.33.30; 1.33.34; 1.33.38;
typo.
|
| 1.32 | 10-Dec-2005 |
elad | branches: 1.32.20; 1.32.22;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.31 | 07-Aug-2003 |
agc | branches: 1.31.16;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.30 | 20-Nov-1999 |
thorpej | branches: 1.30.28;
Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.29 | 10-Feb-1998 |
perry | branches: 1.29.14; 1.29.20;
add/cleanup multiple inclusion protection.
|
| 1.28 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.27 | 29-Jul-1997 |
is | Include file in place of the old netinet/if_ether.h, including stuff from
where it is now, and adding the specialized for Ethernet version of the ARP
structure, for the benefit of programs which are externally (to us) maintained
and not (yet) ported.
XXX This should NOT be used inside the kernel.
|
| 1.26 | 15-Mar-1997 |
is | New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.25 | 17-Jan-1997 |
mikel | branches: 1.25.2;
fix my typo; found by Klaus Klein <kleink@layla.inka.de>
|
| 1.24 | 17-Jan-1997 |
mikel | add prototypes for ethers(3) functions; fixes PR 2471.
fix suggested by Jason Thorpe.
|
| 1.23 | 09-Oct-1996 |
thorpej | Merge netbsd-1-2 branch back into mainline.
|
| 1.22 | 11-May-1996 |
mycroft | branches: 1.22.4;
When sending an ARP request, use the interface address for the route, rather
than the first address assigned. This gives slightly different behaviour in
the presence of aliases. From Bill Fenner, via Pete Bentley.
|
| 1.21 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.20 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.19 | 16-May-1995 |
cgd | parenthesize macro arg usage
|
| 1.18 | 15-May-1995 |
cgd | spacing fixups and KNF. #define ether address size, so it's not
hardcoded as '6' all over.
|
| 1.17 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.16 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.15 | 10-Apr-1995 |
mycroft | Remove now unneeded #ifdef. Prototype new function.
|
| 1.14 | 29-Mar-1995 |
briggs | KERNEL -> _KERNEL
|
| 1.13 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.12 | 06-Mar-1995 |
glass | remove references to arptnew. fix spelling error
|
| 1.11 | 27-Feb-1995 |
glass | fix some typos. from frank@fwi.uva.nl (Frank van der Linden)
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.7 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.6 | 30-Dec-1993 |
deraadt | "struct ether_addr" for ethers(3) functions.
|
| 1.5 | 13-Dec-1993 |
hpeyerl | >From cmaeda@cs.washington.edu; part of the multicast patches derived
from the Multicast patches for BSDI.
(I am a "big dopey bear" for having forgotten this. Thanx Havard.)
|
| 1.4 | 05-Sep-1993 |
cassidy | Add definition for reverse address resolution protocol.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.22.4.1 | 17-Jun-1996 |
gwr | Pad the struct arpcom to avoid unnecessary misalignments on m68k ports.
|
| 1.25.2.4 | 09-Mar-1997 |
is | netinet/if_ether.h -> netinet/if_inarp.h
|
| 1.25.2.3 | 06-Mar-1997 |
is | Wipe out double function prototypes.
|
| 1.25.2.2 | 18-Feb-1997 |
is | Having converted everything, remove the struct ether_arp definition completely.
Some small cleanup.
STILLTODO: some sanity checks of the (now) variable link level address length
in incoming packets..
|
| 1.25.2.1 | 07-Feb-1997 |
is | Snapshot of new ARP code.
Our old ARP code was hardwired for 6-byte length medium
addresses, while the protocol is designed for any size.
This snapshot contains a first hack at getting rid of
Ethernet specific data structures. The ep driver is updated
(and tested on the PCI bus), the iy and fpa drivers have been
updated, but not real life tested yet.
If you want to test this with other drivers, you have to update
them first yourself, and probably tag the relevant directories.
Better contact me if you want to do this.
|
| 1.29.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.29.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.30.28.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.30.28.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.30.28.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.30.28.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.31.16.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.31.16.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.31.16.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.32.22.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.32.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.33.38.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.33.34.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.33.30.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.33.24.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.34.110.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.9 | 24-Jun-2002 |
itojun | integrate IEEE1394 ARP into generic ARP logic.
XXX there's no check at all in ar_hrd, and we don't set ar_hrd on outgoing.
it seems like a bad thing.
|
| 1.8 | 09-Jun-2002 |
itojun | whitespace
|
| 1.7 | 15-Nov-2001 |
lukem | branches: 1.7.8;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.6 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.5 | 04-Jul-2001 |
itojun | branches: 1.5.2;
better support for multiple IPv4 addresses on a single interface.
- consider non-primary (2nd and beyond) IPv4 address as "local", and prevent
outgoing ARP.
- for routing entries generated by ARP, make sure to set rt->rt_ifa equal to
rt_key, to help IPv4 source address selection for traffic to myself.
PR 13311.
caveats/TODOs:
- interface routes ("connected routes" in cisco terminlogy) is tied with the
primary (1st) IPv4 address on the interface. should be fixed with updates
to rt_ifinit().
- source address selection for offlink locations. 1st address tend to be used
with the current code
(you can configure it right by setting rt->rt_ifa accordingly).
|
| 1.4 | 12-Jun-2001 |
wiz | receive, not recieve
|
| 1.3 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.2 | 17-Jan-2001 |
itojun | branches: 1.2.2;
pull post-4.4BSD change to sys/net/route.c from BSD/OS 4.2 (UCB copyrighted).
have sys/net/route.c:rtrequest1(), which takes rt_addrinfo * as the argument.
pass rt_addrinfo all the way down to rtrequest, and ifa->ifa_rtrequest.
3rd arg of ifa->ifa_rtrequest is now rt_addrinfo * instead of sockaddr *
(almost noone is using it anyways).
benefit: the follwoing command now works. previously we need two route(8)
invocations, "add" then "change".
# route add -inet6 default ::1 -ifp gif0
remove unsafe typecast in rtrequest(), from rtentry * to sockaddr *. it was
introduced by 4.3BSD-reno and never corrected.
XXX is eon_rtrequest() change correct regarding to 3rd arg?
eon_rtrequest() and rtrequest() were incorrect since 4.3BSD-reno,
so i do not have correct answer in the source code.
someone with more clue about netiso-over-ip, please help.
|
| 1.1 | 05-Nov-2000 |
onoe | branches: 1.1.2;
First Prototype implementation of network interface part for IEEE1394 (if_fw).
Current status:
Only OHCI chip is supported (fwohci).
ping (IPv4) works with Sony's implementation (SmartConnect) on Win98.
sometimes works but not stable.
Not implemented yet:
IRM (Isochronous Resource Manager) functionality.
Link layer fragmentation.
Topology map.
More to do:
clean ups
MCAP
charactor device part
dhcp
There is no entry in GENERIC config file yet.
Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.
|
| 1.1.2.4 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.1.2.3 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.1.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.1.2.1 | 05-Nov-2000 |
bouyer | file if_ieee1394arp.c was added on branch thorpej_scsipi on 2000-11-22 16:06:07 +0000
|
| 1.2.2.6 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.2.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.2.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.5.2.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.5.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.5.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.7.8.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.7.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.3 | 24-Jun-2002 |
itojun | integrate IEEE1394 ARP into generic ARP logic.
XXX there's no check at all in ar_hrd, and we don't set ar_hrd on outgoing.
it seems like a bad thing.
|
| 1.2 | 17-Jan-2001 |
itojun | branches: 1.2.2; 1.2.4; 1.2.16;
pull post-4.4BSD change to sys/net/route.c from BSD/OS 4.2 (UCB copyrighted).
have sys/net/route.c:rtrequest1(), which takes rt_addrinfo * as the argument.
pass rt_addrinfo all the way down to rtrequest, and ifa->ifa_rtrequest.
3rd arg of ifa->ifa_rtrequest is now rt_addrinfo * instead of sockaddr *
(almost noone is using it anyways).
benefit: the follwoing command now works. previously we need two route(8)
invocations, "add" then "change".
# route add -inet6 default ::1 -ifp gif0
remove unsafe typecast in rtrequest(), from rtentry * to sockaddr *. it was
introduced by 4.3BSD-reno and never corrected.
XXX is eon_rtrequest() change correct regarding to 3rd arg?
eon_rtrequest() and rtrequest() were incorrect since 4.3BSD-reno,
so i do not have correct answer in the source code.
someone with more clue about netiso-over-ip, please help.
|
| 1.1 | 05-Nov-2000 |
onoe | branches: 1.1.2;
First Prototype implementation of network interface part for IEEE1394 (if_fw).
Current status:
Only OHCI chip is supported (fwohci).
ping (IPv4) works with Sony's implementation (SmartConnect) on Win98.
sometimes works but not stable.
Not implemented yet:
IRM (Isochronous Resource Manager) functionality.
Link layer fragmentation.
Topology map.
More to do:
clean ups
MCAP
charactor device part
dhcp
There is no entry in GENERIC config file yet.
Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.
|
| 1.1.2.3 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.1.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.1.2.1 | 05-Nov-2000 |
bouyer | file if_ieee1394arp.h was added on branch thorpej_scsipi on 2000-11-22 16:06:07 +0000
|
| 1.2.16.1 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.2.4.1 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.2.2.1 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.53 | 03-Sep-2022 |
thorpej | Convert ARP from a legacy netisr to pktqueue.
|
| 1.52 | 11-Sep-2020 |
roy | ARP: Use ND rather than our own.
This brings the benefit of Neighbour Unreachability Detection which is
something ARP sorely lacks.
The new timings mirror those of IPv6 and are adjustable via sysctl(8).
Unlike IPv6 ND, these are global and not per interface.
|
| 1.51 | 21-Feb-2017 |
ozaki-r | Replace malloc for DAD with kmem and move them out of the lock for DAD
|
| 1.50 | 11-Oct-2016 |
roy | branches: 1.50.2;
Mark arprequest static and introduce arpannounce so that gratuitous
ARP requests are only send from valid addresses.
|
| 1.49 | 19-Apr-2016 |
ozaki-r | branches: 1.49.2;
Constify rtentry of arpresolve
We don't need to (rather shouldn't) modify rtentry in there.
|
| 1.48 | 07-Apr-2016 |
christos | - tidy up error messages
- add a length argument to arpresolve()
- add KASSERT for overflow
|
| 1.47 | 21-May-2015 |
ozaki-r | Make arp_init, in_revarpinput and revarprequest static
|
| 1.46 | 16-May-2015 |
roy | Separate ARP handling DAD from inet.
This is done by signalling the intent to try tentative addresses
and then clearing the intent once the address is setup.
When the ARP handler is installed (arp_ifinit) then it adds
dad start and stop functions to the address which are used instead
of calling ARP directly.
|
| 1.45 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.44 | 30-Sep-2012 |
dholland | branches: 1.44.14;
Requires <sys/queue.h> for LIST_ENTRY and netinet/in.h for struct in_addr.
|
| 1.43 | 11-Nov-2011 |
gdt | branches: 1.43.10;
Move RTF_ANNOUNCE flag so that it no longer conflicts with RTF_PROTO2.
RTF_ANNOUNCE was defined as RTF_PROTO2. The flag is used to indicated
that host should act as a proxy for a link level arp or ndp request.
(If RTF_PROTO2 is used as an experimental flag (as advertised),
various problems can occur.)
This commit provides a first-class definition with its own bit for
RTF_ANNOUNCE, removes the old aliasing definitions, and adds support
for the new RTF_ANNOUNCE flag to netstat(8) and route(8).,
Also, remove unused RTF_ flags that collide with RTF_PROTO1:
netinet/icmp6.h defined RTF_PROBEMTU as RTF_PROTO1
netinet/if_inarp.h defined RTF_USETRAILERS as RTF_PROTO1
(Neither of these flags are used anywhere. Both have been removed
to reduce chances of collision with RTF_PROTO1.)
Figuring this out and the diff are the work of Beverly Schwartz of
BBN.
(Passed release build, boot in VM, with no apparently related atf
failures.)
Approved for Public Release, Distribution Unlimited
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073.
|
| 1.42 | 18-Feb-2009 |
yamt | branches: 1.42.12;
remove unused #define.
|
| 1.41 | 24-Oct-2008 |
dyoung | branches: 1.41.2; 1.41.8;
Constify the rt_addrinfo argument to the ifa_rtrequest member
function of struct ifaddr.
|
| 1.40 | 15-Apr-2008 |
thorpej | branches: 1.40.4; 1.40.10;
Make ARP stats per-cpu.
|
| 1.39 | 04-Mar-2007 |
christos | branches: 1.39.36;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.38 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.37 | 18-May-2006 |
liamjfoy | branches: 1.37.14;
Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.36 | 10-Dec-2005 |
elad | branches: 1.36.4; 1.36.6; 1.36.8; 1.36.12;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.35 | 21-Apr-2004 |
itojun | branches: 1.35.12;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.34 | 18-Apr-2004 |
matt | De __P()
|
| 1.33 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.32 | 17-Jan-2001 |
itojun | branches: 1.32.24;
pull post-4.4BSD change to sys/net/route.c from BSD/OS 4.2 (UCB copyrighted).
have sys/net/route.c:rtrequest1(), which takes rt_addrinfo * as the argument.
pass rt_addrinfo all the way down to rtrequest, and ifa->ifa_rtrequest.
3rd arg of ifa->ifa_rtrequest is now rt_addrinfo * instead of sockaddr *
(almost noone is using it anyways).
benefit: the follwoing command now works. previously we need two route(8)
invocations, "add" then "change".
# route add -inet6 default ::1 -ifp gif0
remove unsafe typecast in rtrequest(), from rtentry * to sockaddr *. it was
introduced by 4.3BSD-reno and never corrected.
XXX is eon_rtrequest() change correct regarding to 3rd arg?
eon_rtrequest() and rtrequest() were incorrect since 4.3BSD-reno,
so i do not have correct answer in the source code.
someone with more clue about netiso-over-ip, please help.
|
| 1.31 | 20-May-2000 |
jhawk | Install "show arptab" (db_show_arptab) in the ddb command tree.
Move prototype from netinet/if_inarp.h to ddb/db_interface.h.
Change function to have standard ddb parameters (though they're
ignored).
|
| 1.30 | 30-Mar-2000 |
simonb | Extern decl of arpintrq.
|
| 1.29 | 05-Aug-1999 |
sommerfeld | branches: 1.29.2;
Implement arp_drain(), which frees packets tied up in the arp cache if
mbufs are in short supply.
Create a (trivial) protocol domain for arp so that the drain routine will
be called from m_reclaim()
|
| 1.28 | 21-Feb-1999 |
drochner | -always do an RARP if revarpwhoarewe() is called, it might be for another
interface or the server's configuration has changed
-g/c revarpwhoami()
|
| 1.27 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.26 | 15-Mar-1997 |
is | New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.25 | 17-Jan-1997 |
mikel | branches: 1.25.2;
fix my typo; found by Klaus Klein <kleink@layla.inka.de>
|
| 1.24 | 17-Jan-1997 |
mikel | add prototypes for ethers(3) functions; fixes PR 2471.
fix suggested by Jason Thorpe.
|
| 1.23 | 09-Oct-1996 |
thorpej | Merge netbsd-1-2 branch back into mainline.
|
| 1.22 | 11-May-1996 |
mycroft | branches: 1.22.4;
When sending an ARP request, use the interface address for the route, rather
than the first address assigned. This gives slightly different behaviour in
the presence of aliases. From Bill Fenner, via Pete Bentley.
|
| 1.21 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.20 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.19 | 16-May-1995 |
cgd | parenthesize macro arg usage
|
| 1.18 | 15-May-1995 |
cgd | spacing fixups and KNF. #define ether address size, so it's not
hardcoded as '6' all over.
|
| 1.17 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.16 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.15 | 10-Apr-1995 |
mycroft | Remove now unneeded #ifdef. Prototype new function.
|
| 1.14 | 29-Mar-1995 |
briggs | KERNEL -> _KERNEL
|
| 1.13 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.12 | 06-Mar-1995 |
glass | remove references to arptnew. fix spelling error
|
| 1.11 | 27-Feb-1995 |
glass | fix some typos. from frank@fwi.uva.nl (Frank van der Linden)
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.7 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.6 | 30-Dec-1993 |
deraadt | "struct ether_addr" for ethers(3) functions.
|
| 1.5 | 13-Dec-1993 |
hpeyerl | >From cmaeda@cs.washington.edu; part of the multicast patches derived
from the Multicast patches for BSDI.
(I am a "big dopey bear" for having forgotten this. Thanx Havard.)
|
| 1.4 | 05-Sep-1993 |
cassidy | Add definition for reverse address resolution protocol.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.22.4.1 | 17-Jun-1996 |
gwr | Pad the struct arpcom to avoid unnecessary misalignments on m68k ports.
|
| 1.25.2.3 | 06-Mar-1997 |
is | Wipe out double function prototypes.
|
| 1.25.2.2 | 18-Feb-1997 |
is | Having converted everything, remove the struct ether_arp definition completely.
Some small cleanup.
STILLTODO: some sanity checks of the (now) variable link level address length
in incoming packets..
|
| 1.25.2.1 | 07-Feb-1997 |
is | Snapshot of new ARP code.
Our old ARP code was hardwired for 6-byte length medium
addresses, while the protocol is designed for any size.
This snapshot contains a first hack at getting rid of
Ethernet specific data structures. The ep driver is updated
(and tested on the PCI bus), the iy and fpa drivers have been
updated, but not real life tested yet.
If you want to test this with other drivers, you have to update
them first yourself, and probably tag the relevant directories.
Better contact me if you want to do this.
|
| 1.29.2.2 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.29.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.32.24.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.32.24.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.32.24.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.32.24.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.35.12.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.35.12.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.35.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.36.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.36.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.36.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.36.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.37.14.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.37.14.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.39.36.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.39.36.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.40.10.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.40.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.41.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.41.2.1 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.42.12.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.42.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.43.10.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.43.10.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.44.14.4 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.44.14.3 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.44.14.2 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.44.14.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.49.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.49.2.1 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.50.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.70 | 15-May-2020 |
maxv | igmp_sendpkt() expects ip_output() to set 'imo.imo_multicast_ttl' into
'ip->ip_ttl'; but ip_output() won't if the target is not a multicast
address, meaning that the uninitialized 'ip->ip_ttl' byte gets sent to
the network. This leaks one byte of kernel heap.
Fix this by filling 'ip->ip_ttl' with a TTL of one.
Found by KMSAN.
Reported-by: syzbot+e49f7b8a8fec5a477c9a@syzkaller.appspotmail.com
|
| 1.69 | 14-Sep-2018 |
maxv | branches: 1.69.4;
Use non-variadic function pointer in protosw::pr_input.
|
| 1.68 | 21-Jun-2018 |
knakahara | branches: 1.68.2;
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.67 | 10-Apr-2018 |
maxv | Replace comment by KASSERT.
|
| 1.66 | 07-Feb-2018 |
maxv | branches: 1.66.2;
Remove RSVP_ISI, that's mostly dead code. FreeBSD and OpenBSD too removed
it; FreeBSD kept some pieces but they are mostly no-opts.
Sent on tech-net@, no comment.
|
| 1.65 | 17-Nov-2017 |
ozaki-r | Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
|
| 1.64 | 24-Jan-2017 |
ozaki-r | branches: 1.64.6;
Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.63 | 11-Jan-2017 |
ozaki-r | branches: 1.63.2;
Get rid of unnecessary header inclusions
|
| 1.62 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.61 | 08-Jul-2016 |
ozaki-r | branches: 1.61.2;
Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.60 | 28-Jun-2016 |
ozaki-r | Add missing NULL checks for m_get_rcvif_psref
|
| 1.59 | 21-Jun-2016 |
ozaki-r | Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.58 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.57 | 26-Apr-2016 |
ozaki-r | Sweep unnecessary route.h inclusions
|
| 1.56 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.55 | 29-May-2014 |
rmind | branches: 1.55.4;
Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.54 | 25-Feb-2014 |
pooka | branches: 1.54.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.53 | 09-Jan-2012 |
liamjfoy | branches: 1.53.6; 1.53.10;
check against NULL
|
| 1.52 | 17-Jul-2011 |
joerg | branches: 1.52.2; 1.52.6;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.51 | 16-Sep-2009 |
pooka | Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.50 | 13-Sep-2009 |
pooka | Wipe out the last vestiges of POOL_INIT with one swift stroke. In
most cases, use a proper constructor. For proplib, give a local
equivalent of POOL_INIT for the kernel object implementation. This
way the code structure can be preserved, and a local link set is
not hazardous anyway (unless proplib is split to several modules,
but that'll be the day).
tested by booting a kernel in qemu and compile-testing i386/ALL
|
| 1.49 | 04-May-2008 |
thorpej | Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.48 | 24-Apr-2008 |
ad | branches: 1.48.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.47 | 23-Apr-2008 |
thorpej | Use <net/net_stats.h> / netstat_sysctl().
|
| 1.46 | 15-Apr-2008 |
thorpej | branches: 1.46.2;
Make IGMP stats per-cpu.
|
| 1.45 | 25-Apr-2007 |
dyoung | branches: 1.45.28;
Get rid of some gratuitous casts and join some lines.
|
| 1.44 | 12-Mar-2007 |
ad | branches: 1.44.2;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.43 | 05-Oct-2006 |
tls | branches: 1.43.4; 1.43.8;
Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.42 | 11-Dec-2005 |
christos | branches: 1.42.20; 1.42.22;
merge ktrace-lwp.
|
| 1.41 | 03-Feb-2005 |
perry | branches: 1.41.6;
ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.40 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.39 | 13-Nov-2004 |
christos | branches: 1.39.4; 1.39.6;
PR/25749: Peter Postma: missing splx() in kernel.
|
| 1.38 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.37 | 25-Apr-2004 |
simonb | Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.36 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.35 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.34 | 26-Jun-2003 |
itojun | branches: 1.34.2;
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.33 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.32 | 07-Nov-2002 |
thorpej | Fix signed/unsigned comparison warnings.
|
| 1.31 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.30 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.29 | 09-Jun-2002 |
itojun | whitespace
|
| 1.28 | 12-May-2002 |
matt | branches: 1.28.2; 1.28.4;
Eliminate commons.
|
| 1.27 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.26 | 25-Jul-2001 |
enami | Remove an obsolete comment.
|
| 1.25 | 16-Jun-2000 |
matt | branches: 1.25.2; 1.25.4; 1.25.6;
Don't copy M_EXT mbufs unless in "dhcp" mode. Do a mtod after the pullup
to make sure the ip pointer is still valid.
|
| 1.24 | 30-Mar-2000 |
augustss | branches: 1.24.2;
Remove register declarations.
|
| 1.23 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.22 | 09-Jul-1999 |
thorpej | branches: 1.22.2;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.21 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.20 | 25-Apr-1999 |
hwr | If the incoming code value is 0, timer gets 0, which would mean
a divide by zero afterwards.
This is also, what Bill Fenner seems to have done in the FreeBSD igmp
code.
This should fix kern/6541.
|
| 1.19 | 19-Jan-1999 |
mycroft | branches: 1.19.2;
Don't screw with ip_len; just subtract from it where we actually use the
value.
|
| 1.18 | 13-Feb-1998 |
tls | branches: 1.18.6;
Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.17 | 12-Jan-1998 |
scottr | Use option header file for MROUTING
|
| 1.16 | 09-Sep-1996 |
mycroft | branches: 1.16.14;
Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.15 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.14 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.13 | 04-Jun-1995 |
mycroft | Don't cast things unnecessarily.
|
| 1.12 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.11 | 01-Jun-1995 |
mycroft | Add missing ntohl() in multicast test.
|
| 1.10 | 31-May-1995 |
mycroft | Implement IGMP v2. Based on the Multicast 3.5 distribution.
|
| 1.9 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.8 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.7 | 09-Jun-1994 |
brezak | Patch to fix ip cksum errors. From mccanne@ee.lbl.gov (Steven McCanne).
|
| 1.6 | 04-Jun-1994 |
mycroft | Modify the loopback checks to deal with multiple interfaces.
|
| 1.5 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.4 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.3 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.2 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.1 | 06-Dec-1993 |
hpeyerl | branches: 1.1.1;
multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.1 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.16.14.1 | 01-Oct-1998 |
cgd | pull up revision 1.18 from trunk. (tls)
|
| 1.18.6.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.19.2.1 | 26-Apr-1999 |
perry | branches: 1.19.2.1.2; 1.19.2.1.4;
pullup 1.19->1.20 (Heiko Rupp): fix handling of packets with 0 timer values.
|
| 1.19.2.1.4.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.19.2.1.4.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.19.2.1.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.19.2.1.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.19.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.22.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.24.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.25.6.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.25.6.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.25.6.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.25.6.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.25.4.6 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.25.4.5 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.25.4.4 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.25.4.3 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.25.4.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.25.4.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.25.2.1 | 04-Aug-2003 |
msaitoh | Pull up revision 1.34 (requested by itojun in ticket #50):
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.28.4.2 | 02-Jul-2003 |
tron | Apply patch (requested by itojun in ticket #1360):
Make it compilable.
|
| 1.28.4.1 | 30-Jun-2003 |
grant | Pull up revision 1.34 (requested by itojun in ticket #1341):
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.28.2.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.28.2.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.28.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.34.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.34.2.4 | 29-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.34.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.34.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.34.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.39.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.39.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.41.6.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.41.6.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.42.22.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.42.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.43.8.2 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.43.8.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.43.4.2 | 07-May-2007 |
yamt | sync with head.
|
| 1.43.4.1 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.44.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.45.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.46.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.48.2.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.48.2.2 | 16-Sep-2009 |
yamt | sync with head
|
| 1.48.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.52.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.52.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.52.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.53.10.1 | 18-May-2014 |
rmind | sync with head
|
| 1.53.6.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.53.6.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.54.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.55.4.5 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.55.4.4 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.55.4.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.55.4.2 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.55.4.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.61.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.61.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.63.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.64.6.3 | 25-May-2020 |
martin | Pull up following revision(s) (requested by christos in ticket #1549):
sys/netinet/igmp.c: revision 1.70
sys/kern/kern_time.c: revision 1.204
igmp_sendpkt() expects ip_output() to set 'imo.imo_multicast_ttl' into
'ip->ip_ttl'; but ip_output() won't if the target is not a multicast
address, meaning that the uninitialized 'ip->ip_ttl' byte gets sent to
the network. This leaks one byte of kernel heap.
Fix this by filling 'ip->ip_ttl' with a TTL of one.
Found by KMSAN.
-
Fix uninitialized memory access. Found by KMSAN.
|
| 1.64.6.2 | 13-Jul-2018 |
martin | Pull up following revision(s) via patch (requested by knakahara in ticket #905):
sys/netinet/ip_mroute.c: revision 1.160
sys/netinet6/in6_l2tp.c: revision 1.16
sys/net/if.h: revision 1.263
sys/netinet/in_l2tp.c: revision 1.15
sys/netinet/ip_icmp.c: revision 1.172
sys/netinet/igmp.c: revision 1.68
sys/netinet/ip_encap.c: revision 1.69
sys/netinet6/ip6_mroute.c: revision 1.129
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.64.6.1 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.66.2.3 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.66.2.2 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.66.2.1 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.68.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.69.4.1 | 18-May-2020 |
martin | Pull up following revision(s) (requested by maxv in ticket #915):
sys/netinet/igmp.c: revision 1.70
igmp_sendpkt() expects ip_output() to set 'imo.imo_multicast_ttl' into
'ip->ip_ttl'; but ip_output() won't if the target is not a multicast
address, meaning that the uninitialized 'ip->ip_ttl' byte gets sent to
the network. This leaks one byte of kernel heap.
Fix this by filling 'ip->ip_ttl' with a TTL of one.
Found by KMSAN.
|
| 1.15 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.14 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.13 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.12 | 29-May-2014 |
rmind | branches: 1.12.40;
Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.11 | 25-Dec-2007 |
perry | branches: 1.11.54; 1.11.70;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.10 | 10-Dec-2005 |
elad | branches: 1.10.46; 1.10.52; 1.10.56; 1.10.60;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.9 | 07-Aug-2003 |
agc | branches: 1.9.16;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.8 | 20-Nov-1999 |
thorpej | branches: 1.8.28;
Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.7 | 10-Feb-1998 |
perry | branches: 1.7.14; 1.7.20;
add/cleanup multiple inclusion protection.
|
| 1.6 | 31-May-1995 |
mycroft | Implement IGMP v2. Based on the Multicast 3.5 distribution.
|
| 1.5 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.4 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.3 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.2 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.1 | 06-Dec-1993 |
hpeyerl | branches: 1.1.1;
multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.1.1.1 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.7.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.7.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.8.28.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.8.28.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.8.28.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.8.28.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.9.16.2 | 21-Jan-2008 |
yamt | sync with head
|
| 1.9.16.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.10.60.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.10.56.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.10.52.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.10.46.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.11.70.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.11.54.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.12.40.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.27 | 17-Feb-2021 |
christos | - pass the alignment instead of the mask (as Roy asked and to match the
other macro)
- use alignof to determine that alignment and CTASSERT what we expect
- remove unused macros
|
| 1.26 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.25 | 14-Sep-2018 |
maxv | branches: 1.25.12;
Use non-variadic function pointer in protosw::pr_input.
|
| 1.24 | 29-May-2014 |
rmind | branches: 1.24.26; 1.24.28;
Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.23 | 15-Apr-2008 |
thorpej | branches: 1.23.48; 1.23.64;
Make IGMP stats per-cpu.
|
| 1.22 | 10-Dec-2005 |
elad | branches: 1.22.70;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.21 | 25-Apr-2004 |
simonb | branches: 1.21.12;
Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.20 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.19 | 18-Apr-2004 |
matt | De __P()
|
| 1.18 | 07-Oct-2003 |
mycroft | There is also no reason to use arc4random() here.
|
| 1.17 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.16 | 26-Jun-2003 |
itojun | branches: 1.16.2;
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.15 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.14 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.13 | 29-May-2002 |
itojun | use arc4random
|
| 1.12 | 12-May-2002 |
matt | branches: 1.12.2; 1.12.4;
Eliminate commons.
|
| 1.11 | 19-Nov-1999 |
bouyer | branches: 1.11.4; 1.11.6; 1.11.8;
Update protocoles and interfaces stats counters to 64bit.
RTM_IFINFO is now 0xf, 0xe is RTM_OIFINFO which returns the old (if_msghdr14)
struct with 32bit counters (binary compat, conditioned on COMPAT_14).
Same for sysctl: node 3 is renamed NET_RT_OIFLIST, NET_RT_IFLIST is now node 4.
Change rt_msg1() to add an mbuf to the mbuf chain instead of just panic()
when the message is larger than MHLEN.
|
| 1.10 | 10-Feb-1998 |
perry | branches: 1.10.14; 1.10.20;
add/cleanup multiple inclusion protection.
|
| 1.9 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.8 | 31-May-1995 |
mycroft | Implement IGMP v2. Based on the Multicast 3.5 distribution.
|
| 1.7 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.6 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.5 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.4 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.3 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.2 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.1 | 06-Dec-1993 |
hpeyerl | branches: 1.1.1;
multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.1.1.1 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.10.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.10.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.11.8.2 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.11.8.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.11.6.2 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.11.6.1 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.11.4.1 | 04-Aug-2003 |
msaitoh | Pullup revision 1.16 (requested by itojun in ticket #50):
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.12.4.1 | 30-Jun-2003 |
grant | Pull up revision 1.16 (requested by itojun in ticket #1341):
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.12.2.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.12.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.16.2.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.16.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.16.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.16.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.21.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.22.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.23.64.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.23.48.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.24.28.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.24.26.1 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.25.12.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.248 | 20-Aug-2024 |
ozaki-r | inet: send GARP on link up if DAD is disabled
This behavior was accidentally removed at rev 1.233.
|
| 1.247 | 25-Nov-2022 |
knakahara | branches: 1.247.2; 1.247.8;
Support explicit unnumbered interface.
Currently, NetBSD supports implicit unnumbered interface by setting
the same IP address to two interfaces. However, such interface is not
treated as unnumbered when one of the interfaces is being changed and
has been changed IP address. That behavior can be harmful for some
routing daemons.
|
| 1.246 | 19-Nov-2022 |
yamt | Make arp have its own mowner
This helped me to debug mbuf leaks in arp.
(if_arp.c rev. 1.298)
|
| 1.245 | 17-Nov-2022 |
knakahara | Fix sending broken RTM_DELADDR message in some operations.
Here is mininum reproduction operation.
|
| 1.244 | 04-Nov-2022 |
ozaki-r | inpcb: use in_port_t for port numbers
|
| 1.243 | 20-Sep-2022 |
knakahara | Remove routes on an address removal if the routes referencing to the address. Implemented by ozaki-r@n.o.
A route that has a gateway is on a connected route can be invalid if the
connected route is deleted, i.e., an associated address is removed.
Traditionally NetBSD doesn't sweep such a route on the address removal. Sending
packets over the route fails with "No route to host". Also the route holds an
orphan ifaddr as rt_ifa that is destructed say by in_purgeaddr.
If the same address is assgined again in such a state, there can be two
different ifaddr objects with the same address. Until recently it's not a
big problem because we can send packets anyway. However after MP-ification
of the network stack, we can't send packets because we strictly check if rt_ifa
(i.e., the (old) ifaddr) is valid.
This change automatically removes such routes on a removal of an associated
address to avoid keeping inconsistent routes.
|
| 1.242 | 21-Sep-2021 |
christos | don't opencode kauth_cred_get()
|
| 1.241 | 29-Sep-2020 |
roy | inet: Treat LINK_STATE_UNKNOWN as LINK_STATE_UP when changing
It's something we have always done.
it's really rare for anything to transition to UNKNOWN from either
UP or DOWN, but technically it is possible.
|
| 1.240 | 11-Sep-2020 |
roy | inet: Add SIOCGNBRINFO to retrieve neighbor state about an address
|
| 1.239 | 11-Sep-2020 |
roy | in: No need to set expire here anymore
|
| 1.238 | 29-Aug-2020 |
christos | Partially revert previous: set RTF_HOST regardless of mask for point-to-point
links. Unbreaks IPSEC/L2TP configurations.
|
| 1.237 | 20-Aug-2020 |
riastradh | [ozaki-r] Changes to the kernel core for wireguard
|
| 1.236 | 18-Dec-2019 |
roy | inet: Add support for IPv4 /31 prefixes, as described in RFC 3021.
To run a /31 network, participating hosts MUST drop support
for directed broadcasts, and treat the first and last addresses
on subnet as unicast. The broadcast address for the prefix
should be the link local broadcast address, INADDR_BROADCAST.
Taken from FreeBSD, r226402.
Fixes PR kern/51388.
|
| 1.235 | 25-Sep-2019 |
ozaki-r | Make panic messages more informative
|
| 1.234 | 29-Apr-2019 |
roy | branches: 1.234.2;
rtsock: Route address message simplification
Rename rt_newaddrmsg to rt_addrmsg_rt.
Add rt_addrmsg which drops the error and route arguments which are only
needed by one caller.
|
| 1.233 | 29-Nov-2018 |
ozaki-r | Don't run DAD on link-up if it's explicitly disabled
|
| 1.232 | 29-Nov-2018 |
ozaki-r | Introduce and use ip_dad_enabled() and ip6_dad_enabled() functions
|
| 1.231 | 13-May-2018 |
khorben | branches: 1.231.2;
Fix spello in a comment
|
| 1.230 | 24-Apr-2018 |
knakahara | Fix sys/netinet/in.c:r1.229 problem. I have missed FALLTHROUGH, sorry.
|
| 1.229 | 20-Apr-2018 |
knakahara | SIOCSIFDSTADDR uses struct ifreq instead of struct ifaddr or struct in_aliasreq.
SIOCSIFDSTADDR is not used by base package commands...
I checked sys/net*/* only.
|
| 1.228 | 08-Apr-2018 |
christos | Protect ip_dad_count with if NARP > 0 to fix compilation
|
| 1.227 | 06-Apr-2018 |
ozaki-r | Make GARP work again when DAD is disabled
The change avoids setting an IP address tentative on initializing it when the
IPv4 DAD is disabled (net.inet.ip.dad_count=0), which allows a GARP packet to be
sent (see arpannounce). This is the same behavior of NetBSD 7, i.e., before
introducing the IPv4 DAD.
Additionally do the same change to IPv6 DAD for consistency.
The change is suggested by roy@
|
| 1.226 | 06-Apr-2018 |
ozaki-r | Revert the previous two commits as per roy@'s request
It broke the ip_dad_count > 0 case unexpectedly.
|
| 1.225 | 06-Apr-2018 |
ozaki-r | Don't set IN_IFF_* flags to ia4_flags if DAD is disabled
This fix allows that a GARP packet is sent when adding an IP address to an
interface with IFF_UP on a kernel with IPv4 DAD is disabled
(net.inet.ip.dad_count=0), which is the same behavior of NetBSD 7, i.e.,
before introducing the IPv4 DAD.
|
| 1.224 | 06-Apr-2018 |
ozaki-r | Simplify; clear then set flags to ia4_flags (NFCI)
|
| 1.223 | 06-Mar-2018 |
ozaki-r | Use pool(9) for llentry allocations
llentry is easy to be leaked and pool suits for it because pool is usable to
detect leaks.
Also sweep unnecessary wrappers for llentry, in_llentry and in6_llentry.
|
| 1.222 | 06-Mar-2018 |
ozaki-r | Fix memory leaks on arp -d and ndp -d for static entries
We have to delete entries on in_lltable_delete and in6_lltable_delete
unconditionally. Note that we don't need to worry about LLE_IFADDR because
there is no such entries now.
|
| 1.221 | 06-Mar-2018 |
ozaki-r | Fix reference leaks of llentry
callout_reset and callout_halt can cancel a pending callout without telling us.
Detect a cancel and remove a reference by using callout_pending and
callout_stop (it's a bit tricy though, we can detect it).
While here, we can remove remaining abuses of mutex_owned for softnet_lock.
|
| 1.220 | 06-Mar-2018 |
ozaki-r | Add assertions
We must not destroy llentries holding mbufs.
|
| 1.219 | 24-Feb-2018 |
ozaki-r | branches: 1.219.2;
Avoid a deadlock between softnet_lock and IFNET_LOCK
A deadlock occurs because there is a violation of the rule of lock ordering;
softnet_lock is held with hodling IFNET_LOCK, which violates the rule.
To avoid the deadlock, replace softnet_lock in in_control and in6_control
with KERNEL_LOCK.
We also need to add some KERNEL_LOCKs to protect the network stack surely.
This is required, for example, for PR kern/51356.
Fix PR kern/53043
|
| 1.218 | 14-Feb-2018 |
maxv | Remove IFF_STATICARP, we don't support this, and the code is useless in its
current form.
ok ozaki-r@
|
| 1.217 | 08-Feb-2018 |
ozaki-r | Don't call lltable_purge_entries from in_if_down if ARP isn't enabled
Reported by bouyer@
|
| 1.216 | 19-Jan-2018 |
ozaki-r | Suppress noisy debugging outputs
Even if DEBUG they are too noisy under load.
|
| 1.215 | 15-Jan-2018 |
ozaki-r | Remove extra pserialize_perform from in_purgeaddr
It's already performed in ifa_remove. Note so there (in in6_unlink_ifa too).
|
| 1.214 | 10-Jan-2018 |
knakahara | add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
|
| 1.213 | 27-Dec-2017 |
ozaki-r | Don't pass rwlock to callout_halt
|
| 1.212 | 25-Dec-2017 |
ozaki-r | Fix wrong usage of psref_held
We can't use it for checking if a caller does NOT hold a given target.
If you want to do it you should have psref_not_held or something.
|
| 1.211 | 15-Dec-2017 |
ozaki-r | Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
|
| 1.210 | 17-Nov-2017 |
ozaki-r | Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
|
| 1.209 | 10-Nov-2017 |
ozaki-r | Fix a deadlock between a route update and lltable
It happens because rtalloc1 is called from lltable with holding
IF_AFDATA_WLOCK.
If a route update is in action, rtalloc1 would wait for its completion with
holding IF_AFDATA_WLOCK. At the same moment, a softint (e.g., arpintr) may try
to take IF_AFDATA_WLOCK and get stuck on it. Unfortunately the stuck softint
prevents the route update from progressing because the route update calls
psref_target_destroy that needs the softint to complete.
A resource allocation graph of the senario looks like this:
route update =(psref_target_destroy)=> softint => IF_AFDATA_WLOCK
=(rt_update_wait)=> route update
Fix the deadlock by pulling rtalloc1 out of the lltable codes inside
IF_AFDATA_WLOCK.
Note that the deadlock happens only if NET_MPSAFE is enabled.
|
| 1.208 | 10-Nov-2017 |
ozaki-r | Remove redundant KASSERTMSG
The function is static, has just one caller and the caller does the same check.
|
| 1.207 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.206 | 04-Aug-2017 |
uwe | Fix it's -> its in a comment.
|
| 1.205 | 22-Jun-2017 |
ozaki-r | Purge ARP/NDP entries on an interface when the interface is down
Fix PR kern/51179
|
| 1.204 | 22-Jun-2017 |
ozaki-r | Fix in_lltable_match_prefix
The function has not been used but will be used soon.
|
| 1.203 | 01-Jun-2017 |
chs | branches: 1.203.2;
remove checks for failure after memory allocation calls that cannot fail:
kmem_alloc() with KM_SLEEP
kmem_zalloc() with KM_SLEEP
percpu_alloc()
pserialize_create()
psref_class_create()
all of these paths include an assertion that the allocation has not failed,
so callers should not assert that again.
|
| 1.202 | 25-May-2017 |
ozaki-r | Fix that a fresh in_ifaddr is unexpectedly freed before activating it
An in_ifaddr object is initialized with refcnt=0 and the refcnt
is incremented when being enqueued to the lists. However before
enqueuing it, in_ifinit can hold and refelease a reference to
it, i.e., call ifaref and ifafree, resulting in that the object
is freed in ifafree because its refcnt is decremented to 0.
It can be reproduced by doing:
ifconfig tun0 create
ifconfig tun1 create
ifconfig tun0 10.1 10.2
ifconfig tun1 10.2 10.1
ifconfig # Cause a kernel panic (may depend on environmemts)
We need to initialize a created in_ifaddr object with refcnt=1
to make the object survive over in_ifinit.
The issue is found by ryo@
|
| 1.201 | 12-May-2017 |
ryo | replace in_fmtaddr() by IN_PRINT(), and delete function in_fmtaddr()
|
| 1.200 | 28-Apr-2017 |
ozaki-r | Don't output debugging logs just if DIAGNOSTIC
Also make log messages informative.
|
| 1.199 | 17-Mar-2017 |
roy | branches: 1.199.4;
Add the local route after finishing the configuration of the address.
This fixes the issue where the initial address announced had an
invalid broadcast address.
|
| 1.198 | 02-Mar-2017 |
ozaki-r | Protect ia_allhosts by in_ifaddr_lock
|
| 1.197 | 23-Jan-2017 |
ozaki-r | Replace some splnet with splsoftnet
|
| 1.196 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.195 | 02-Jan-2017 |
christos | branches: 1.195.2;
- You can't just call the pfil hook to remove an address before an address
is removed! Hold a reference instead, remove it, and then free it.
- GC iatoifa()
|
| 1.194 | 31-Dec-2016 |
ryo | In the case of SIOCDIFADDR, call pfil_run_addrhooks before release ia.
|
| 1.193 | 27-Dec-2016 |
ozaki-r | Fix panic in pfil_run_hooks on bootup
XXX a kernel with pf still fails to boot up. Please someone fix it.
|
| 1.192 | 26-Dec-2016 |
knakahara | pserialize_perform() is required an additionally serialization. see pserialize(9).
ok by ozaki-r@n.o.
|
| 1.191 | 12-Dec-2016 |
ozaki-r | Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.190 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.189 | 06-Dec-2016 |
knakahara | add API to manipulate ifa->ia_hash and ia_hash_pslist_entry, and fix ia_hash_pslist_entry race by using them.
in_ifaddr_lock is required before writing ifa->ia_hash and
ia_hash_pslist_entry to serialize writer processings.
reviewed by ozaki-r@n.o.
|
| 1.188 | 18-Nov-2016 |
knakahara | We must use PSLIST_ENTRY_DESTROY after PSLIST_WRITER_REMOVE and waiting all readers done.
And then, if we want to re-insert the removed pslist element, we need to
call PSLIST_ENTERY_INIT again.
advised by riastradh@n.o and reviewed by ozaki-r@n.o, thanks.
|
| 1.187 | 18-Oct-2016 |
ozaki-r | Don't hold global locks if NET_MPSAFE is enabled
If NET_MPSAFE is enabled, don't hold KERNEL_LOCK and softnet_lock in
part of the network stack such as IP forwarding paths. The aim of the
change is to make it easy to test the network stack without the locks
and reduce our local diffs.
By default (i.e., if NET_MPSAFE isn't enabled), the locks are held
as they used to be.
Reviewed by knakahara@
|
| 1.186 | 01-Oct-2016 |
roy | Default netmask to /32 for INET on POINTOPOINT links if not specified.
|
| 1.185 | 29-Sep-2016 |
roy | in_ifscrub is no longer needed.
|
| 1.184 | 29-Sep-2016 |
roy | Set dstaddr in in_ifinit so that sppp consumers announce the correct
dstaddr in routing messages.
|
| 1.183 | 29-Sep-2016 |
roy | When changing an address via in_ifinit, ensure that the old address
is correctly scrubbed.
This allows sppp consumers to announce removal of the old address.
|
| 1.182 | 16-Sep-2016 |
roy | Drop hostIsNew from in_ifinit, let the function work out if the address
has changed.
Sync address flag setup with the IPv6 counterpart.
When scrubbing the address, or setting up the address fails, restore the
old address flags as well as the old address.
|
| 1.181 | 13-Sep-2016 |
christos | revert previous, roy says it breaks DaD.
|
| 1.180 | 13-Sep-2016 |
christos | When initializing addresses, reset the interface flags to 0. This fixes
an issue where point to point addresses that started down, and then came
up, were left with stale flags on one side of the point to point link.
|
| 1.179 | 01-Sep-2016 |
ozaki-r | Apply psz/psref to remaining IFADDR_READER_FOREACH
Pointed out by ryo@
|
| 1.178 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.177 | 28-Jul-2016 |
ozaki-r | Fix panic on adding/deleting IP addresses under network load
Adding and deleting IP addresses aren't serialized with other network
opeartions, e.g., forwarding packets. So if we add or delete an IP
address under network load, a kernel panic may happen on manipulating
network-related shared objects such as rtentry and rtcache.
To avoid such panicks, we still need to hold softnet_lock in in_control
and in6_control that are called via ioctl and do network-related operations
including IP address additions/deletions.
Fix PR kern/51356
|
| 1.176 | 20-Jul-2016 |
ozaki-r | Get rid of unnecessary satosin
|
| 1.175 | 14-Jul-2016 |
christos | branches: 1.175.2;
provide net.inet.multicast, like we have net.inet6.multicast to be used
by netstat.
|
| 1.174 | 13-Jul-2016 |
ozaki-r | Get rid of wrongly added TAILQ_INSERT_TAIL
|
| 1.173 | 08-Jul-2016 |
ozaki-r | Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.172 | 07-Jul-2016 |
ozaki-r | Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.171 | 06-Jul-2016 |
ozaki-r | Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.170 | 06-Jul-2016 |
ozaki-r | Add and use pslist(9)-based hashtable for IPv4 addresses
Note that we leave the old hashtable to keep vmstat -H working.
|
| 1.169 | 30-Jun-2016 |
ozaki-r | Make sure that ifaddr is published after its initialization finished
Basically we should insert an item to a collection (say a list) after
item's initialization has been completed to avoid accessing an item
that is initialized halfway. ifaddr (in{,6}_ifaddr) isn't processed
like so and needs to be fixed.
In order to do so, we need to tweak {arp,nd6}_rtrequest that depend
on that an ifaddr is inserted during its initialization; they explore
interface's address list to determine that rt_getkey(rt) of a given
rtentry is in the list to know whether the route's interface should
be a loopback, which doesn't work after the change. To make it work,
first check RTF_LOCAL flag that is set in rt_ifa_addlocal that calls
{arp,nd6}_rtrequest eventually. Note that we still need the original
code for the case to remove and re-add a local interface route.
|
| 1.168 | 23-Jun-2016 |
ozaki-r | Fix typo in a comment
|
| 1.167 | 21-Jun-2016 |
ozaki-r | Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.166 | 27-May-2016 |
christos | make hostzerobroadcast default to "no".
|
| 1.165 | 04-Apr-2016 |
ozaki-r | Separate nexthop caches from the routing table
By this change, nexthop caches (IP-MAC address pair) are not stored
in the routing table anymore. Instead nexthop caches are stored in
each network interface; we already have lltable/llentry data structure
for this purpose. This change also obsoletes the concept of cloning/cloned
routes. Cloned routes no longer exist while cloning routes still exist
with renamed to connected routes.
Noticeable changes are:
- Nexthop caches aren't listed in route show/netstat -r
- sysctl(NET_RT_DUMP) doesn't return them
- If RTF_LLDATA is specified, it returns nexthop caches
- Several definitions of routing flags and messages are removed
- RTF_CLONING, RTF_XRESOLVE, RTF_LLINFO, RTF_CLONED and RTM_RESOLVE
- RTF_CONNECTED is added
- It has the same value of RTF_CLONING for backward compatibility
- route's -xresolve, -[no]cloned and -llinfo options are removed
- -[no]cloning remains because it seems there are users
- -[no]connected is introduced and recommended
to be used instead of -[no]cloning
- route show/netstat -r drops some flags
- 'L' and 'c' are not seen anymore
- 'C' now indicates a connected route
- Gateway value of a route of an interface address is now not
a L2 address but "link#N" like a connected (cloning) route
- Proxy ARP: "arp -s ... pub" doesn't create a route
You can know details of behavior changes by seeing diffs under tests/.
Proposed on tech-net and tech-kern:
http://mail-index.netbsd.org/tech-net/2016/03/11/msg005701.html
|
| 1.164 | 25-Feb-2016 |
ozaki-r | Use callout_halt instead of callout_stop
|
| 1.163 | 26-Nov-2015 |
ozaki-r | Fix build dependency of if_llatbl.c
if_llatbl.c is required if inet or inet6 is enabled. Depending on ether
doesn't suit for NDP case.
|
| 1.162 | 16-Nov-2015 |
ozaki-r | Add missing rtfree
|
| 1.161 | 31-Aug-2015 |
ozaki-r | Fix building kernels w/o ether
|
| 1.160 | 31-Aug-2015 |
ozaki-r | Fix building kernels w/o DIAGNOSTIC
|
| 1.159 | 31-Aug-2015 |
ozaki-r | Replace ARP cache (llinfo) with lltable/llentry
Highlights of the change are:
- Use llentry instead of llinfo to manage ARP caches
- ARP specific data are stored in the hashed list
of an interface instead of the global list (llinfo_arp)
- Fine-grain locking on llentry
- arptimer (callout) per ARP cache
- the global timer callout with the big locks can be
removed (though softnet_lock is still required for now)
- net.inet.arp.prune is now obsoleted
- it was the interval of the global timer callout
- net.inet.arp.refresh is now obsoleted
- it was a parameter that prevents expiration of active caches
- Removed to simplify the timer logic, but we may be able to
restore the feature if really needed
Proposed on tech-kern and tech-net.
|
| 1.158 | 31-Aug-2015 |
ozaki-r | Hook up lltable/llentry with the kernel (and rumpkernel)
It is built and initialized on bootup, but there is no user for now.
Most codes in in.c are imported from FreeBSD as well as lltable/llentry.
|
| 1.157 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.156 | 16-May-2015 |
roy | Separate ARP handling DAD from inet.
This is done by signalling the intent to try tentative addresses
and then clearing the intent once the address is setup.
When the ARP handler is installed (arp_ifinit) then it adds
dad start and stop functions to the address which are used instead
of calling ARP directly.
|
| 1.155 | 05-May-2015 |
roy | If we don't have ARP, don't set IN_IFF_TENTATIVE.
|
| 1.154 | 02-May-2015 |
joerg | Fix !ARP build.
|
| 1.153 | 02-May-2015 |
roy | Appease gcc.
|
| 1.152 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.151 | 26-Feb-2015 |
roy | Don't add local routes for the any address or p2p addresses where the address matches the destination.
|
| 1.150 | 26-Feb-2015 |
roy | Introduce the routing flag RTF_LOCAL to track local address routes.
Add functions rt_ifa_addlocal() and rt_ifa_remlocal() to add and remove
local routes for the address and announce the new address and route
to the routing socket.
Add in_ifaddlocal() and in_ifremlocal() to use these functions.
Rename in6_if{add,rem}loop() to in6_if{add,rem}local() and use these
functions.
rtinit() no longer announces the address, just the network route for the
address. As such, calls to rt_newaddrmsg() have been removed from
in_addprefix() and in_scrubprefix().
This solves the problem of potentially more than one announcement, or no
announcement at all for the address in certain situations.
|
| 1.149 | 01-Dec-2014 |
christos | Don't pass junk in sin_family and sin_len for SIOCGIFNETMASK, and explain why.
XXX: pullup 7?
|
| 1.148 | 09-Sep-2014 |
rmind | branches: 1.148.2;
Eliminate IFAREF() and IFAFREE() macros in favour of functions.
|
| 1.147 | 01-Jul-2014 |
rtr | branches: 1.147.2;
fix parameter types in pr_ioctl, called xx_control() functions and remove
abuse of pointer to struct mbuf type.
param2 changed to u_long type and uses parameter name 'cmd' (ioctl command)
param3 changed to void * type and uses parameter name 'data'
param4 changed to struct ifnet * and uses parameter name 'ifp'
param5 has been removed (formerly struct lwp *) and uses of 'l' have been
replaced with curlwp from curproc(9).
callers have had (now unnecessary) casts to struct mbuf * removed, called
code has had (now unnecessary) casts to u_long, void * and struct ifnet *
respectively removed.
reviewed by rmind@
|
| 1.146 | 29-May-2014 |
rmind | Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.145 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.144 | 29-Jun-2013 |
rmind | branches: 1.144.4;
- Rewrite parts of pfil(9): use array to store hooks and thus be more cache
friendly (there are only few hooks in the system). Make the structures
opaque and the interface more strict.
- Remove PFIL_HOOKS option by making pfil(9) mandatory.
|
| 1.143 | 08-Jun-2012 |
gdt | branches: 1.143.2; 1.143.4;
Simply use the ifa_addr pointer, rather than taking its address.
Resolves failure to match addresses in SIOC[GD]LIFADDR calls.
Diagnosis and fix is due to Mark Keaton of BBN.
|
| 1.142 | 12-Dec-2011 |
roy | branches: 1.142.2;
When adding or scrubbing a prefix, always notify userland even if the
prefix does not have IFA_ROUTE.
Don't scrub the interface in SIOCAIFADDR if the new address does't
have IFA_ROUTE. If more functions are added to in_ifscrub then this logic
might need to be revisited.
Fixes PR/26450.
|
| 1.141 | 19-Nov-2011 |
tls | branches: 1.141.2;
First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>. This change includes
the following:
An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are
fixed. Some effort is made to accumulate entropy more quickly at
boot time.
A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.
The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.
An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c). This
is based on Greg Rose's implementation from Qualcomm.
A new random stream generator, nist_ctr_drbg, is provided. It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim. This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.
An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness. The arc4random/arc4randbytes API is deprecated for
in-kernel use. It is replaced by "cprng_strong". The current
cprng_fast implementation wraps the existing arc4random
implementation. The current cprng_strong implementation wraps the
new CTR_DRBG implementation. Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.
In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data(). Performance is expected to improve.
The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.
The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot. There is
approximately a 3/10000 chance of a false positive from these
tests. Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.
A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32). A new kernel will require a new rndctl.
The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.
Manual pages for the new kernel interfaces are forthcoming.
|
| 1.140 | 28-Oct-2011 |
dyoung | branches: 1.140.2;
Remove the #if 1 / #endif around some code that appears to be
responsible deleting the 'first' AF_INET address on the interface if the
target address has family == AF_UNSPEC.
|
| 1.139 | 19-Oct-2011 |
dyoung | Use if_addr_init() and if_mcast_op() instead of ifp->if_ioctl().
|
| 1.138 | 15-May-2010 |
oki | Backout rev.1.137. It causes troubles, see PR kern/43294.
We needs more discussion/a more general solution.
|
| 1.137 | 12-Mar-2010 |
oki | branches: 1.137.2;
Fixed a number of race conditions in the case of receiving ipv4 packet.
found by iij seil team.
|
| 1.136 | 07-Dec-2009 |
dyoung | branches: 1.136.2;
Initialize/compare pointers with NULL instead of 0.
|
| 1.135 | 11-Sep-2009 |
dyoung | Make ifconfig(8) set and display preference numbers for IPv6
addresses. Make the kernel support SIOC[SG]IFADDRPREF for IPv6
interface addresses.
In in6ifa_ifpforlinklocal(), consult preference numbers before
making an otherwise arbitrary choice of in6_ifaddr. Otherwise,
preference numbers are *not* consulted by the kernel, but that will
be rather easy for somebody with a little bit of free time to fix.
Please note that setting the preference number for a link-local
IPv6 address does not work right, yet, but that ought to be fixed
soon.
In support of the changes above,
1 Add a method to struct domain for "externalizing" a sockaddr, and
provide an implementation for IPv6. Expect more work in this area: it
may be more proper to say that the IPv6 implementation "internalizes"
a sockaddr. Add sockaddr_externalize().
2 Add a subroutine, sofamily(), that returns a struct socket's address
family or AF_UNSPEC.
3 Make a lot of IPv4-specific code generic, and move it from
sys/netinet/ to sys/net/ for re-use by IPv6 parts of the kernel and
ifconfig(8).
|
| 1.134 | 18-Apr-2009 |
tsutsui | Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.133 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.132 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.131 | 12-Feb-2009 |
christos | PR/40603: Christoph Badura: unprivileged users can add and delete interface
link addresses. Fixed by centralizing the test as suggested. Will pull up
to 5.0 once submitter tests the fix.
|
| 1.130 | 21-Dec-2008 |
roy | branches: 1.130.2;
The automatic addition of a subnet route should not error if a manually
added route already exists. Fixes PR kern/40133.
|
| 1.129 | 17-Dec-2008 |
cegger | kill MALLOC and FREE macros.
|
| 1.128 | 07-Nov-2008 |
dyoung | *** Summary ***
When a link-layer address changes (e.g., ifconfig ex0 link
02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor
Advertisement to update the network-/link-layer address bindings
on our LAN peers.
Refuse a change of ethernet address to the address 00:00:00:00:00:00
or to any multicast/broadcast address. (Thanks matt@.)
Reorder ifnet ioctl operations so that driver ioctls may inherit
the functions of their "class"---ether_ioctl(), fddi_ioctl(), et
cetera---and the class ioctls may inherit from the generic ioctl,
ifioctl_common(), but both driver- and class-ioctls may override
the generic behavior. Make network drivers share more code.
Distinguish a "factory" link-layer address from others for the
purposes of both protecting that address from deletion and computing
EUI64.
Return consistent, appropriate error codes from network drivers.
Improve readability. KNF.
*** Details ***
In if_attach(), always initialize the interface ioctl routine,
ifnet->if_ioctl, if the driver has not already initialized it.
Delete if_ioctl == NULL tests everywhere else, because it cannot
happen.
In the ioctl routines of network interfaces, inherit common ioctl
behaviors by calling either ifioctl_common() or whichever ioctl
routine is appropriate for the class of interface---e.g., ether_ioctl()
for ethernets.
Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In
the user->kernel interface, SIOCSIFADDR's argument was an ifreq,
but on the protocol->ifnet interface, SIOCSIFADDR's argument was
an ifaddr. That was confusing, and it would work against me as I
make it possible for a network interface to overload most ioctls.
On the protocol->ifnet interface, replace SIOCSIFADDR with
SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to
invoke SIOCINITIFADDR.
In ifioctl(), give the interface the first shot at handling most
interface ioctls, and give the protocol the second shot, instead
of the other way around. Finally, let compatibility code (COMPAT_OSOCK)
take a shot.
Pull device initialization out of switch statements under
SIOCINITIFADDR. For example, pull ..._init() out of any switch
statement that looks like this:
switch (...->sa_family) {
case ...:
..._init();
...
break;
...
default:
..._init();
...
break;
}
Rewrite many if-else clauses that handle all permutations of IFF_UP
and IFF_RUNNING to use a switch statement,
switch (x & (IFF_UP|IFF_RUNNING)) {
case 0:
...
break;
case IFF_RUNNING:
...
break;
case IFF_UP:
...
break;
case IFF_UP|IFF_RUNNING:
...
break;
}
unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and
#ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4).
In ipw(4), remove an if_set_sadl() call that is out of place.
In nfe(4), reuse the jumbo MTU logic in ether_ioctl().
Let ethernets register a callback for setting h/w state such as
promiscuous mode and the multicast filter in accord with a change
in the if_flags: ether_set_ifflags_cb() registers a callback that
returns ENETRESET if the caller should reset the ethernet by calling
if_init(), 0 on success, != 0 on failure. Pull common code from
ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(),
and register if_flags callbacks for those drivers.
Return ENOTTY instead of EINVAL for inappropriate ioctls. In
zyd(4), use ENXIO instead of ENOTTY to indicate that the device is
not any longer attached.
Add to if_set_sadl() a boolean 'factory' argument that indicates
whether a link-layer address was assigned by the factory or some
other source. In a comment, recommend using the factory address
for generating an EUI64, and update in6_get_hw_ifid() to prefer a
factory address to any other link-layer address.
Add a routing message, RTM_LLINFO_UPD, that tells protocols to
update the binding of network-layer addresses to link-layer addresses.
Implement this message in IPv4 and IPv6 by sending a gratuitous
ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD
messages on a change of an interface's link-layer address.
In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address
that is broadcast/multicast or equal to 00:00:00:00:00:00.
Make ether_ioctl() call ifioctl_common() to handle ioctls that it
does not understand.
In gif(4), initialize if_softc and use it, instead of assuming that
the gif_softc and ifp overlap.
Let ifioctl_common() handle SIOCGIFADDR.
Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels
that certain invariants on a struct route are satisfied.
In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit
about the ioctls that we do not allow on an agr(4) member interface.
bzero -> memset. Delete unnecessary casts to void *. Use
sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with
NULL instead of "testing truth". Replace some instances of (type
*)0 with NULL. Change some K&R prototypes to ANSI C, and join
lines.
|
| 1.127 | 25-Sep-2008 |
pooka | branches: 1.127.2; 1.127.4;
Don't wrap whole file in INET.
|
| 1.126 | 11-May-2008 |
dyoung | branches: 1.126.4;
Cosmetic: compare sa_family with AF_UNSPEC instead of testing truth.
Join a line. Compare sa_len with 0 instead of testing truth.
|
| 1.125 | 28-Apr-2008 |
martin | branches: 1.125.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.124 | 10-Apr-2008 |
dyoung | branches: 1.124.2; 1.124.4;
s/8/NBBY/
|
| 1.123 | 06-Feb-2008 |
matt | branches: 1.123.6;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.122 | 06-Dec-2007 |
dyoung | Use ifa_insert(), ifa_remove().
|
| 1.121 | 05-Dec-2007 |
dyoung | Extract common code, creating a subroutine if_purgeaddrs(ifp,
family, purgeaddr) which applies function `purgeaddr' to each
address on `ifp' belonging to `family'.
|
| 1.120 | 05-Dec-2007 |
dyoung | Use IFADDR_FIRST() and IFADDR_NEXT().
|
| 1.119 | 09-Nov-2007 |
dyoung | branches: 1.119.2;
KNF. Remove superfluous casts and parentheses.
|
| 1.118 | 01-Sep-2007 |
dyoung | branches: 1.118.4; 1.118.6;
Use ifreq_setaddr(), ifreq_getaddr(), sockaddr_in_init(), and
sockaddr_copy(). Constify. Compare pointers with NULL, not 0.
Don't "test truth" of pointers, but compare with NULL.
|
| 1.117 | 15-Apr-2007 |
dyoung | branches: 1.117.2; 1.117.6; 1.117.8;
Cosmetic: shorten a staircase. bzero -> memset. KNF.
|
| 1.116 | 04-Mar-2007 |
christos | branches: 1.116.2; 1.116.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.115 | 16-Nov-2006 |
christos | branches: 1.115.4;
__unused removal on arguments; approved by core.
|
| 1.114 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.113 | 25-Oct-2006 |
elad | Kill some KAUTH_GENERIC_ISSUSER.
|
| 1.112 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.111 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.110 | 23-Sep-2006 |
elad | PR/29766: Chris Ross: Incomplete correction of comments in netinet/in.c
Patch applied, thanks!
|
| 1.109 | 23-Jul-2006 |
ad | branches: 1.109.4; 1.109.6;
Use the LWP cached credentials where sane.
|
| 1.108 | 14-May-2006 |
elad | integrate kauth.
|
| 1.107 | 10-May-2006 |
mrg | quell GCC 4.1 uninitialised variable warnings.
XXX: we should audit the tree for which old ones are no longer needed
after getting the older compilers out of the tree..
|
| 1.106 | 11-Dec-2005 |
christos | branches: 1.106.4; 1.106.6; 1.106.8; 1.106.10; 1.106.12;
merge ktrace-lwp.
|
| 1.105 | 28-Sep-2005 |
seanb | - Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.104 | 26-Feb-2005 |
perry | branches: 1.104.2; 1.104.4; 1.104.6;
nuke trailing whitespace
|
| 1.103 | 03-Feb-2005 |
perry | ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.102 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.101 | 24-Jan-2005 |
matt | branches: 1.101.2;
Add IFNET_FOREACH and IFADDR_FOREACH macros and start using them.
|
| 1.100 | 08-Aug-2004 |
yamt | branches: 1.100.4;
in_control: fix address leaks on error, which causes a panic
("no domain for AF 0") on if_detach.
- SIOCAIFADDR, SIOCSIFADDR: free an address on error.
- SIOCSIFNETMASK, SIOCSIFDSTADDR: reject operations for an interface which
has no AF_INET addresses.
partly from OpenBSD and FreeBSD.
reviewed by Christos Zoulas on tech-net@.
|
| 1.99 | 26-Jul-2004 |
yamt | call PFIL_NEWIF hooks at a correct place.
(on SIOCAIFADDR rather than SIOCGIFALIAS.)
from Peter Postma, PR/26402.
ok'ed by itojun.
|
| 1.98 | 18-Jul-2004 |
yamt | fix typos. PFIL_HOOK -> PFIL_HOOKS
|
| 1.97 | 07-Jul-2004 |
mycroft | Fix SIOCSIFNETMASK -- it needs to use in_ifscrub() and in_ifinit() to update
the interface route and various internal state. Also, it should use an ifreq,
not an if_aliasreq. Addresses PR 9604. (Nothing in our source tree uses
SIOCSIFNETMASK, though. Perhaps it should be deprecated.)
|
| 1.96 | 22-Jun-2004 |
itojun | prepare PF-related hooks. reviewed by matt, perry, christos
|
| 1.95 | 30-May-2004 |
itojun | fix SIOC*LIFADDR for IPv4. markus friedl
|
| 1.94 | 21-Apr-2004 |
itojun | kill sprintf, use snprintf
|
| 1.93 | 11-Nov-2003 |
jonathan | branches: 1.93.2;
Change global head-of-local-IP-address list from in_ifaddr to
in_ifaddrhead. Recent changes in struct names caused a namespace
collision in fast-ipsec, which are most cleanly fixed by using
"in_ifaddrhead" as the listhead name.
|
| 1.92 | 23-Oct-2003 |
mycroft | Remove all the code to maintain ia_inpcbs. This information was only used to
close sockets on address changes, which was deemed to be a bad idea and was
summarily removed, so there is no point in wasting effort on maintaining it
any more.
|
| 1.91 | 16-Aug-2003 |
itojun | do not disconnect L4 connections on IP address removal. the behavior
is too extreme (consider DHCP/PPP-based fixed address allocation).
see tech-net for more info.
|
| 1.90 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.89 | 26-Jun-2003 |
itojun | branches: 1.89.2;
cleanup multicast routing stuff on if_detach().
XXX sideeffect to running instance of multicast routing daemon unknown
|
| 1.88 | 26-Jun-2003 |
itojun | put meaningful count into in_multientries.
(or we could remove this variable - noone seem to use it)
|
| 1.87 | 26-Jun-2003 |
itojun | purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.86 | 26-Jun-2003 |
itojun | tabify
|
| 1.85 | 18-Jun-2003 |
itojun | install host route for p2p interface even if there's connected net route
by broadcast interface. PR 21903.
|
| 1.84 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.83 | 13-Jun-2003 |
onoe | For loopback interface, assign ia_dstaddr instead of just changing reference
to ifa_dstaddr. This fixes the problem that assigning more than 2 IPv4
aliases to loopback interface fails to create routing table entry.
|
| 1.82 | 07-Nov-2002 |
thorpej | Fix signed/unsigned comparison warnings.
|
| 1.81 | 22-Oct-2002 |
simonb | "newifaddr" in in_control() was set but never used, remove it.
|
| 1.80 | 11-Sep-2002 |
itojun | KNF - return is not a function. sync w/kame.
|
| 1.79 | 04-Sep-2002 |
itojun | backout 1.78, ioctl(SIOCSIFADDR) is needed to test if the interface
supports AF_INET or not
|
| 1.78 | 04-Sep-2002 |
itojun | avoid SIOCSIFADDR if there's an IPv4 address already.
the comment doesn't match the behavior, it seems that the code assumed that
there's only one IPv4 address on an interface. sync w/kame
|
| 1.77 | 09-Jun-2002 |
itojun | whitespace
|
| 1.76 | 09-May-2002 |
itojun | branches: 1.76.2; 1.76.4;
backout 1.72. it is not correct for the kernel to remove routes by itself,
and the code was buggy (dereferenced null pointer when IFAFREE removes the
route).
|
| 1.75 | 30-Mar-2002 |
itojun | do not consider /32 address itself as broadcast.
with /32 address, in_addr == in_broadaddr.
|
| 1.74 | 01-Mar-2002 |
thorpej | In in_savemkludge() and in_restoremkludge(), don't insert into a new
list without removing from the old one first.
From Matt Thomas.
|
| 1.73 | 21-Feb-2002 |
christos | Sean amended his patch not to include the IFAFREE()
|
| 1.72 | 21-Feb-2002 |
christos | PR/15662: Sean Boudreau: make sure we clean all routes of an interface when
we change its ip address.
|
| 1.71 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.70 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.69 | 16-Sep-2001 |
martin | branches: 1.69.2;
Fix typo in comment.
|
| 1.68 | 27-Jul-2001 |
itojun | branches: 1.68.2;
do not check in_dstaddr on in_{add,scrub}prefix, otherwise linklocal
address manipulation could choke. sync with kame
|
| 1.67 | 22-Jul-2001 |
itojun | manage IFA_ROUTE on interface address better, so that we can
provide a better support for multiple address with the same prefix better.
(like 10.0.0.1/8 and 10.0.0.2/8 on the same interface)
continuation of PR 13311.
remove irrelevant #if 0'ed segment for PR 10427.
|
| 1.66 | 13-Apr-2001 |
thorpej | branches: 1.66.2;
Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.65 | 08-Oct-2000 |
enami | branches: 1.65.2;
- Keep track of allhost multicast address record we joined into
each in_ifaddr and delete it when an address is purged.
- Don't simply try to delete a multicast address record listed in the
ia_multiaddrs. It results a dangling pointer. Let who holds a
reference to it to delete it.
|
| 1.64 | 08-Oct-2000 |
itojun | implement multicast kludge table for IPv4.
- when all the interface address is removed from an interface, and there's
multicast groups still left joined, keep it in kludge table.
- when an interface address is added again, recover multicast groups from
kludge table.
this will avoid problem with dangling in_ifaddr on pcmcia card removal,
due to the link from multicast group info (in_multi).
the code is basically from sys/netinet6/in6.c (jinmei@kame).
pointed out by: Shiva Shenoy <shiva_s@yahoo.com>
|
| 1.63 | 06-Oct-2000 |
itojun | remove obsolete handling code for SIOCSIFPHY*. they are now in ifioctl().
sync with kame.
|
| 1.62 | 02-Aug-2000 |
itojun | inhibit error code from rtinit(). this happens when we try to assign
multiple addresses from same prefix, onto single interface. PR 10427.
more info:
- 4.4BSD did not check return code from in_ifinit() at all.
4.4BSD does not support multiple address from same prefix.
- past KAME change passed in{,6}_ifinit() to upwards, toward ifconfig(8).
the behavior is filed as PR 10427.
- the commit inhibits EEXIST from rtinit(), hence partially recovers old
4.4BSD behavior.
- the right thing to happen is to properly support multiple address assignment
from the same prefix. KAME tree has more extensive change, however, it needs
much more time to get stabilized (rtentry refcnt change can cause serious
issue, we really need to bake it before bring it to netbsd)
|
| 1.61 | 06-May-2000 |
mycroft | branches: 1.61.4;
GC in_interfaces.
|
| 1.60 | 03-Apr-2000 |
enami | Bump the reference count of ifaddr while it is refered through in_multi.
|
| 1.59 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.58 | 21-Mar-2000 |
itojun | improve comment (about undo'ing code on in{,6}_ifinit failure)
|
| 1.57 | 18-Mar-2000 |
itojun | #if 0'ed undo code for interface address addition failure.
it was a bit too strong, and forbids multiple addresses from
same prefix to be assigned.
now the behavior is the same as previous - memory leak on interface address
addition failure.
http://orange.kame.net/dev/query-pr.cgi?pr=218
|
| 1.56 | 12-Mar-2000 |
itojun | undo interface address addition attempt, when in_ifinit fails.
(this basically avoids memory leakage)
|
| 1.55 | 06-Mar-2000 |
itojun | allow SIOCDIFADDR with AF_UNSPEC address by default, until we fix ifconfig(8).
(should be COMPAT_43)
|
| 1.54 | 25-Feb-2000 |
itojun | allow AF_UNSPEC for SIOCDIFADDR. ISC DHCP client depends on this behavior.
|
| 1.53 | 25-Feb-2000 |
itojun | backout previous commit (sanity check for family) - it seems to be doing
something wrong. i'll revise it soon.
|
| 1.52 | 25-Feb-2000 |
itojun | reject non-AF_INET addresses on ioctl.
without this, we can configure invalid sockaddrs, for example,
sa_family == 0 (and we can never remove them!)
|
| 1.51 | 02-Feb-2000 |
thorpej | PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.50 | 01-Feb-2000 |
thorpej | First-draft if_detach() implementation, originally from Bill Studnemund,
although this version has been changed somewhat:
- reference counting on ifaddrs isn't as complete as Bill's original
work was. This is hard to get right, and we should attack one
protocol at a time.
- This doesn't do reference counting or dynamic allocation of ifnets yet.
- This version introduces a new PRU -- PRU_PURGEADDR, which is used to
purge an ifaddr from a protocol. The old method Bill used didn't work
on all protocols, and it only worked on some because it was Very Lucky.
This mostly works ... i.e. works for my USB Ethernet, except for a dangling
ifaddr reference left by the IPv6 code; have not yet tracked this down.
|
| 1.49 | 12-Dec-1999 |
itojun | do not use member of sockaddr_storage directly.
(if the definition is like in rfc2553) they are not supposed to be used.
XXX i'm trying to change rfc2553 sockaddr_storage definition to include
"ss_len" and "ss_family". see ipngwg. situation might change soon.
|
| 1.48 | 01-Jul-1999 |
itojun | branches: 1.48.2; 1.48.8;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.47 | 26-Jun-1999 |
sommerfeld | If the new global variable hostzerobroadcast is zero, no longer assume
address zero of each net/subnet is a broadcast address.
(The default value is nonzero, which preserves the current behavior).
This can be set using sysctl; the boot-time default can also be
configured using the HOSTZEROBROADCAST kernel config option.
While we're here, defopt HOSTZEROBROADCAST and SUBNETSARELOCAL
|
| 1.46 | 19-Dec-1998 |
thorpej | branches: 1.46.4; 1.46.6;
Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.45 | 30-Sep-1998 |
tls | branches: 1.45.4;
Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.44 | 28-Sep-1998 |
christos | SIOCGIFALIAS should not be restricted to the superuser.
|
| 1.43 | 06-Sep-1998 |
christos | Add SIOCGIFALIAS
|
| 1.42 | 05-Jul-1998 |
jonathan | Garbage-collect ``needs-flag'' from attributes ether, fddi, arc:
NETHER, NFDDI, NARC are not used anywhere. Remove #include "ether.h",
which had no effect.
Removes clash with "options NATM" for native-ATM network protocol stack.
|
| 1.41 | 05-Jul-1998 |
jonathan | defopt INET, NETATALK.
|
| 1.40 | 29-May-1998 |
matt | Change arp so its console log messages print out IP addresses in
dotted quad format instead of hex.
|
| 1.39 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.38 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.37 | 12-Jan-1998 |
scottr | Use option header file for MROUTING
|
| 1.36 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.35 | 23-Jul-1997 |
thorpej | branches: 1.35.6;
Pull SYN_cache_branch down into the main line.
|
| 1.34 | 15-Mar-1997 |
is | branches: 1.34.2;
New ARP system, supports IPv4 over any hardware link.
Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will
only support Ethernet. Tcpdump itself should be ok, but libpcap needs
lot of work.
For the detailed change history, look at the commit log entries for
the is-newarp branch.
|
| 1.33 | 14-Sep-1996 |
mrg | branches: 1.33.4;
move the packet filter hooks in to a saner location. while i'm here, rename
PACKET_FILTER to PFIL_HOOKS.
|
| 1.32 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.31 | 07-Sep-1996 |
mrg | fix a couple of minor nits after discussions with jason.
|
| 1.30 | 06-Sep-1996 |
mrg | add packet filter interface code. see pfil(9) for more details. you
need the PACKET_FILTER option to enable this code. currently, ipfilter
version 3.1.1-beta has been converted to use this new interface.
|
| 1.29 | 23-Jun-1996 |
mycroft | Return ENOPROTOOPT rather than picking pseudo-random error values.
Don't allow SIOCGET{VIF,SG}CNT from sockets other than the multicast router.
Restructure rip_ctloutput() like ip_ctloutput(), and fix memory leaks.
|
| 1.28 | 22-May-1996 |
mycroft | A few style changes to match netiso and netns.
|
| 1.27 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.26 | 13-Feb-1996 |
christos | branches: 1.26.4;
netinet prototypes
|
| 1.25 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.24 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.23 | 04-Jun-1995 |
mycroft | For consistency, set sin_len for SIOC{ADD,DEL}MULTI.
|
| 1.22 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.21 | 04-Jun-1995 |
mycroft | Clean up a lot of ugly casts.
|
| 1.20 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.19 | 31-May-1995 |
mycroft | Integrate multicast 3.5 distribution, with several bugs fixed and general
cleanup. This is a (working) snapshot of work in progress.
|
| 1.18 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.17 | 11-Apr-1995 |
mycroft | Remove some explicit references to loif.
|
| 1.16 | 10-Apr-1995 |
mycroft | Remove now unneeded #ifdef. Prototype new function.
|
| 1.15 | 03-Nov-1994 |
mycroft | Fix off by one error in in_socktrim(), reported by Karn Fox.
|
| 1.14 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.13 | 21-Jun-1994 |
chopps | config.new hack for if_ether.c for lack of an `and' in the grammer
and protect some ether specific code in in.c
|
| 1.12 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.11 | 14-Mar-1994 |
glass | add missing explicit type declaration for func argument
|
| 1.10 | 10-Feb-1994 |
mycroft | Deprecate af.h.
|
| 1.9 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.8 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.7 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.6 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.5 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.4 | 10-Jun-1993 |
deraadt | branches: 1.4.4;
patch from Yuval Yarom, sent to me by <andrew@werple.apana.org.au>
The check that the destination of a forwarded ip packet is not on
the loopback net is wrong, and will always fail. The following patch
fixes the problem.
[allows "route add $hostname localhost" to be added to /etc/netstart to
keep things for $hostname away from the ethernet driver]
|
| 1.3 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.4.4.2 | 08-Nov-1993 |
mycroft | Remove references to af.h.
|
| 1.4.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.26.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.33.4.2 | 09-Mar-1997 |
is | netinet/if_ether.h -> netinet/if_inarp.h
|
| 1.33.4.1 | 07-Feb-1997 |
is | Snapshot of new ARP code.
Our old ARP code was hardwired for 6-byte length medium
addresses, while the protocol is designed for any size.
This snapshot contains a first hack at getting rid of
Ethernet specific data structures. The ep driver is updated
(and tested on the PCI bus), the iy and fpa drivers have been
updated, but not real life tested yet.
If you want to test this with other drivers, you have to update
them first yourself, and probably tag the relevant directories.
Better contact me if you want to do this.
|
| 1.34.2.1 | 14-May-1997 |
mellon | Keep track of maximum mtu of non-local interface
|
| 1.35.6.1 | 01-Oct-1998 |
cgd | pull up revisions 1.38-1.39, 1.45 (via patch) from trunk. (tls)
|
| 1.45.4.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.46.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.46.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.46.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.46.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.48.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.48.2.2 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.48.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.61.4.6 | 30-Nov-2003 |
he | Pull up revision 1.89 via patch, requested by itojun in ticket #54:
Clean up multicast routing when an interface is detached.
|
| 1.61.4.5 | 04-Aug-2003 |
msaitoh | Pull up revision 1.87 (requested by itojun in ticket #50):
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.61.4.4 | 03-Apr-2002 |
he | Pull up revision 1.75 (requested by itojun):
Reject TCP SYN packets sent to the broadcast address.
|
| 1.61.4.3 | 17-Oct-2000 |
tv | Pullup 1.65 [enami]:
- Keep track of allhost multicast address record we joined into
each in_ifaddr and delete it when an address is purged.
- Don't simply try to delete a multicast address record listed in the
ia_multiaddrs. It results a dangling pointer. Let who holds a
reference to it to delete it.
Also 1.64 [itojun, req by enami]:
implement multicast kludge table for IPv4.
- when all the interface address is removed from an interface, and there's
multicast groups still left joined, keep it in kludge table.
- when an interface address is added again, recover multicast groups from
kludge table.
this will avoid problem with dangling in_ifaddr on pcmcia card removal,
due to the link from multicast group info (in_multi).
|
| 1.61.4.2 | 06-Oct-2000 |
itojun | pullup (approved by releng-1-5)
move privilege check for SIOCSIFPHY* from in{,6}_control to ifioctl.
fix privilege check mistakes (which allows non-root user to modify gif
physical address in some cases). sync with kame.
> cvs rdiff -r1.62 -r1.63 syssrc/sys/netinet/in.c
> cvs rdiff -r1.34 -r1.35 syssrc/sys/netinet6/in6.c
> cvs rdiff -r1.71 -r1.73 syssrc/sys/net/if.c
|
| 1.61.4.1 | 18-Aug-2000 |
itojun | pullup (approved by releng-1-5)
sys/netinet6/in6.c 1.33 -> 1.34
sys/netinet/in.c 1.61 -> 1.62
> inhibit error code from rtinit(). this happens when we try to assign
> multiple addresses from same prefix, onto single interface. PR 10427.
|
| 1.65.2.10 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.65.2.9 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.7 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.65.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.65.2.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.65.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.65.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.66.2.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.66.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.66.2.3 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.66.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.66.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.68.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.69.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.76.4.4 | 02-Apr-2006 |
riz | Pull up following revision(s) (requested by seanb in ticket #6040):
sys/netinet/in.c: revision 1.105
- Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.76.4.3 | 02-Jul-2003 |
tron | Apply patch (requested by itojun in ticket #1361):
Make it compilable.
|
| 1.76.4.2 | 30-Jun-2003 |
grant | Pull up revision 1.89 (requested by itojun in ticket #1342):
cleanup multicast routing stuff on if_detach().
|
| 1.76.4.1 | 30-Jun-2003 |
grant | Pull up revision 1.87 (requested by itojun in ticket #1341):
purge rti structure (in igmp.c) for removed ifp on if_detach().
|
| 1.76.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.89.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.89.2.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.89.2.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.89.2.3 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.89.2.2 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.89.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.93.2.2 | 02-Apr-2006 |
riz | Pull up following revision(s) (requested by seanb in ticket #10411):
sys/netinet/in.c: revision 1.105
- Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.93.2.1 | 10-Jul-2004 |
tron | branches: 1.93.2.1.2; 1.93.2.1.4;
Pull up revision 1.97 (requested by mycroft in ticket #615):
Fix SIOCSIFNETMASK -- it needs to use in_ifscrub() and in_ifinit() to update
the interface route and various internal state. Also, it should use an ifreq,
not an if_aliasreq. Addresses PR 9604. (Nothing in our source tree uses
SIOCSIFNETMASK, though. Perhaps it should be deprecated.)
|
| 1.93.2.1.4.1 | 02-Apr-2006 |
riz | Pull up following revision(s) (requested by seanb in ticket #10411):
sys/netinet/in.c: revision 1.105
- Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.93.2.1.2.1 | 02-Apr-2006 |
riz | Pull up following revision(s) (requested by seanb in ticket #10411):
sys/netinet/in.c: revision 1.105
- Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.100.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.101.2.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.101.2.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.104.6.1 | 02-Apr-2006 |
riz | Pull up following revision(s) (requested by seanb in ticket #1235):
sys/netinet/in.c: revision 1.105
- Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.104.4.6 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.104.4.5 | 07-Dec-2007 |
yamt | sync with head
|
| 1.104.4.4 | 15-Nov-2007 |
yamt | sync with head.
|
| 1.104.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.104.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.104.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.104.2.1 | 02-Apr-2006 |
riz | Pull up following revision(s) (requested by seanb in ticket #1235):
sys/netinet/in.c: revision 1.105
- Close NULL dereference when a GIFALIAS is performed on
a non existant address.
- Code review: christos
|
| 1.106.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.106.10.4 | 11-May-2006 |
elad | sync with head
|
| 1.106.10.3 | 06-May-2006 |
christos | - Move kauth_cred_t declaration to <sys/types.h>
- Cleanup struct ucred; forward declarations that are unused.
- Don't include <sys/kauth.h> in any header, but include it in the c files
that need it.
Approved by core.
|
| 1.106.10.2 | 10-Mar-2006 |
elad | generic_authorize() -> kauth_authorize_generic().
|
| 1.106.10.1 | 08-Mar-2006 |
elad | Adapt to kernel authorization KPI.
|
| 1.106.8.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.106.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.106.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.106.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.109.6.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.109.6.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.109.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.115.4.2 | 15-Apr-2007 |
yamt | sync with head.
|
| 1.115.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.116.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.116.2.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.116.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.117.8.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.117.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.117.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.117.6.3 | 09-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.117.6.2 | 11-Nov-2007 |
joerg | Sync with HEAD.
|
| 1.117.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.117.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.118.6.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.118.6.2 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.118.6.1 | 19-Nov-2007 |
mjf | Sync with HEAD.
|
| 1.118.4.1 | 13-Nov-2007 |
bouyer | Sync with HEAD
|
| 1.119.2.1 | 08-Dec-2007 |
ad | Sync with head.
|
| 1.123.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.123.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.123.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.124.4.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.124.4.3 | 16-Sep-2009 |
yamt | sync with head
|
| 1.124.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.124.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.124.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.125.2.2 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.125.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.126.4.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.126.4.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.127.4.4 | 22-Aug-2012 |
bouyer | Pull up following revision(s) (requested by gdt in ticket #1776):
sys/netinet/in.c: revision 1.143
Simply use the ifa_addr pointer, rather than taking its address.
Resolves failure to match addresses in SIOC[GD]LIFADDR calls.
Diagnosis and fix is due to Mark Keaton of BBN.
|
| 1.127.4.3 | 20-May-2010 |
snj | Revert ticket 1357.
|
| 1.127.4.2 | 29-Mar-2010 |
snj | Pull up following revision(s) (requested by bouyer in ticket #1357):
sys/netinet/in.c: revision 1.137 via patch
Fixed a number of race conditions in the case of receiving ipv4 packet.
found by iij seil team.
|
| 1.127.4.1 | 09-Jan-2009 |
snj | branches: 1.127.4.1.4;
Pull up following revision(s) (requested by roy in ticket #239):
sys/netinet/in.c: revision 1.130
The automatic addition of a subnet route should not error if a manually
added route already exists. Fixes PR kern/40133.
|
| 1.127.4.1.4.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.127.2.3 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.127.2.2 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.127.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.130.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.136.2.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.136.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.137.2.1 | 30-May-2010 |
rmind | sync with head
|
| 1.140.2.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.140.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.140.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.141.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.142.2.1 | 13-Jun-2012 |
riz | Pull up following revision(s) (requested by gdt in ticket #330):
sys/netinet/in.c: revision 1.143
Simply use the ifa_addr pointer, rather than taking its address.
Resolves failure to match addresses in SIOC[GD]LIFADDR calls.
Diagnosis and fix is due to Mark Keaton of BBN.
|
| 1.143.4.2 | 28-Aug-2013 |
rmind | sync with head
|
| 1.143.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.143.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.143.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.144.4.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.147.2.1 | 14-Apr-2015 |
snj | Pull up following revision(s) (requested by christos in ticket #689):
sys/netinet/in.c: revision 1.149
Don't pass junk in sin_family and sin_len for SIOCGIFNETMASK, and explain why.
XXX: pullup 7?
|
| 1.148.2.12 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.148.2.11 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.148.2.10 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.148.2.9 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.148.2.8 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.148.2.7 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.148.2.6 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.148.2.5 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.148.2.4 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.148.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.148.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.148.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.175.2.5 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.175.2.4 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.175.2.3 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.175.2.2 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.175.2.1 | 26-Jul-2016 |
pgoyette | Sync with HEAD
|
| 1.195.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.199.4.2 | 19-May-2017 |
pgoyette | Resolve conflicts from previous merge (all resulting from $NetBSD
keywork expansion)
|
| 1.199.4.1 | 02-May-2017 |
pgoyette | Sync with HEAD - tag prg-localcount2-base1
|
| 1.203.2.14 | 08-Oct-2020 |
martin | Pull up following revision(s) (requested by roy in ticket #1613):
sys/netinet/in.c: revision 1.241
sys/netinet6/in6.c: revision 1.282
inet: Treat LINK_STATE_UNKNOWN as LINK_STATE_UP when changing
It's something we have always done.
it's really rare for anything to transition to UNKNOWN from either
UP or DOWN, but technically it is possible.
|
| 1.203.2.13 | 09-Apr-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #701):
sys/netinet/in.c: 1.228
Protect ip_dad_count with if NARP > 0 to fix compilation
|
| 1.203.2.12 | 08-Apr-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #701):
sys/netinet/in.c: 1.227
sys/netinet6/in6.c: 1.265
tests/net/arp/t_arp.sh: 1.35-1.36
Make GARP work again when DAD is disabled
The change avoids setting an IP address tentative on initializing it when
the IPv4 DAD is disabled (net.inet.ip.dad_count=0), which allows a GARP packet
to be sent (see arpannounce). This is the same behavior of NetBSD 7, i.e.,
before introducing the IPv4 DAD.
Additionally do the same change to IPv6 DAD for consistency.
The change is suggested by roy@
--
Improve packet checks and error reporting
--
Add tests for GARP without DAD
Additionally make the existing tests for GARP more explicit.
|
| 1.203.2.11 | 13-Mar-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #622):
sys/netinet/if_arp.c: revision 1.270
sys/net/if_llatbl.c: revision 1.24 (patch)
sys/net/if_llatbl.c: revision 1.25
sys/net/if_llatbl.c: revision 1.26
sys/net/route.c: revision 1.204
sys/netinet6/in6.c: revision 1.261
sys/netinet6/in6.c: revision 1.262 (patch)
sys/netinet6/in6.c: revision 1.263
sys/netinet/in.c: revision 1.216
sys/netinet6/in6.c: revision 1.264
sys/netinet6/nd6.c: revision 1.246 (patch)
sys/netinet/if_arp.c: revision 1.269
sys/net/if_llatbl.h: revision 1.14
sys/netinet6/in6.c: revision 1.259
sys/netinet/in.c: revision 1.220
sys/netinet/in.c: revision 1.221 (patch)
sys/netinet/in.c: revision 1.222
sys/netinet/in.c: revision 1.223
Suppress noisy debugging outputs
Even if DEBUG they are too noisy under load.
Tweak sanity checks
Scheduling a timer of static entries is wrong.
Add assertions
We must not destroy llentries holding mbufs.
Fix reference leaks of llentry
callout_reset and callout_halt can cancel a pending callout without telling us.
Detect a cancel and remove a reference by using callout_pending and
callout_stop (it's a bit tricy though, we can detect it).
While here, we can remove remaining abuses of mutex_owned for softnet_lock.
Fix memory leaks on arp -d and ndp -d for static entries
We have to delete entries on in_lltable_delete and in6_lltable_delete
unconditionally. Note that we don't need to worry about LLE_IFADDR because
there is no such entries now.
Use pool(9) for llentry allocations
llentry is easy to be leaked and pool suits for it because pool is usable to
detect leaks.
Also sweep unnecessary wrappers for llentry, in_llentry and in6_llentry.
|
| 1.203.2.10 | 26-Feb-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #588):
sys/netinet6/in6.c: revision 1.260
sys/netinet/in.c: revision 1.219
sys/netinet/wqinput.c: revision 1.4
sys/rump/net/lib/libnetinet/netinet_component.c: revision 1.11
sys/netinet/ip_input.c: revision 1.376
sys/netinet6/ip6_input.c: revision 1.193
Avoid a deadlock between softnet_lock and IFNET_LOCK
A deadlock occurs because there is a violation of the rule of lock ordering;
softnet_lock is held with hodling IFNET_LOCK, which violates the rule.
To avoid the deadlock, replace softnet_lock in in_control and in6_control
with KERNEL_LOCK.
We also need to add some KERNEL_LOCKs to protect the network stack surely.
This is required, for example, for PR kern/51356.
Fix PR kern/53043
|
| 1.203.2.9 | 11-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #539):
sys/netinet/in.c: 1.217
Don't call lltable_purge_entries from in_if_down if ARP isn't enabled
Reported by bouyer@
|
| 1.203.2.8 | 11-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #536):
distrib/sets/lists/base/shl.mi: 1.825
distrib/sets/lists/comp/mi: 1.2168-1.2169
distrib/sets/lists/comp/shl.mi: 1.310
distrib/sets/lists/debug/mi: 1.234
distrib/sets/lists/debug/shl.mi: 1.188
distrib/sets/lists/man/mi: 1.1570
distrib/sets/lists/tests/mi: 1.772
etc/mtree/NetBSD.dist.tests: 1.150
share/man/man4/Makefile: 1.650
share/man/man4/ipsec.4: 1.42-1.43
share/man/man4/ipsecif.4: 1.1-1.5
sys/arch/amd64/conf/ALL: 1.77
sys/arch/amd64/conf/GENERIC: 1.480
sys/conf/files: 1.1191
sys/net/Makefile: 1.34
sys/net/files.net: 1.14
sys/net/if.c: 1.404
sys/net/if.h: 1.248
sys/net/if_gif.c: 1.135
sys/net/if_ipsec.c: 1.1-1.3
sys/net/if_ipsec.h: 1.1
sys/net/if_l2tp.c: 1.16
sys/net/if_types.h: 1.28
sys/netinet/in.c: 1.214
sys/netinet/in.h: 1.103
sys/netinet/in_gif.c: 1.92
sys/netinet/ip_var.h: 1.122
sys/netinet6/in6.c: 1.257
sys/netinet6/in6.h: 1.88
sys/netinet6/in6_gif.c: 1.90
sys/netinet6/ip6_var.h: 1.75
sys/netipsec/Makefile: 1.6
sys/netipsec/files.netipsec: 1.13
sys/netipsec/ipsec.h: 1.62
sys/netipsec/ipsecif.c: 1.1
sys/netipsec/ipsecif.h: 1.1
sys/netipsec/key.c: 1.246-1.247
sys/netipsec/key.h: 1.34
sys/rump/net/Makefile.rumpnetcomp: 1.20
sys/rump/net/lib/libipsec/IPSEC.ioconf: 1.1
sys/rump/net/lib/libipsec/Makefile: 1.1
sys/rump/net/lib/libipsec/ipsec_component.c: 1.1
tests/net/Makefile: 1.34
tests/net/if_ipsec/Makefile: 1.1
tests/net/if_ipsec/t_ipsec.sh: 1.1-1.2
Don't touch an SP without a reference to it
unify processing to check nesting count for some tunnel protocols.
add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
ipsec(4) interface supports rump now.
add ipsec(4) interface ATF.
add ipsec(4) interface man as ipsecif.4.
add ipsec(4) interface to amd64/GENERIC and amd64/ALL configs.
apply in{,6}_tunnel_validate() to gif(4).
Spell IPsec that way. Simplify macro usage. Sort SEE ALSO. Bump
date for previous.
Improve wording and macro use.
Some parts are not clear to me, so someone with knowledge of ipsecif(4)
should improve this some more.
Improve ipsecif.4. Default port ipsec(4) NAT-T is tested now.
pointed out by wiz@n.o and suggested by ozaki-r@n.o, thanks.
Change the prefix of test names to ipsecif_ to distinguish from tests for ipsec(4)
New sentence, new line. Remove empty macro.
Fix PR kern/52920. Pointed out by David Binderman, thanks.
Improve wording, and put a new drawing, from me and Kengo Nakahara.
apply a little more #ifdef INET/INET6. fixes !INET6 builds.
|
| 1.203.2.7 | 03-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #514):
sys/net/route.c: 1.205
sys/net/rtsock.c: 1.237-1.238
sys/netinet/in.c: 1.215
sys/netinet/tcp_subr.c: 1.272
sys/netinet/tcp_timer.c: 1.93
sys/netinet/tcp_timer.h: 1.29
sys/netinet/tcp_var.h: 1.182
sys/netinet6/in6.c: 1.258
Remove extra pserialize_perform from in_purgeaddr
It's already performed in ifa_remove. Note so there (in in6_unlink_ifa too).
Release rt_so_mtx on updating a rtentry to avoid a deadlock with route_intr
The deadlock happened only if NET_MPSAFE on.
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
Fix a return value of rt_update_prepare
Callers expect it to be an errno.
Fix another deadlock
When waiting for a route update to finish, a waiter has to release its reference
to the route to avoid a deadlock. Because a updater tries to wait for references
to a target route (except for a reference by the updater itself) to be released.
|
| 1.203.2.6 | 13-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #488):
sys/netinet/in.c: revision 1.213
Don't pass rwlock to callout_halt
|
| 1.203.2.5 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #463):
sys/netinet/in.c: revision 1.212
sys/netinet/ip_output.c: revision 1.288
sys/netinet6/in6.c: revision 1.256
sys/netinet6/in6_pcb.c: revision 1.163
sys/sys/lwp.h: revision 1.176
Add missing curlwp_bindx
--
Add missing curlwp_bindx
--
Check LP_BOUND is surely set in curlwp_bindx
This may find an extra call of curlwp_bindx.
--
Fix usage of curlwp_bind in ip_output
curlwp_bindx must be called in LIFO order, i.e., we can't call curlwp_bind
and curlwp_bindx like this:
bound1 = curlwp_bind();
bound2 = curlwp_bind();
curlwp_bindx(bound1);
curlwp_bindx(bound2);
ip_outout did so if NET_MPSAFE. Fix it.
--
Fix wrong usage of psref_held
We can't use it for checking if a caller does NOT hold a given target.
If you want to do it you should have psref_not_held or something.
|
| 1.203.2.4 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.203.2.3 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.203.2.2 | 17-Nov-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #353):
sys/net/if_llatbl.c: 1.22
sys/net/if_llatbl.h: 1.13
sys/netinet/if_arp.c: 1.254
sys/netinet/in.c: 1.208-1.209
sys/netinet6/in6.c: 1.249-1.250
sys/netinet6/nd6.c: 1.237
Remove redundant KASSERTMSG
The function is static, has just one caller and the caller does the same check.
--
Fix a deadlock between a route update and lltable
It happens because rtalloc1 is called from lltable with holding
IF_AFDATA_WLOCK.
If a route update is in action, rtalloc1 would wait for its completion with
holding IF_AFDATA_WLOCK. At the same moment, a softint (e.g., arpintr) may try
to take IF_AFDATA_WLOCK and get stuck on it. Unfortunately the stuck softint
prevents the route update from progressing because the route update calls
psref_target_destroy that needs the softint to complete.
A resource allocation graph of the senario looks like this:
route update =(psref_target_destroy)=> softint => IF_AFDATA_WLOCK
=(rt_update_wait)=> route update
Fix the deadlock by pulling rtalloc1 out of the lltable codes inside
IF_AFDATA_WLOCK.
Note that the deadlock happens only if NET_MPSAFE is enabled.
|
| 1.203.2.1 | 07-Jul-2017 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #107):
usr.sbin/arp/arp.c: revision 1.56
sys/net/rtsock.c: revision 1.218
sys/net/if_llatbl.c: revision 1.20
usr.sbin/arp/arp.c: revision 1.57
sys/net/rtsock.c: revision 1.219
sys/net/if_llatbl.c: revision 1.21
usr.sbin/arp/arp.c: revision 1.58
tests/net/net_common.sh: revision 1.19
sys/netinet6/nd6.h: revision 1.84
sys/netinet6/nd6.h: revision 1.85
tests/net/arp/t_arp.sh: revision 1.23
sys/netinet6/in6.c: revision 1.246
tests/net/arp/t_arp.sh: revision 1.24
sys/netinet6/in6.c: revision 1.247
tests/net/arp/t_arp.sh: revision 1.25
sys/netinet6/in6.c: revision 1.248
tests/net/arp/t_arp.sh: revision 1.26
usr.sbin/ndp/ndp.c: revision 1.49
tests/net/arp/t_arp.sh: revision 1.27
tests/net/ndp/t_ndp.sh: revision 1.20
tests/net/arp/t_arp.sh: revision 1.28
tests/net/ndp/t_ndp.sh: revision 1.21
tests/net/arp/t_arp.sh: revision 1.29
tests/net/ndp/t_ndp.sh: revision 1.22
tests/net/ndp/t_ndp.sh: revision 1.23
tests/net/route/t_flags6.sh: revision 1.13
tests/net/ndp/t_ndp.sh: revision 1.24
tests/net/route/t_flags6.sh: revision 1.14
tests/net/ndp/t_ndp.sh: revision 1.25
tests/net/route/t_flags6.sh: revision 1.15
tests/net/ndp/t_ndp.sh: revision 1.26
sbin/route/rtutil.c: revision 1.9
tests/net/ndp/t_ndp.sh: revision 1.27
tests/net/ndp/t_ndp.sh: revision 1.28
tests/net/net/t_ipv6address.sh: revision 1.14
tests/net/ndp/t_ra.sh: revision 1.28
tests/net/ndp/t_ndp.sh: revision 1.29
sys/net/route.h: revision 1.113
tests/net/ndp/t_ra.sh: revision 1.29
sys/net/rtsock.c: revision 1.220
sys/net/rtsock.c: revision 1.221
sys/net/rtsock.c: revision 1.222
sys/net/rtsock.c: revision 1.223
tests/net/route/t_route.sh: revision 1.13
sys/net/rtsock.c: revision 1.224
sys/net/route.c: revision 1.196
sys/net/if_llatbl.c: revision 1.19
sys/net/route.c: revision 1.197
sbin/route/route.c: revision 1.156
tests/net/route/t_flags.sh: revision 1.16
tests/net/route/t_flags.sh: revision 1.17
usr.sbin/ndp/ndp.c: revision 1.50
tests/net/route/t_flags.sh: revision 1.18
sys/netinet/in.c: revision 1.204
tests/net/route/t_flags.sh: revision 1.19
sys/netinet/in.c: revision 1.205
tests/net/arp/t_arp.sh: revision 1.30
tests/net/arp/t_arp.sh: revision 1.31
sys/net/if_llatbl.h: revision 1.11
tests/net/arp/t_arp.sh: revision 1.32
sys/net/if_llatbl.h: revision 1.12
tests/net/arp/t_arp.sh: revision 1.33
sys/netinet6/nd6.c: revision 1.233
sys/netinet6/nd6.c: revision 1.234
sys/netinet/if_arp.c: revision 1.251
sys/netinet6/nd6.c: revision 1.235
sys/netinet/if_arp.c: revision 1.252
sbin/route/route.8: revision 1.57
sys/net/rtsock.c: revision 1.214
sys/net/rtsock.c: revision 1.215
sys/net/rtsock.c: revision 1.216
sys/net/rtsock.c: revision 1.217
whitespace police
Simplify
We can assume that rt_ifp is always non-NULL.
Sending a routing message (RTM_ADD) on adding an llentry
A message used to be sent on adding a cloned route. Restore the
behavior for backward compatibility.
Requested by ryo@
Drop RTF_CONNECTED from a result of RTM_GET for ARP/NDP entries
ARP/NDP entries aren't connected routes.
Reported by ryo@
Support -c <count> option for route monitor
route command exits if it receives <count> routing messages where
<count> is a value specified by -c.
The option is useful to get only particular message(s) in a test script.
Test routing messages emitted on operations of ARP/NDP entries
Do netstat -a for an appropriate protocol
Add missing declarations for cleanup
Set net.inet.arp.keep only if it's required
Don't create a permanent L2 cache entry on adding an address to an interface
It was created to copy FreeBSD, however actually the cache isn't
necessary. Remove it to simplify the code and reduce the cost to
maintain it (e.g., keep a consistency with a corresponding local
route).
Fix typo
Fix in_lltable_match_prefix
The function has not been used but will be used soon.
Remove unused function (nd6_rem_ifa_lle)
Allow in6_lltable_free_entry to be called without holding the afdata lock of ifp as well as in_lltable_free_entry
This behavior is a bit odd and should be fixed in the future...
Purge ARP/NDP entries on an interface when the interface is down
Fix PR kern/51179
Purge all related L2 caches on removing a route
The change addresses situations similar to PR 51179.
Purge L2 caches on changing an interface of a route
The change addresses situations similar to PR 51179.
Test implicit removals of ARP/NDP entries
One test case reproudces PR 51179.
Fix build of kernels without both INET and INET6
Tweak lltable_sysctl_dumparp
- Rename lltable_sysctl_dumparp to lltable_sysctl_dump
because it's not only for ARP
- Enable it not only for INET but also for INET6
Fix usage of routing messages on arp -d and ndp -d
It didn't work as we expected; we should set RTA_GATEWAY not
RTA_IFP on RTM_GET to return an if_index and the kernel should
use it on RTM_DELETE.
Improve backward compatibility of (fake) routing messages on adding an ARP/NDP entry
A message originally included only DST and GATEWAY. Restore it.
Fix ifdef; care about a case w/ INET6 and w/o INET
Drop RTF_UP from a routing message of a deleted ARP/NDP entry
Check existence of ARP/NDP entries
Checking ARP/NDP entries is valid rather than checking routes.
Fix wrong comment
Drop RTF_LLINFO flag (now it's RTF_LLDATA) from local routes
They don't have llinfo anymore. And also the change fixes unexpected
behavior of ARP proxy.
Restore ARP/NDP entries to route show and netstat -r
Requested by dyoung@ some time ago
Enable to remove multiple ARP/NDP entries for one destination
The kernel can have multiple ARP/NDP entries which have an indentical
destination on different interfaces. This is normal and can be
reproduce easily by ping -I or ping6 -S. We should be able to remove
such entries.
arp -d <ip> and ndp -d <ip> are changed to fetch all ARP/NDP entries
and remove matched entries. So we can remove multiple entries
described above. This fetch all and selective removal behavior is
the same as arp <ip> and ndp <ip>; they also do fetch all entries
and show only matched entries.
Related to PR 51179
Check if ARP/NDP entries are purged when a related route is deleted
|
| 1.219.2.7 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.219.2.6 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.219.2.5 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.219.2.4 | 22-Apr-2018 |
pgoyette | Sync with HEAD
|
| 1.219.2.3 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.219.2.2 | 07-Apr-2018 |
pgoyette | Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
|
| 1.219.2.1 | 15-Mar-2018 |
pgoyette | Synch with HEAD
|
| 1.231.2.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.231.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.231.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.234.2.2 | 24-Aug-2024 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1883):
tests/net/arp/t_dad.sh: revision 1.16
sys/netinet/in.c: revision 1.248
tests/net/arp/t_arp.sh: revision 1.46
sys/netinet/if_arp.c: revision 1.314
arp: fix the behavior on detecting an address duplication without IPv4 DAD
On receiving an ARP request that has the same source protocol address as
the own address, i.e., address duplication, the original behavior of
a kernel prior to supporing IPv4 DAD is to send an ARP reply. It is
the same with a latest kernel with DAD enabled. However, a latest
kernel without DAD sends back an GARP packet. Restore the original
behavior.
inet: send GARP on link up if DAD is disabled
This behavior was accidentally removed at rev 1.233.
tests, arp: add tests of address duplications without DAD
tests, arp: add tests for GARP on link up
|
| 1.234.2.1 | 08-Oct-2020 |
martin | Pull up following revision(s) (requested by roy in ticket #1104):
sys/netinet/in.c: revision 1.241
sys/netinet6/in6.c: revision 1.282
inet: Treat LINK_STATE_UNKNOWN as LINK_STATE_UP when changing
It's something we have always done.
it's really rare for anything to transition to UNKNOWN from either
UP or DOWN, but technically it is possible.
|
| 1.247.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.247.2.1 | 24-Aug-2024 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #812):
tests/net/arp/t_dad.sh: revision 1.16
sys/netinet/in.c: revision 1.248
tests/net/arp/t_arp.sh: revision 1.46
sys/netinet/if_arp.c: revision 1.314
arp: fix the behavior on detecting an address duplication without IPv4 DAD
On receiving an ARP request that has the same source protocol address as
the own address, i.e., address duplication, the original behavior of
a kernel prior to supporing IPv4 DAD is to send an ARP reply. It is
the same with a latest kernel with DAD enabled. However, a latest
kernel without DAD sends back an GARP packet. Restore the original
behavior.
inet: send GARP on link up if DAD is disabled
This behavior was accidentally removed at rev 1.233.
tests, arp: add tests of address duplications without DAD
tests, arp: add tests for GARP on link up
|
| 1.115 | 16-Jun-2023 |
rin | White space fixes. No binary changes.
|
| 1.114 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.113 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.112 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.111 | 08-Sep-2020 |
christos | branches: 1.111.2;
Add IP_BINDANY, IPV6_BINDANY which can be used to bind to any address in
order to implement transparent proxies.
|
| 1.110 | 20-Aug-2020 |
riastradh | [ozaki-r] Changes to the kernel core for wireguard
|
| 1.109 | 18-Dec-2019 |
roy | inet: Add support for IPv4 /31 prefixes, as described in RFC 3021.
To run a /31 network, participating hosts MUST drop support
for directed broadcasts, and treat the first and last addresses
on subnet as unicast. The broadcast address for the prefix
should be the link local broadcast address, INADDR_BROADCAST.
Taken from FreeBSD, r226402.
Fixes PR kern/51388.
|
| 1.108 | 09-Nov-2018 |
maya | Use the same type redefinition guards as stdint.h since rev1.8
PR pkg/53713
|
| 1.107 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.106 | 11-Jul-2018 |
maxv | Rename
ip_undefer_csum -> in_undefer_cksum
in_delayed_cksum -> in_undefer_cksum_tcpudp
The two previous names were inconsistent and misleading.
Put the two functions into in_offload.c. Add comments to explain what
we're doing.
The same could be done for IPv6.
|
| 1.105 | 19-Apr-2018 |
christos | branches: 1.105.2;
s/static inline/static __inline/g for consistency.
|
| 1.104 | 09-Feb-2018 |
maxv | branches: 1.104.2;
Remove dead code.
|
| 1.103 | 10-Jan-2018 |
knakahara | add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
|
| 1.102 | 01-Jan-2018 |
christos | 1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
|
| 1.101 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.100 | 16-Feb-2017 |
knakahara | branches: 1.100.6;
add l2tp(4) L2TPv3 interface.
originally implemented by IIJ SEIL team.
|
| 1.99 | 01-Aug-2016 |
ozaki-r | branches: 1.99.2;
Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.98 | 13-Oct-2015 |
rjs | branches: 1.98.2;
Add core networking support for SCTP.
|
| 1.97 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.96 | 10-Feb-2015 |
rjs | Add DCCP protocol support from KAME.
|
| 1.95 | 02-Dec-2014 |
christos | use the new printing code.
|
| 1.94 | 02-Dec-2014 |
christos | add routines to print in_addr and sockaddr_in (in_print and sin_print)
|
| 1.93 | 12-Oct-2014 |
christos | branches: 1.93.2;
document that we depend on the option numbers matching.
|
| 1.92 | 05-Jun-2014 |
rmind | - Implement pktqueue interface for lockless IP input queue.
- Replace ipintrq and ip6intrq with the pktqueue mechanism.
- Eliminate kernel-lock from ipintr() and ip6intr().
- Some preparation work to push softnet_lock out of ipintr().
Discussed on tech-net.
|
| 1.91 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.90 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.89 | 27-Jun-2013 |
christos | branches: 1.89.2; 1.89.6;
implement IP_PKTINFO and IP_RECVPKTINFO.
|
| 1.88 | 27-Apr-2013 |
joerg | Systematically include sys/featuretest.h when _NETBSD_SOURCE is used.
Some are redundant, but make verification with grep much easier.
|
| 1.87 | 22-Jun-2012 |
christos | branches: 1.87.2;
PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.86 | 14-Sep-2009 |
degroote | branches: 1.86.12;
Import pfsync support from OpenBSD 4.2
Pfsync interface exposes change in the pf(4) over a pseudo-interface, and can
be used to synchronise different pf.
This work was part of my 2009 GSoC
No objection on tech-net@
|
| 1.85 | 17-Jul-2009 |
minskim | Add the IP_MINTTL socket option.
The IP_MINTTL option may be used on SOCK_STREAM sockets to discard
packets with a TTL lower than the option value. This can be used to
implement the Generalized TTL Security Mechanism (GTSM) according to
RFC 3682.
OK'ed by christos@.
|
| 1.84 | 16-Jul-2009 |
minskim | Add the IP_RECVTTL option support.
If the IP_RECVTTL option is enabled on a SOCK_DGRAM socket, the
recvmsg(2) call will return the TTL of the received datagram. The
msg_control field in the msghdr structure points to a buffer that
contains a cmsghdr structure followed by the TTL value.
Modeled after FreeBSD implementation.
|
| 1.83 | 25-Jan-2008 |
joerg | branches: 1.83.2; 1.83.10; 1.83.24;
Refactor in_cksum/in4_cksum/in6_cksum implementations:
- All three functions are included in the kernel by default.
They call a backend function cpu_in_cksum after possibly
computing the checksum of the pseudo header.
- cpu_in_cksum is the core to implement the one-complement sum.
The default implementation is moderate fast on most platforms
and provides a 32bit accumulator with 16bit addends for L32 platforms
and a 64bit accumulator with 32bit addends for L64 platforms.
It handles edge cases like very large mbuf chains (could happen with
native IPv6 in the future) and provides a good base for new native
implementations.
- Modify i386 and amd64 assembly to use the new interface.
This disables the MD implementations on !x86 until the conversion is
done. For Alpha, the portable version is faster.
|
| 1.82 | 25-Dec-2007 |
perry | Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.81 | 19-Sep-2007 |
dyoung | branches: 1.81.6; 1.81.8; 1.81.12;
1) Introduce a new socket option, (SOL_SOCKET, SO_NOHEADER), that
tells a socket that it should both add a protocol header to tx'd
datagrams and remove the header from rx'd datagrams:
int onoff = 1, s = socket(...);
setsockopt(s, SOL_SOCKET, SO_NOHEADER, &onoff);
2) Add an implementation of (SOL_SOCKET, SO_NOHEADER) for raw IPv4
sockets.
3) Reorganize the protocols' pr_ctloutput implementations a bit.
Consistently return ENOPROTOOPT when an option is unsupported,
and EINVAL if a supported option's arguments are incorrect.
Reorganize the flow of code so that it's more clear how/when
options are passed down the stack until they are handled.
Shorten some pr_ctloutput staircases for readability.
4) Extract common mbuf code into subroutines, add new sockaddr
methods, and introduce a new subroutine, fsocreate(), for reuse
later; use it first in sys_socket():
struct mbuf *m_getsombuf(struct socket *so)
Create an mbuf and make its owner the socket `so'.
struct mbuf *m_intopt(struct socket *so, int val)
Create an mbuf, make its owner the socket `so', put the
int `val' into it, and set its length to sizeof(int).
int fsocreate(..., int *fd)
Create a socket, a la socreate(9), put the socket into the
given LWP's descriptor table, return the descriptor at `fd'
on success.
void *sockaddr_addr(struct sockaddr *sa, socklen_t *slenp)
const void *sockaddr_const_addr(const struct sockaddr *sa, socklen_t *slenp)
Extract a pointer to the address part of a sockaddr. Write
the length of the address part at `slenp', if `slenp' is
not NULL.
socklen_t sockaddr_getlen(const struct sockaddr *sa)
Return the length of a sockaddr. This just evaluates to
sa->sa_len. I only add this for consistency with code that
appears in a portable userland library that I am going to
import.
const struct sockaddr *sockaddr_any(const struct sockaddr *sa)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
const void *sockaddr_anyaddr(const struct sockaddr *sa, socklen_t *slenp)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
|
| 1.80 | 30-Aug-2007 |
dyoung | Use malloc(9) for sockaddrs instead of pool(9), and remove dom_sa_pool
and dom_sa_len members from struct domain. Pools of fixed-size
objects are too rigid for sockaddr_dls, whose size can vary over
a wide range.
Return sockaddr_dl to its "historical" size. Now that I'm using
malloc(9) instead of pool(9) to allocate sockaddr_dl, I can create
a sockaddr_dl of any size in the kernel, so expanding sockaddr_dl
is useless.
Avoid using sizeof(struct sockaddr_dl) in the kernel.
Introduce sockaddr_dl_alloc() for allocating & initializing an
arbitrary sockaddr_dl on the heap.
Add an argument, the sockaddr length, to sockaddr_alloc(),
sockaddr_copy(), and sockaddr_dl_setaddr().
Constify: LLADDR() -> CLLADDR().
Where the kernel overwrites LLADDR(), use sockaddr_dl_setaddr(),
instead. Used properly, sockaddr_dl_setaddr() will not overrun
the end of the sockaddr.
|
| 1.79 | 02-May-2007 |
dyoung | branches: 1.79.2; 1.79.6; 1.79.8;
Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.78 | 17-Feb-2007 |
dyoung | branches: 1.78.4; 1.78.6;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.77 | 13-Nov-2006 |
dyoung | branches: 1.77.4;
Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.76 | 18-May-2006 |
liamjfoy | branches: 1.76.8; 1.76.10;
Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.75 | 16-Feb-2006 |
perry | branches: 1.75.2; 1.75.6;
Change "inline" back to "__inline" in .h files -- C99 is still too
new, and some apps compile things in C89 mode. C89 keywords stay.
As per core@.
|
| 1.74 | 24-Dec-2005 |
perry | branches: 1.74.2; 1.74.4; 1.74.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.73 | 20-Dec-2005 |
christos | Define INADDR_NONE when we are in the kernel too.
|
| 1.72 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.71 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.70 | 31-Jan-2005 |
kim | branches: 1.70.6;
Add RFC 3378 EtherIP support, ported from OpenBSD to NetBSD by
Hans Rosenfeld (rosenfeld at grumpf.hope-2000.org)
This change makes it possible to add gif interfaces to bridges, which
will then send and receive IP protocol 97 packets. Packets are Ethernet
frames with an EtherIP header prepended.
|
| 1.69 | 15-Dec-2004 |
thorpej | branches: 1.69.2; 1.69.4;
Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.68 | 04-Sep-2004 |
manu | IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.67 | 07-May-2004 |
jonathan | Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.66 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.65 | 18-Apr-2004 |
matt | De __P()
|
| 1.64 | 19-Nov-2003 |
jonathan | branches: 1.64.2;
Patch back support for (badly) randomized IP ids, by request:
* Include "opt_inet.h" everywhere IP-ids are generated with ip_newid(),
so the RANDOM_IP_ID option is visible. Also in ip_id(), to ensure
the prototype for ip_randomid() is made visible.
* Add new sysctl to enable randomized IP-ids, provided the kernel was
configured with RANDOM_IP_ID. (The sysctl defaults to zero, and is
a read-only zero if RANDOM_IP_ID is not configured).
Note that the implementation of randomized IP ids is still defective,
and should not be enabled at all (even if configured) without
very careful deliberation. Caveat emptor.
|
| 1.63 | 10-Nov-2003 |
jonathan | Allocate sysctl oid for ipv4 sysctl node "ifq", define symbolic name, and
bump IPCTL_MAXID. (Should have been committed with other ifq sysctl changes).
|
| 1.62 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.61 | 28-Apr-2003 |
bjh21 | branches: 1.61.2;
Add a new feature-test macro, _NETBSD_SOURCE. If this is defined
by the application, all NetBSD interfaces are made visible, even
if some other feature-test macro (like _POSIX_C_SOURCE) is defined.
<sys/featuretest.h> defined _NETBSD_SOURCE if none of _ANSI_SOURCE,
_POSIX_C_SOURCE and _XOPEN_SOURCE is defined, so as to preserve
existing behaviour.
This has two major advantages:
+ Programs that require non-POSIX facilities but define _POSIX_C_SOURCE
can trivially be overruled by putting -D_NETBSD_SOURCE in their CFLAGS.
+ It makes most of the #ifs simpler, in that they're all now ORs of the
various macros, rather than having checks for (!defined(_ANSI_SOURCE) ||
!defined(_POSIX_C_SOURCE) || !defined(_XOPEN_SOURCE)) all over the place.
I've tried not to change the semantics of the headers in any case where
_NETBSD_SOURCE wasn't defined, but there were some places where the
current semantics were clearly mad, and retaining them was harder than
correcting them. In particular, I've mostly normalised things so that
_ANSI_SOURCE gets you the smallest set of stuff, then _POSIX_C_SOURCE,
_XOPEN_SOURCE and _NETBSD_SOURCE in that order.
Tested by building for vax, encouraged by thorpej, and uncontested in
tech-userlevel for a week.
|
| 1.60 | 12-Apr-2003 |
dogcow | PR/991: Darren Reed: Add a sysctl (checkinteface) to implement this. This
implementation is taken from FreeBSD, but we default to off.
XXX: We should really do this on a per ifaddr basis as jason suggested.
|
| 1.59 | 27-Jan-2003 |
kleink | C++ does not permit static a data member to have the same name as its
class, so in a C++ environment rename the ip_opts member to Ip_opts as
observed in several other implementations; from Jon Olsson in
PR toolchain/19880.
|
| 1.58 | 13-May-2002 |
kleink | branches: 1.58.4;
Define uint{8,32}_t locally, per XNS5.2/POSIX-2001, and use them in this
header where applicable; use private fixed-width integer types otherwise.
|
| 1.57 | 12-May-2002 |
kleink | Provide local definitions of in_{addr,port}_t in <netinet/in.h> and use
them where deemed appropriate by XNS5.2/POSIX-2001.
|
| 1.56 | 24-Feb-2002 |
martin | Clear M_BCAST and M_MCAST on outgoing mbufs.
Don't copy ttl from the inner packet to the encapsulating packet. Make
the outer ttl sysctl'able. This should close PR 14269 from Jasper Wallace
(change partly from there) and it makes traceroute work over gre tunnels.
|
| 1.55 | 02-Jun-2001 |
thorpej | branches: 1.55.2;
Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.54 | 27-May-2001 |
itojun | typo in comment
|
| 1.53 | 27-Mar-2001 |
itojun | net.inet.ip.maxfragpackets defines the maximum size of ip reass queue
(prevents fragment flood from chewing up mbuf memory space).
derived from KAME net.inet6.ip6.maxfragpackets.
|
| 1.52 | 19-Jan-2001 |
kleink | branches: 1.52.2;
Add IPPROTO_VRRP.
|
| 1.51 | 28-Aug-2000 |
simonb | #define<tab> cleanup.
|
| 1.50 | 25-Aug-2000 |
tron | Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
|
| 1.49 | 28-Jul-2000 |
kleink | Avoid recursion with traditional cpp.
|
| 1.48 | 26-Jun-2000 |
kleink | XNS5.2: define sa_family_t and use it where specified by the standard.
|
| 1.47 | 10-Mar-2000 |
itojun | branches: 1.47.4;
move IPPROTO_DONE to IPPROTO_xx group
|
| 1.46 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.45 | 09-Feb-2000 |
itojun | to improve RFC2553/2292 compliance, and promote use of
RFC2553/2292-compliant header file path, now the following headers are
forbidden:
netinet6/ip6.h
netinet6/icmp6.h
netinet6/in6.h
if you want netinet6/{ip6,icmp6}.h, use netinet/{ip6,icmp6}.h.
if you want netinet6/in6.h, you just need to include netinet/in.h.
it pulls it in.
(we may need to integrate them into netinet/in.h, but for cross-BSD code
sharing i'd like to keep it like this for now)
|
| 1.44 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.43 | 20-Nov-1999 |
thorpej | Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.42 | 02-Jul-1999 |
itojun | branches: 1.42.2; 1.42.8;
move ipsec sysctl index to IPPROTO_AH (instead of IPPROTO_ESP),
so that you can perform sysctl operation when ESP is not compiled in.
|
| 1.41 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.40 | 26-Jun-1999 |
sommerfeld | If the new global variable hostzerobroadcast is zero, no longer assume
address zero of each net/subnet is a broadcast address.
(The default value is nonzero, which preserves the current behavior).
This can be set using sysctl; the boot-time default can also be
configured using the HOSTZEROBROADCAST kernel config option.
While we're here, defopt HOSTZEROBROADCAST and SUBNETSARELOCAL
|
| 1.39 | 14-Sep-1998 |
hwr | branches: 1.39.6; 1.39.8; 1.39.10;
Typo. :(
|
| 1.38 | 14-Sep-1998 |
hwr | Some additions.
And IDPR-CMTP is 38 not 39 according to IANA.
|
| 1.37 | 13-Sep-1998 |
hwr | Add a gre tunnel pseudo network device. Gre = generic route encapsulation.
This device shows up like any other network interface and can be used to
tunnel L3 protocols as e.g. IP over IP.
|
| 1.36 | 05-Sep-1998 |
kleink | Protect _XOPEN_SOURCE against sysctl MIB identifiers.
|
| 1.35 | 04-May-1998 |
matt | Default IP flow to being enabled. Add a sysctl to control the maximum
number of flows (net.inet.ip.maxflows). If set to 0, will disable fast
path forwarding.
|
| 1.34 | 29-Apr-1998 |
kml | Add support for deletion of routes added by path MTU discovery;
uses new generic route timeout code. Add sysctl for timeout period.
|
| 1.33 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.32 | 07-Jan-1998 |
lukem | add the following, derived from FreeBSD:
* IP_PORTRANGE socket option, which controls how the ephemeral ports
are allocated. it takes the following settings:
IP_PORTRANGE_DEFAULT use anonportmin (49152) -> anonportmax (65535)
IP_PORTRANGE_HIGH as IP_PORTRANGE_DEFAULT (retained for FreeBSD
compat reasons, where these are separate)
IP_PORTRANGE_LOW use 600 -> 1023. only works if uid==0.
* in_pcb flag INP_ANONPORT. set if port was allocated ephmerally
|
| 1.31 | 05-Jan-1998 |
lukem | enhance ephemeral port allocation code:
* support sysctl net.inet.ip.anonportmin (lowest ephemeral port)
and net.inet.ip.anonportmax (highest ephemeral port).
these can't be set to >65535, < IPPORT_RESERVED (unless IPNOPRIVPORTS
is defined), and anonportmin has to be < anonportmax.
* use a cleaner way of only cycling through the available set once;
this will be useful for when a random allocation scheme is used
* define IPPORT_ANON{MIN,MAX} instead of IPPORT_USER{LOW,HIGH}
|
| 1.30 | 30-Dec-1997 |
lukem | as per the IANA assigned ports numbers document, use ports
49152..65535 for ephemeral ports (instead of 1024..5000).
closes my [kern/4440], but with correct code :)
|
| 1.29 | 16-Dec-1997 |
thorpej | Add INADDR_ALLRTRS_GROUP and INADDR_MAX_LOCAL_GROUP.
|
| 1.28 | 18-Oct-1997 |
kml | branches: 1.28.2;
change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.27 | 17-Oct-1997 |
thorpej | Allow `subnetsarelocal' to be changed via sysctl.
|
| 1.26 | 14-Oct-1997 |
matt | Add support for returning maximum supported MTU when ip_output fails with
EMSGSIZE.
|
| 1.25 | 27-Aug-1997 |
matt | Add IPPROTO_ESP and IPPROTO_AH defines.
|
| 1.24 | 25-Feb-1997 |
cjs | branches: 1.24.4;
Add net.inet.ip.allowsrcrt option which allows/drops all source
routed packets. This currently defaults to `drop,' but once we
verify that all applications that rely on determining remote IP
addresses for authentication are dropping the connection when they
see a source route option (not just disabling the source route
option), we can turn this back on and conform with the host
requirements.
|
| 1.23 | 11-Jan-1997 |
thorpej | branches: 1.23.4;
Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.22 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.21 | 22-May-1996 |
mycroft | A few style changes to match netiso and netns.
|
| 1.20 | 13-Feb-1996 |
christos | branches: 1.20.4;
netinet prototypes
|
| 1.19 | 16-Jan-1996 |
thorpej | Add a net.inet.ip.directed-broadcast sysctl as suggested by
Darren Reed <darrenr@vitruvius.arbld.unimelb.edu.au> in PR #1227.
This change is slightly different than the one submitted by Darren in
that the DIRECTED_BROADCAST compile-time option will behave like it used
to so that existing configurations utilizing it won't have to change.
|
| 1.18 | 15-Jan-1996 |
thorpej | Add net.inet.ip.forwsrcrt: if zero, the system will not forward
source-routed packets. Note this value is protected by kernel security
level; it can only be changed if securelevel < 1.
|
| 1.17 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.16 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.15 | 31-May-1995 |
mycroft | Add IPPROTO_IP. Fix comment for IP_MULTICAST_IF.
|
| 1.14 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.13 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.12 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.11 | 14-May-1994 |
cgd | i forgot a four letter word...
|
| 1.10 | 14-May-1994 |
cgd | multiple inclusion protection, for the rpc headers
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 28-Jan-1994 |
deraadt | need a stub 'struct socket;' for a prototype
|
| 1.7 | 26-Jan-1994 |
cgd | renumber the IP setsockopt options back to the Reno/Net2 versions,
moving the multicast options after them
From: Mike Karels <karels@BSDI.COM>
(grr.)
|
| 1.6 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 06-Dec-1993 |
hpeyerl | multicast patches
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.20.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.23.4.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.24.4.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.24.4.1 | 28-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.28.2.2 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.28.2.1 | 29-Jan-1998 |
mellon | Pull up 1.29 (thorpej)
|
| 1.39.10.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.39.10.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.39.10.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.39.8.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.39.8.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.39.6.1 | 30-May-2001 |
he | Pull up revision 1.53 (via patch, requested by he):
Introduce net.inet.ip.maxfragpackets, which controls the maximum
number of IPv4 fragment reassembly queue entries. Defends against
certain DoS attacks. Fixes SA#2001-006.
|
| 1.42.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.42.2.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.42.2.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.42.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.47.4.3 | 26-Feb-2002 |
he | Pull up revision 1.56 (via patch, requested by martin):
Clear M_BCAST and M_MCAST on encapsulated packets on outgoing
mbufs. Also do not copy TTL from the inner packet, and make the
outer TTL sysctl'able. Fixes PR#14269, and makes traceroute work
over GRE tunnels.
|
| 1.47.4.2 | 24-Apr-2001 |
he | Pull up revision 1.53 (via patch, requested by itojun):
Introduce net.inet.ip.maxfragpackets, which controls the maximum
number of IPv4 fragment reassembly queue entries. Defends against
certain DoS attacks.
|
| 1.47.4.1 | 26-Aug-2000 |
tron | Pull up from current (approved by thorpej):
Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
syssrc/sys/netinet/in.h 1.49 -> 1.50
syssrc/sys/netinet/in_pcb.c 1.66 -> 1.67
syssrc/sys/netinet/ip_input.c 1.116 -> 1.117
syssrc/sys/netinet/ip_var.h 1.41 -> 1.42
|
| 1.52.2.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.52.2.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.52.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.52.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.55.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.55.2.1 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.58.4.1 | 28-Apr-2003 |
tron | Pull up revision 1.59 (requested by kleink in ticket #1119):
C++ does not permit static a data member to have the same name as its
class, so in a C++ environment rename the ip_opts member to Ip_opts as
observed in several other implementations; from Jon Olsson in
PR toolchain/19880.
|
| 1.61.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.61.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.61.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.61.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.61.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.61.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.64.2.1 | 10-May-2004 |
tron | Pull up revision 1.67 (requested by jonathan in ticket #280):
Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.69.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.69.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.70.6.7 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.70.6.6 | 21-Jan-2008 |
yamt | sync with head
|
| 1.70.6.5 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.70.6.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.70.6.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.70.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.70.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.74.6.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.74.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.74.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.74.2.1 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.75.6.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.75.2.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.76.10.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.76.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.77.4.2 | 07-May-2007 |
yamt | sync with head.
|
| 1.77.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.78.6.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.78.4.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.78.4.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.79.8.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.79.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.79.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.79.6.2 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.79.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.79.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.81.12.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.81.8.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.81.6.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.83.24.1 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.83.10.3 | 16-Sep-2009 |
yamt | sync with head
|
| 1.83.10.2 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.83.10.1 | 18-Jul-2009 |
yamt | sync with head.
|
| 1.83.2.1 | 22-Feb-2008 |
keiichi | imported Mobile IPv6 code developed by the SHISA project
(http://www.mobileip.jp/).
|
| 1.86.12.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.86.12.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.87.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.87.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.87.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.89.6.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.89.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.93.2.5 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.93.2.4 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.93.2.3 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.93.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.93.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.98.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.98.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.99.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.100.6.3 | 18-Mar-2018 |
martin | Pull up following revision(s) (requested by tih in ticket #639):
sys/kern/uipc_socket.c: revision 1.258
sys/kern/uipc_socket.c: revision 1.259
sys/netinet/ip_input.c: revision 1.364 (via patch)
sys/netinet/ip_output.c: revision 1.289
sys/netinet/in.h: revision 1.102
sys/netinet/in_pcb.c: revision 1.181
share/man/man9/sockopt.9: revision 1.11
sys/netinet/in_pcb.h: revision 1.65
sys/sys/socketvar.h: revision 1.146
sys/kern/uipc_syscalls.c: revision 1.189
sys/netinet/ip_output.c: revision 1.290
share/man/man4/ip.4: revision 1.41
share/man/man4/ip.4: revision 1.42
sys/kern/uipc_syscalls.c: revision 1.190
pass valsize for getsockopt like we do for setsockopt
make sure that we have enough space, don't require the exact size
(Tom Ivar Helbekkmo)
1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
new sentence-new line
Remove comment now that the getsockopt code passes the size.
Add a new sockopt member to keep track of the actual size of the option
that should be returned to the caller in getsockopt(2).
(Tom Ivar Helbekkmo)
|
| 1.100.6.2 | 11-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #536):
distrib/sets/lists/base/shl.mi: 1.825
distrib/sets/lists/comp/mi: 1.2168-1.2169
distrib/sets/lists/comp/shl.mi: 1.310
distrib/sets/lists/debug/mi: 1.234
distrib/sets/lists/debug/shl.mi: 1.188
distrib/sets/lists/man/mi: 1.1570
distrib/sets/lists/tests/mi: 1.772
etc/mtree/NetBSD.dist.tests: 1.150
share/man/man4/Makefile: 1.650
share/man/man4/ipsec.4: 1.42-1.43
share/man/man4/ipsecif.4: 1.1-1.5
sys/arch/amd64/conf/ALL: 1.77
sys/arch/amd64/conf/GENERIC: 1.480
sys/conf/files: 1.1191
sys/net/Makefile: 1.34
sys/net/files.net: 1.14
sys/net/if.c: 1.404
sys/net/if.h: 1.248
sys/net/if_gif.c: 1.135
sys/net/if_ipsec.c: 1.1-1.3
sys/net/if_ipsec.h: 1.1
sys/net/if_l2tp.c: 1.16
sys/net/if_types.h: 1.28
sys/netinet/in.c: 1.214
sys/netinet/in.h: 1.103
sys/netinet/in_gif.c: 1.92
sys/netinet/ip_var.h: 1.122
sys/netinet6/in6.c: 1.257
sys/netinet6/in6.h: 1.88
sys/netinet6/in6_gif.c: 1.90
sys/netinet6/ip6_var.h: 1.75
sys/netipsec/Makefile: 1.6
sys/netipsec/files.netipsec: 1.13
sys/netipsec/ipsec.h: 1.62
sys/netipsec/ipsecif.c: 1.1
sys/netipsec/ipsecif.h: 1.1
sys/netipsec/key.c: 1.246-1.247
sys/netipsec/key.h: 1.34
sys/rump/net/Makefile.rumpnetcomp: 1.20
sys/rump/net/lib/libipsec/IPSEC.ioconf: 1.1
sys/rump/net/lib/libipsec/Makefile: 1.1
sys/rump/net/lib/libipsec/ipsec_component.c: 1.1
tests/net/Makefile: 1.34
tests/net/if_ipsec/Makefile: 1.1
tests/net/if_ipsec/t_ipsec.sh: 1.1-1.2
Don't touch an SP without a reference to it
unify processing to check nesting count for some tunnel protocols.
add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
ipsec(4) interface supports rump now.
add ipsec(4) interface ATF.
add ipsec(4) interface man as ipsecif.4.
add ipsec(4) interface to amd64/GENERIC and amd64/ALL configs.
apply in{,6}_tunnel_validate() to gif(4).
Spell IPsec that way. Simplify macro usage. Sort SEE ALSO. Bump
date for previous.
Improve wording and macro use.
Some parts are not clear to me, so someone with knowledge of ipsecif(4)
should improve this some more.
Improve ipsecif.4. Default port ipsec(4) NAT-T is tested now.
pointed out by wiz@n.o and suggested by ozaki-r@n.o, thanks.
Change the prefix of test names to ipsecif_ to distinguish from tests for ipsec(4)
New sentence, new line. Remove empty macro.
Fix PR kern/52920. Pointed out by David Binderman, thanks.
Improve wording, and put a new drawing, from me and Kengo Nakahara.
apply a little more #ifdef INET/INET6. fixes !INET6 builds.
|
| 1.100.6.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.104.2.4 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.104.2.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.104.2.2 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.104.2.1 | 22-Apr-2018 |
pgoyette | Sync with HEAD
|
| 1.105.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.105.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.111.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.20 | 30-Nov-2014 |
christos | Only check that the offset < sizeof(struct ip) if nxt != 0, i.e. in the
tcp and udp cases. From kre.
XXX: pullup 7
|
| 1.19 | 12-Mar-2013 |
christos | branches: 1.19.12; 1.19.14;
- Do the size checks before calling the cpu checksum code. Otherwise you'll
crash there and not panic.
- Don't panic on short packets unless DIAGNOSTIC. In general we should try
to make the kernel survive errors...
|
| 1.18 | 25-Apr-2011 |
yamt | branches: 1.18.4; 1.18.14;
fix assertions
|
| 1.17 | 12-Feb-2008 |
joerg | branches: 1.17.32; 1.17.38;
Explicitly predict panic conditions as false.
|
| 1.16 | 07-Feb-2008 |
joerg | Reimplement in4_cksum to not copy data, but sum up directly.
Tested on sparc and m68k by martin@.
|
| 1.15 | 25-Jan-2008 |
joerg | Refactor in_cksum/in4_cksum/in6_cksum implementations:
- All three functions are included in the kernel by default.
They call a backend function cpu_in_cksum after possibly
computing the checksum of the pseudo header.
- cpu_in_cksum is the core to implement the one-complement sum.
The default implementation is moderate fast on most platforms
and provides a 32bit accumulator with 16bit addends for L32 platforms
and a 64bit accumulator with 32bit addends for L64 platforms.
It handles edge cases like very large mbuf chains (could happen with
native IPv6 in the future) and provides a good base for new native
implementations.
- Modify i386 and amd64 assembly to use the new interface.
This disables the MD implementations on !x86 until the conversion is
done. For Alpha, the portable version is faster.
|
| 1.14 | 04-Mar-2007 |
tsutsui | branches: 1.14.16; 1.14.22;
Pass (char *) to mtod(9) on address calculation.
|
| 1.13 | 04-Mar-2007 |
christos | Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.12 | 11-Dec-2005 |
christos | branches: 1.12.26;
merge ktrace-lwp.
|
| 1.11 | 03-Feb-2005 |
perry | branches: 1.11.6;
ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.10 | 07-Aug-2003 |
agc | branches: 1.10.8; 1.10.10;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.9 | 29-Jul-2002 |
itojun | branches: 1.9.6;
be friendly with gcc-3.1.1 -O2, which takes advantage of ANSI C
pointer aliasing rule (gcc optimization/7427). from tsubai, sync w/kame
|
| 1.8 | 21-Dec-2001 |
itojun | branches: 1.8.8;
whitespace. sync with kame
|
| 1.7 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.6 | 19-May-2001 |
thorpej | branches: 1.6.2;
Brain'o in last. Pointed out by Steve Woodford <scw@netbsd.org>.
|
| 1.5 | 19-May-2001 |
thorpej | Don't compute psuedo header checksum if nxt == 0.
|
| 1.4 | 30-Mar-2000 |
augustss | branches: 1.4.6; 1.4.8;
Remove register declarations.
|
| 1.3 | 15-Feb-2000 |
itojun | make assumption on mbuf explicit (m->m_len >= sizeof (struct ip)).
|
| 1.2 | 13-Dec-1999 |
itojun | branches: 1.2.2;
sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.1 | 30-Nov-1999 |
itojun | branches: 1.1.2;
file in4_cksum.c was initially added on branch kame.
|
| 1.1.2.1 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.2.2.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.4.8.4 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.4.8.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.4.8.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.4.8.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.4.6.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.4.6.1 | 30-Mar-2000 |
bouyer | file in4_cksum.c was added on branch thorpej_scsipi on 2000-11-20 18:10:22 +0000
|
| 1.6.2.2 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.6.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.8.8.1 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.9.6.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.9.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.9.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.9.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.10.10.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.10.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.11.6.4 | 27-Feb-2008 |
yamt | sync with head.
|
| 1.11.6.3 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.11.6.2 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.11.6.1 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.12.26.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.14.22.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.14.16.1 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.17.38.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.17.32.1 | 31-May-2011 |
rmind | sync with head
|
| 1.18.14.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.18.14.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.18.4.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.19.14.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.19.12.1 | 01-Dec-2014 |
martin | Pull up following revision(s) (requested by christos in ticket #281):
sys/netinet/in4_cksum.c: revision 1.20
Only check that the offset < sizeof(struct ip) if nxt != 0, i.e. in the
tcp and udp cases. From kre.
|
| 1.22 | 25-Jan-2008 |
joerg | Refactor in_cksum/in4_cksum/in6_cksum implementations:
- All three functions are included in the kernel by default.
They call a backend function cpu_in_cksum after possibly
computing the checksum of the pseudo header.
- cpu_in_cksum is the core to implement the one-complement sum.
The default implementation is moderate fast on most platforms
and provides a 32bit accumulator with 16bit addends for L32 platforms
and a 64bit accumulator with 32bit addends for L64 platforms.
It handles edge cases like very large mbuf chains (could happen with
native IPv6 in the future) and provides a good base for new native
implementations.
- Modify i386 and amd64 assembly to use the new interface.
This disables the MD implementations on !x86 until the conversion is
done. For Alpha, the portable version is faster.
|
| 1.21 | 09-Jan-2008 |
joerg | When not compiling for the kernel, use stdio.h instead of sys/systm.h
(printf) and locally define the protoype. Makes it possible to use
in_cksum.c for regression testing.
|
| 1.20 | 09-Jan-2008 |
joerg | Anyone seriously interested in implementing in_cksum on a new platform
should read RFC 1071, so point them to it.
|
| 1.19 | 11-Dec-2005 |
christos | branches: 1.19.46; 1.19.52; 1.19.60;
merge ktrace-lwp.
|
| 1.18 | 03-Feb-2005 |
perry | branches: 1.18.6;
ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.17 | 07-Aug-2003 |
agc | branches: 1.17.8; 1.17.10;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.16 | 09-Jun-2002 |
itojun | branches: 1.16.6;
whitespace
|
| 1.15 | 13-Nov-2001 |
lukem | branches: 1.15.8;
add RCSIDs
|
| 1.14 | 30-Mar-2000 |
augustss | branches: 1.14.6; 1.14.8;
Remove register declarations.
|
| 1.13 | 13-Oct-1996 |
christos | branches: 1.13.28;
backout previous kprintf changes
|
| 1.12 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.11 | 08-Apr-1996 |
jonathan | fixes for -Wall -Wmissing-prototypes:
include <sys/systm.h> to get a prototyped declaration of printf().
include <netinet/in.h> to get a prototyped declaration of in_cksum().
|
| 1.10 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.9 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.8 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.7 | 10-Jan-1994 |
mycroft | Fix function name.
|
| 1.6 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 17-Apr-1993 |
glass | this file is never compiled, nor included in 'files' because it is adapted
for the particular architecture. However, it never would've compiled either
as it had the old '../h/foo.h' stuff in it.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.13.28.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.14.8.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.14.8.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.14.6.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.14.6.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.15.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.16.6.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.16.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.16.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.16.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.17.10.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.17.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.18.6.2 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.18.6.1 | 21-Jan-2008 |
yamt | sync with head
|
| 1.19.60.1 | 10-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.19.52.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.19.46.1 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.96 | 07-Dec-2022 |
knakahara | gif(4), ipsec(4) and l2tp(4) use encap_attach_addr().
|
| 1.95 | 19-Sep-2019 |
knakahara | Avoid having a rtcache directly in a percpu storage for tunnel protocols.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
|
| 1.94 | 01-May-2018 |
maxv | branches: 1.94.2; 1.94.6;
Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.93 | 27-Apr-2018 |
knakahara | Fix LOCKDEBUG kernel panic when many(about 200) tunnel interfaces is created.
The tunnel interfaces are gif(4), l2tp(4), and ipsecif(4). They use mutex
itself in percpu area. When percpu_cpu_enlarge() run, the address of the
mutex in percpu area becomes different from the address which lockdebug
saved. That can cause "already initialized" false detection.
|
| 1.92 | 10-Jan-2018 |
knakahara | branches: 1.92.2;
apply in{,6}_tunnel_validate() to gif(4).
|
| 1.91 | 27-Nov-2017 |
knakahara | IFF_RUNNING checking in Rx and Tx processing is unnecessary now.
Because the configs of gif (members of gif_var) are protected by psref(9).
|
| 1.90 | 27-Nov-2017 |
knakahara | preserve gif(4) configs by psref(9) like vlan(4) and l2tp(4).
After Tx side does not use softint, gif(4) can use psref(9) for config
preservation like vlan(4) and l2tp(4).
update locking notes later.
|
| 1.89 | 15-Nov-2017 |
knakahara | Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.88 | 21-Sep-2017 |
knakahara | add lock for percpu route like l2tp(4).
|
| 1.87 | 06-Jan-2017 |
knakahara | branches: 1.87.8;
remove unnecessary conversion.
gif_softc->gif_pdst is already valid sockaddr.
|
| 1.86 | 14-Dec-2016 |
knakahara | fix race of gif_softc->gif_ro when we send multiple flows over gif on NET_MPSAFE enabled kernel.
make gif_softc->gif_ro percpu as well as ipforward_rt to resolve this race.
and add future TODO comment for etherip(4).
|
| 1.85 | 12-Dec-2016 |
ozaki-r | Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.84 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.83 | 18-Aug-2016 |
knakahara | remove unnecessary dependency on gif.h to become friendly with module and rump.
When in_gif.c become compile target, NGIF is always more than 1.
|
| 1.82 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.81 | 06-Jul-2016 |
ozaki-r | branches: 1.81.2;
Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.80 | 06-Jul-2016 |
knakahara | fix build failure when defined GIF_ENCAPCHECK
|
| 1.79 | 06-Jul-2016 |
ozaki-r | Apply m_get_rcvif_psref (kill m_get_rcvif_NOMPSAFE)
|
| 1.78 | 04-Jul-2016 |
knakahara | fix: gif(4) receive side race
A panic cause in rn_match() called by encap[46]_lookup(). The reason is that
gif(4) does not suspend receive packet processing in spite of suspending
transmit packet processing while anyone is doing gif(4) ioctl.
|
| 1.77 | 04-Jul-2016 |
knakahara | let gif(4) promise softint(9) contract (1/2) : gif(4) side
To prevent calling softint_schedule() after called softint_disestablish(),
the following modifications are added
+ ioctl (writing configuration) side
- off IFF_RUNNING flag before changing configuration
- wait softint handler completion before changing configuration
+ packet processing (reading configuraiotn) side
- if IFF_RUNNING flag is on, do nothing
+ in whole
- add gif_list_lock_{enter,exit} to prevent the same configuration is
set to other gif(4) interfaces
|
| 1.76 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.75 | 26-Jan-2016 |
knakahara | eliminate variable argument in encapsw
|
| 1.74 | 26-Jan-2016 |
knakahara | implement encapsw instead of protosw and uniform prototype.
suggested and advised by riastradh@n.o, thanks.
BTW, It seems in_stf_input() had bugs...
|
| 1.73 | 23-Jan-2016 |
riastradh | Those were local changes not meant to be part of the revert. SORRY!
|
| 1.72 | 23-Jan-2016 |
christos | fix compilation
|
| 1.71 | 22-Jan-2016 |
riastradh | Back out previous change to introduce struct encapsw.
This change was intended, but Nakahara-san had already made a better
one locally! So I'll let him commit that one, and I'll try not to
step on anyone's toes again.
|
| 1.70 | 22-Jan-2016 |
riastradh | Don't abuse struct protosw for ip_encap -- introduce struct encapsw.
Mostly mechanical change to replace it, culling some now-needless
boilerplate around all the users.
This does not substantively change the ip_encap API or eliminate
abuse of sketchy pointer casts -- that will come later, and will be
easier now that it is not tangled up with struct protosw.
|
| 1.69 | 20-Jan-2016 |
riastradh | Eliminate struct protosw::pr_output.
You can't use this unless you know what it is a priori: the formal
prototype is variadic, and the different instances (e.g., ip_output,
route_output) have different real prototypes.
Convert the only user of it, raw_send in net/raw_cb.c, to take an
explicit callback argument. Convert the only instances of it,
route_output and key_output, to such explicit callbacks for raw_send.
Use assertions to make sure the conversion to explicit callbacks is
warranted.
Discussed on tech-net with no objections:
https://mail-index.netbsd.org/tech-net/2016/01/16/msg005484.html
|
| 1.68 | 18-Jan-2016 |
knakahara | Refactor protosw codes in gif(4). No functional change.
- remove unnecessary include
- reduce scopes
|
| 1.67 | 25-Dec-2015 |
knakahara | use satosin{,6} macros instead of casts.
|
| 1.66 | 11-Dec-2015 |
knakahara | PR kern/50522: gif(4) ioctl causes panic while someone is using the gif(4) interface.
It is required to wait other CPU's softint completion before disestablishing
the softint handler.
|
| 1.65 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.64 | 18-May-2014 |
rmind | branches: 1.64.4;
Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.63 | 01-Mar-2013 |
joerg | branches: 1.63.6; 1.63.10;
Retire OSI network stack. OK core@
|
| 1.62 | 09-Jan-2012 |
liamjfoy | branches: 1.62.6;
check against NULL
|
| 1.61 | 17-Jul-2011 |
joerg | branches: 1.61.2; 1.61.6;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.60 | 07-Nov-2008 |
dyoung | *** Summary ***
When a link-layer address changes (e.g., ifconfig ex0 link
02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor
Advertisement to update the network-/link-layer address bindings
on our LAN peers.
Refuse a change of ethernet address to the address 00:00:00:00:00:00
or to any multicast/broadcast address. (Thanks matt@.)
Reorder ifnet ioctl operations so that driver ioctls may inherit
the functions of their "class"---ether_ioctl(), fddi_ioctl(), et
cetera---and the class ioctls may inherit from the generic ioctl,
ifioctl_common(), but both driver- and class-ioctls may override
the generic behavior. Make network drivers share more code.
Distinguish a "factory" link-layer address from others for the
purposes of both protecting that address from deletion and computing
EUI64.
Return consistent, appropriate error codes from network drivers.
Improve readability. KNF.
*** Details ***
In if_attach(), always initialize the interface ioctl routine,
ifnet->if_ioctl, if the driver has not already initialized it.
Delete if_ioctl == NULL tests everywhere else, because it cannot
happen.
In the ioctl routines of network interfaces, inherit common ioctl
behaviors by calling either ifioctl_common() or whichever ioctl
routine is appropriate for the class of interface---e.g., ether_ioctl()
for ethernets.
Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In
the user->kernel interface, SIOCSIFADDR's argument was an ifreq,
but on the protocol->ifnet interface, SIOCSIFADDR's argument was
an ifaddr. That was confusing, and it would work against me as I
make it possible for a network interface to overload most ioctls.
On the protocol->ifnet interface, replace SIOCSIFADDR with
SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to
invoke SIOCINITIFADDR.
In ifioctl(), give the interface the first shot at handling most
interface ioctls, and give the protocol the second shot, instead
of the other way around. Finally, let compatibility code (COMPAT_OSOCK)
take a shot.
Pull device initialization out of switch statements under
SIOCINITIFADDR. For example, pull ..._init() out of any switch
statement that looks like this:
switch (...->sa_family) {
case ...:
..._init();
...
break;
...
default:
..._init();
...
break;
}
Rewrite many if-else clauses that handle all permutations of IFF_UP
and IFF_RUNNING to use a switch statement,
switch (x & (IFF_UP|IFF_RUNNING)) {
case 0:
...
break;
case IFF_RUNNING:
...
break;
case IFF_UP:
...
break;
case IFF_UP|IFF_RUNNING:
...
break;
}
unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and
#ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4).
In ipw(4), remove an if_set_sadl() call that is out of place.
In nfe(4), reuse the jumbo MTU logic in ether_ioctl().
Let ethernets register a callback for setting h/w state such as
promiscuous mode and the multicast filter in accord with a change
in the if_flags: ether_set_ifflags_cb() registers a callback that
returns ENETRESET if the caller should reset the ethernet by calling
if_init(), 0 on success, != 0 on failure. Pull common code from
ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(),
and register if_flags callbacks for those drivers.
Return ENOTTY instead of EINVAL for inappropriate ioctls. In
zyd(4), use ENXIO instead of ENOTTY to indicate that the device is
not any longer attached.
Add to if_set_sadl() a boolean 'factory' argument that indicates
whether a link-layer address was assigned by the factory or some
other source. In a comment, recommend using the factory address
for generating an EUI64, and update in6_get_hw_ifid() to prefer a
factory address to any other link-layer address.
Add a routing message, RTM_LLINFO_UPD, that tells protocols to
update the binding of network-layer addresses to link-layer addresses.
Implement this message in IPv4 and IPv6 by sending a gratuitous
ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD
messages on a change of an interface's link-layer address.
In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address
that is broadcast/multicast or equal to 00:00:00:00:00:00.
Make ether_ioctl() call ifioctl_common() to handle ioctls that it
does not understand.
In gif(4), initialize if_softc and use it, instead of assuming that
the gif_softc and ifp overlap.
Let ifioctl_common() handle SIOCGIFADDR.
Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels
that certain invariants on a struct route are satisfied.
In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit
about the ioctls that we do not allow on an agr(4) member interface.
bzero -> memset. Delete unnecessary casts to void *. Use
sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with
NULL instead of "testing truth". Replace some instances of (type
*)0 with NULL. Change some K&R prototypes to ANSI C, and join
lines.
|
| 1.59 | 12-Apr-2008 |
thorpej | branches: 1.59.4; 1.59.10; 1.59.12;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.58 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.57 | 20-Dec-2007 |
dyoung | branches: 1.57.6;
Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.56 | 02-May-2007 |
dyoung | branches: 1.56.8; 1.56.16; 1.56.20;
Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.55 | 04-Mar-2007 |
christos | branches: 1.55.2; 1.55.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.54 | 17-Feb-2007 |
dyoung | branches: 1.54.2;
bzero -> memset
|
| 1.53 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.52 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.51 | 23-Nov-2006 |
rpaulo | New EtherIP driver based on tap(4) and gif(4) by Hans Rosenfeld.
Notable changes:
* Fixes PR 34268.
* Separates the code from gif(4) (which is more cleaner).
* Allows the usage of STP (Spanning Tree Protocol).
* Removed EtherIP implementation from gif(4)/tap(4).
Some input from Christos.
|
| 1.50 | 28-Jul-2006 |
dyoung | branches: 1.50.4; 1.50.6;
Extract predicate M_UNWRITABLE(m, len), which is true iff len
consecutive bytes at the front of m are writable (i.e., neither
shared nor read-only).
|
| 1.49 | 28-Jul-2006 |
dyoung | Fix mtod() usage. If we will write to the mbuf data, check whether
the data is read-only/shared and call m_pullup(). Otherwise,
extract a const pointer to the mbuf data.
XXX I should extract a new macro, M_WRITABLE(m, len), that is true
if m has len consecutive writable bytes at its front.
|
| 1.48 | 28-Jul-2006 |
dyoung | Where mbuf data may be read-only/shared, use mtod(m, const ...).
Annotate a comparison and m_pullup() that seem unnecessary.
|
| 1.47 | 07-Jun-2006 |
kardel | merge FreeBSD timecounters from branch simonb-timecounters
- struct timeval time is gone
time.tv_sec -> time_second
- struct timeval mono_time is gone
mono_time.tv_sec -> time_uptime
- access to time via
{get,}{micro,nano,bin}time()
get* versions are fast but less precise
- support NTP nanokernel implementation (NTP API 4)
- further reading:
Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf
NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
|
| 1.46 | 11-Dec-2005 |
christos | branches: 1.46.4; 1.46.6; 1.46.8; 1.46.14;
merge ktrace-lwp.
|
| 1.45 | 26-Jun-2005 |
mlelstv | branches: 1.45.2;
expire cached route. Fixes PR 22792.
|
| 1.44 | 02-Jun-2005 |
tron | Change the first argument of the encapsulation check function from
"const struct mbuf *" to "struct mbuf *". Without this change the
actual implementation cannot even use m_copydata() on the mbuf chain
which is broken.
|
| 1.43 | 02-Jun-2005 |
tron | Remove type casts and lint directives which are now longer necessary
because the first argument of m_copydata() is "const struct mbuf *" now.
|
| 1.42 | 29-May-2005 |
christos | - add const
- remove bogus casts
- avoid nested variables
|
| 1.41 | 26-Feb-2005 |
perry | branches: 1.41.2;
nuke trailing whitespace
|
| 1.40 | 03-Feb-2005 |
perry | ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.39 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.38 | 01-Feb-2005 |
he | Fix "unused local variable" warning/error if compiling without
bridge support by making variable declaration conditional. Found
while compiling for shark.
|
| 1.37 | 31-Jan-2005 |
kim | Add RFC 3378 EtherIP support, ported from OpenBSD to NetBSD by
Hans Rosenfeld (rosenfeld at grumpf.hope-2000.org)
This change makes it possible to add gif interfaces to bridges, which
will then send and receive IP protocol 97 packets. Packets are Ethernet
frames with an EtherIP header prepended.
|
| 1.36 | 26-Apr-2004 |
matt | branches: 1.36.4; 1.36.6;
Remove #else clause of __STDC__
|
| 1.35 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.34 | 11-Nov-2003 |
jonathan | branches: 1.34.4;
Change global head-of-local-IP-address list from in_ifaddr to
in_ifaddrhead. Recent changes in struct names caused a namespace
collision in fast-ipsec, which are most cleanly fixed by using
"in_ifaddrhead" as the listhead name.
|
| 1.33 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.32 | 11-Nov-2002 |
itojun | branches: 1.32.6;
make USE_ENCAPCHECK (in netinet*/*gif.c) to global option, GIF_ENCAPCHECK.
#ifdef out unneeded code when possible.
From: Krister Walfridsson <cato@df.lth.se>
|
| 1.31 | 05-Nov-2002 |
itojun | improve gif lookup performance, when there are many of those,
by using radix tree for lookups. tested by yshimizu@iij.
|
| 1.30 | 11-Sep-2002 |
itojun | KNF - return is not a function. sync w/kame.
|
| 1.29 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.28 | 14-Jul-2002 |
itojun | need to bzero() before rtalloc. KAME PR 432
|
| 1.27 | 09-Jun-2002 |
itojun | whitespace
|
| 1.26 | 21-Dec-2001 |
itojun | branches: 1.26.8; 1.26.10;
move protosw fragment for gif/stf to their own source code.
reduce #ifdef in stf code. sync with kame
|
| 1.25 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.24 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.23 | 16-Aug-2001 |
itojun | branches: 1.23.4;
gif interface now uses generic software interrupt
(on archs that support it). also, make gif ALTQ-capable on outgoing.
sync with kame, comments from thorpej.
|
| 1.22 | 29-Jul-2001 |
itojun | sync gif interface code with latest kame.
IFF_RUNNING is clearified. attach/detach logic is more clearner.
the old code mistakenly set IFF_UP by itself, now the behavior is gone.
|
| 1.21 | 14-May-2001 |
itojun | branches: 1.21.2;
drop multi destination mode (IFF_LINK0).
|
| 1.20 | 10-May-2001 |
itojun | correct ecn consideration on tunnel encap/decap. sync with kame.
|
| 1.19 | 20-Feb-2001 |
itojun | branches: 1.19.2;
add AF_ISO case to output. from chopps.
|
| 1.18 | 20-Feb-2001 |
itojun | ISO over IPv4/v6 by EON encapsulation. from chopps, sync with kame.
|
| 1.17 | 22-Jan-2001 |
itojun | revert revision 1.15 (on ingress, DF bit copied from inner to outer).
since we do not have feedback mechanism from path MTU to tunnel MTU
(not sure if we should), and inner packet source will not get informed of
outer PMTUD (we shouldn't do this), 1.15 behavior can lead us to
blackhole behavior.
configurable behavior (as suggested in RFC2401 6.1) would be nice to have,
however, reusing net.inet.ipsec.dfbit would be hairy.
|
| 1.16 | 22-Jan-2001 |
itojun | make it possible to turn off ingress filter on gif/stf tunnel egress,
by using IFF_LINK2. (part of) PR 11163 from Ken Raeburn.
|
| 1.15 | 05-Jul-2000 |
thorpej | RFCs 1853, 2003, 2401 -- copy the DF bit.
|
| 1.14 | 26-Apr-2000 |
itojun | branches: 1.14.4;
sync with more recent kame. defer inclusion of net/if_gif.h.
|
| 1.13 | 20-Apr-2000 |
enami | IN_MULTICAST() takes in_addr.s_addr as argument, not pointer to it.
|
| 1.12 | 19-Apr-2000 |
itojun | introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.11 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.10 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.9 | 07-Feb-2000 |
itojun | s/DIAGNOSTIC/DEBUG/
|
| 1.8 | 06-Jan-2000 |
itojun | remove too much portability code in KAME, to improve readability.
|
| 1.7 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.6 | 20-Aug-1999 |
itojun | branches: 1.6.2; 1.6.8;
do not capture packets by gif, when gif interface is down.
|
| 1.5 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.4 | 06-Jul-1999 |
itojun | sync with KAME/NetBSD 1.4, SNAP kit 19990705.
key changes are:
- icmp6 redirect fix (dst check)
- revised ip6 multicast check for loopback i/f
- several RCS ID cleanups
|
| 1.3 | 03-Jul-1999 |
thorpej | RCS ID police.
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 28-Jun-1999 |
itojun | branches: 1.1.2;
file in_gif.c was initially added on branch kame.
|
| 1.1.2.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.1.2.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.1.2.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 01-Jul-1999 |
thorpej | file in_gif.c was added on branch chs-ubc2 on 1999-07-01 23:47:00 +0000
|
| 1.6.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.6.2.3 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.6.2.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.6.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.14.4.1 | 01-May-2001 |
he | Pull up revision 1.16 (via patch, requested by itojun):
Make it possible to turn off ingress filter on gif/stf tunnel
egress by using IFF_LINK2. Fixes (part of) PR#11163.
|
| 1.19.2.10 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.19.2.9 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.19.2.8 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.19.2.7 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.19.2.6 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.19.2.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.19.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.19.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.19.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.19.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.21.2.6 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.21.2.5 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.21.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.21.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.21.2.2 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.21.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.23.4.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.26.10.1 | 15-Jul-2002 |
thorpej | pullup-1-6 ticket #506 (itojun).
Original log message:
need to bzero() before rtalloc. KAME PR 432.
|
| 1.26.8.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.26.8.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.26.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.32.6.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.32.6.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.32.6.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.32.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.32.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.32.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.34.4.1 | 09-Jan-2006 |
tron | Pull up following revision(s) (requested by mlelstv in ticket #10214):
sys/netinet6/in6_gif.c: revision 1.43
sys/netinet/in_gif.c: revision 1.45
sys/net/if_gif.h: revision 1.11
expire cached route. Fixes PR 22792.
|
| 1.36.6.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.36.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.36.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.41.2.1 | 08-Jan-2006 |
riz | Pull up following revision(s) (requested by mlelstv in ticket #1092):
sys/netinet6/in6_gif.c: revision 1.43
sys/netinet/in_gif.c: revision 1.45
sys/net/if_gif.h: revision 1.11
expire cached route. Fixes PR 22792.
|
| 1.45.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.45.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.45.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.45.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.45.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.46.14.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.46.8.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.46.8.1 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.46.6.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time(), use bintime's for timeout
calculations and use "time_second" instead of "time.tv_sec".
|
| 1.46.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.50.6.2 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.50.6.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.50.4.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.54.2.3 | 07-May-2007 |
yamt | sync with head.
|
| 1.54.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.54.2.1 | 17-Feb-2007 |
rmind | file in_gif.c was added on branch yamt-idlelwp on 2007-03-12 05:59:36 +0000
|
| 1.55.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.55.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.56.20.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.56.16.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.56.8.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.57.6.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.57.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.59.12.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.59.10.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.59.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.61.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.61.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.61.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.62.6.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.62.6.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.62.6.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.63.10.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.63.6.1 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.64.4.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.64.4.5 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.64.4.4 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.64.4.3 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.64.4.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.64.4.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.81.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.81.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.87.8.6 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1385):
sys/net/if.c 1.461
sys/net/if.h 1.277
sys/net/if_gif.c 1.149
sys/net/if_gif.h 1.33
sys/net/if_ipsec.c 1.19,1.20,1.24
sys/net/if_ipsec.h 1.5
sys/net/if_l2tp.c 1.33,1.36-1.39
sys/net/if_l2tp.h 1.7,1.8
sys/net/route.c 1.220,1.221
sys/net/route.h 1.125
sys/netinet/in_gif.c 1.95
sys/netinet/in_l2tp.c 1.17
sys/netinet/ip_input.c 1.391,1.392
sys/netinet/wqinput.c 1.6
sys/netinet6/in6_gif.c 1.94
sys/netinet6/in6_l2tp.c 1.18
sys/netinet6/ip6_forward.c 1.97
sys/netinet6/ip6_input.c 1.210,1.211
sys/netipsec/ipsec_output.c 1.82,1.83 (patched)
sys/netipsec/ipsecif.c 1.12,1.13,1.15,1.17 (patched)
sys/netipsec/key.c 1.259,1.260
ipsecif(4) support input drop packet counter.
ipsecif(4) should not increment drop counter by errors not related to if_snd. Pointed out by ozaki-r@n.o, thanks.
Remove unnecessary addresses in PF_KEY message.
MOBIKE Extensions for PF_KEY draft-schilcher-mobike-pfkey-extension-01.txt says
|
| 1.87.8.5 | 17-May-2018 |
martin | Pull up following revision(s) (requested by knakahara in ticket #829):
sys/net/if_l2tp.c: revision 1.24
sys/net/if_ipsec.c: revision 1.13
sys/net/if_gif.h: revision 1.31
sys/netipsec/ipsecif.c: revision 1.8
sys/net/if_gif.c: revision 1.140
sys/netinet6/in6_l2tp.c: revision 1.15
sys/net/if_ipsec.h: revision 1.3
sys/netinet6/in6_gif.c: revision 1.92
sys/net/if_l2tp.h: revision 1.5
sys/netinet/in_l2tp.c: revision 1.13
sys/netinet/in_gif.c: revision 1.93
Fix LOCKDEBUG kernel panic when many(about 200) tunnel interfaces is created.
The tunnel interfaces are gif(4), l2tp(4), and ipsecif(4). They use mutex
itself in percpu area. When percpu_cpu_enlarge() run, the address of the
mutex in percpu area becomes different from the address which lockdebug
saved. That can cause "already initialized" false detection.
|
| 1.87.8.4 | 11-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #536):
distrib/sets/lists/base/shl.mi: 1.825
distrib/sets/lists/comp/mi: 1.2168-1.2169
distrib/sets/lists/comp/shl.mi: 1.310
distrib/sets/lists/debug/mi: 1.234
distrib/sets/lists/debug/shl.mi: 1.188
distrib/sets/lists/man/mi: 1.1570
distrib/sets/lists/tests/mi: 1.772
etc/mtree/NetBSD.dist.tests: 1.150
share/man/man4/Makefile: 1.650
share/man/man4/ipsec.4: 1.42-1.43
share/man/man4/ipsecif.4: 1.1-1.5
sys/arch/amd64/conf/ALL: 1.77
sys/arch/amd64/conf/GENERIC: 1.480
sys/conf/files: 1.1191
sys/net/Makefile: 1.34
sys/net/files.net: 1.14
sys/net/if.c: 1.404
sys/net/if.h: 1.248
sys/net/if_gif.c: 1.135
sys/net/if_ipsec.c: 1.1-1.3
sys/net/if_ipsec.h: 1.1
sys/net/if_l2tp.c: 1.16
sys/net/if_types.h: 1.28
sys/netinet/in.c: 1.214
sys/netinet/in.h: 1.103
sys/netinet/in_gif.c: 1.92
sys/netinet/ip_var.h: 1.122
sys/netinet6/in6.c: 1.257
sys/netinet6/in6.h: 1.88
sys/netinet6/in6_gif.c: 1.90
sys/netinet6/ip6_var.h: 1.75
sys/netipsec/Makefile: 1.6
sys/netipsec/files.netipsec: 1.13
sys/netipsec/ipsec.h: 1.62
sys/netipsec/ipsecif.c: 1.1
sys/netipsec/ipsecif.h: 1.1
sys/netipsec/key.c: 1.246-1.247
sys/netipsec/key.h: 1.34
sys/rump/net/Makefile.rumpnetcomp: 1.20
sys/rump/net/lib/libipsec/IPSEC.ioconf: 1.1
sys/rump/net/lib/libipsec/Makefile: 1.1
sys/rump/net/lib/libipsec/ipsec_component.c: 1.1
tests/net/Makefile: 1.34
tests/net/if_ipsec/Makefile: 1.1
tests/net/if_ipsec/t_ipsec.sh: 1.1-1.2
Don't touch an SP without a reference to it
unify processing to check nesting count for some tunnel protocols.
add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
ipsec(4) interface supports rump now.
add ipsec(4) interface ATF.
add ipsec(4) interface man as ipsecif.4.
add ipsec(4) interface to amd64/GENERIC and amd64/ALL configs.
apply in{,6}_tunnel_validate() to gif(4).
Spell IPsec that way. Simplify macro usage. Sort SEE ALSO. Bump
date for previous.
Improve wording and macro use.
Some parts are not clear to me, so someone with knowledge of ipsecif(4)
should improve this some more.
Improve ipsecif.4. Default port ipsec(4) NAT-T is tested now.
pointed out by wiz@n.o and suggested by ozaki-r@n.o, thanks.
Change the prefix of test names to ipsecif_ to distinguish from tests for ipsec(4)
New sentence, new line. Remove empty macro.
Fix PR kern/52920. Pointed out by David Binderman, thanks.
Improve wording, and put a new drawing, from me and Kengo Nakahara.
apply a little more #ifdef INET/INET6. fixes !INET6 builds.
|
| 1.87.8.3 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by knakahara in ticket #462):
sys/net/if_gif.c: revision 1.133, 1.134, 1.137
sys/net/if_gif.h: revision 1.28-1.29
sys/netinet/in_gif.c: revision 1.90-1.91
sys/netinet/in_gif.h: revision 1.18
sys/netinet6/in6_gif.c: revision 1.88-1.89
sys/netinet6/in6_gif.h: revision 1.17
preserve gif(4) configs by psref(9) like vlan(4) and l2tp(4).
After Tx side does not use softint, gif(4) can use psref(9) for config
preservation like vlan(4) and l2tp(4).
update locking notes later.
--
update gif(4) locking notes.
--
IFF_RUNNING checking in Rx and Tx processing is unnecessary now.
Because the configs of gif (members of gif_var) are protected by psref(9).
--
remove duplicated null ckeck
|
| 1.87.8.2 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by knakahara in ticket #419):
sys/net/if_stf.c: revision 1.103
sys/net/if_stf.h: revision 1.8
sys/netinet/in_gif.c: revision 1.89
sys/netinet/in_gif.h: revision 1.17
sys/netinet/in_l2tp.c: revision 1.4
sys/netinet/ip_encap.c: revision 1.66
sys/netinet/ip_encap.h: revision 1.23
sys/netinet/ip_mroute.c: revision 1.148
sys/netinet6/in6_gif.c: revision 1.87
sys/netinet6/in6_gif.h: revision 1.16
sys/netinet6/in6_l2tp.c: revision 1.7
sys/netipsec/xform.h: revision 1.13
sys/netipsec/xform_ipip.c: revision 1.55
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.87.8.1 | 24-Oct-2017 |
snj | Pull up following revision(s) (requested by knahakara in ticket #303):
sys/net/if_gif.c: 1.129-1.130
sys/net/if_gif.h: 1.26-1.27
sys/netinet/in_gif.c: 1.88
sys/netinet6/in6_gif.c: 1.86
add lock for percpu route like l2tp(4).
--
add lock for sclist to exclude ifconfig gifX add/delete and ifconfig gifX tunnel
--
update locking notes.
|
| 1.92.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.94.6.1 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #238):
sys/netipsec/ipsec_output.c: revision 1.83
sys/net/route.h: revision 1.125
sys/netinet6/ip6_input.c: revision 1.210
sys/netinet6/ip6_input.c: revision 1.211
sys/net/if.c: revision 1.461
sys/net/if_gif.h: revision 1.33
sys/net/route.c: revision 1.220
sys/net/route.c: revision 1.221
sys/net/if.h: revision 1.277
sys/netinet6/ip6_forward.c: revision 1.97
sys/netinet/wqinput.c: revision 1.6
sys/net/if_ipsec.h: revision 1.5
sys/netinet6/in6_l2tp.c: revision 1.18
sys/netinet6/in6_gif.c: revision 1.94
sys/net/if_l2tp.h: revision 1.7
sys/net/if_gif.c: revision 1.149
sys/net/if_l2tp.h: revision 1.8
sys/netinet/in_gif.c: revision 1.95
sys/netinet/in_l2tp.c: revision 1.17
sys/netipsec/ipsecif.c: revision 1.17
sys/net/if_ipsec.c: revision 1.24
sys/net/if_l2tp.c: revision 1.37
sys/netinet/ip_input.c: revision 1.391
sys/net/if_l2tp.c: revision 1.38
sys/netinet/ip_input.c: revision 1.392
sys/net/if_l2tp.c: revision 1.39
Avoid having a rtcache directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
wqinput: avoid having struct wqinput_worklist directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Input handlers of wqinput normally involves sleepable operations so we must
avoid dereferencing a percpu data (struct wqinput_worklist) after executing
an input handler. Address this situation by having just a pointer to the data
in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
Add missing #include <sys/kmem.h>
-
Divide Tx context of l2tp(4) to improve performance.
It seems l2tp(4) call path is too long for instruction cache. So, dividing
l2tp(4) Tx context improves CPU use efficiency.
After this commit, l2tp(4) throughput gains 10% on my machine(Atom C3000).
-
Apply some missing changes lost on the previous commit
-
Avoid having a rtcache directly in a percpu storage for tunnel protocols.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
-
l2tp(4): avoid having struct ifqueue directly in a percpu storage.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Tx processing of l2tp(4) uses normally involves sleepable operations so we
must avoid dereferencing a percpu data (struct ifqueue) after executing Tx
processing. Address this situation by having just a pointer to the data in
a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
|
| 1.94.2.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.18 | 27-Nov-2017 |
knakahara | preserve gif(4) configs by psref(9) like vlan(4) and l2tp(4).
After Tx side does not use softint, gif(4) can use psref(9) for config
preservation like vlan(4) and l2tp(4).
update locking notes later.
|
| 1.17 | 15-Nov-2017 |
knakahara | Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.16 | 04-Jul-2016 |
knakahara | branches: 1.16.10;
fix: gif(4) receive side race
A panic cause in rn_match() called by encap[46]_lookup(). The reason is that
gif(4) does not suspend receive packet processing in spite of suspending
transmit packet processing while anyone is doing gif(4) ioctl.
|
| 1.15 | 26-Jan-2016 |
knakahara | eliminate variable argument in encapsw
|
| 1.14 | 23-Nov-2006 |
rpaulo | branches: 1.14.98; 1.14.118;
New EtherIP driver based on tap(4) and gif(4) by Hans Rosenfeld.
Notable changes:
* Fixes PR 34268.
* Separates the code from gif(4) (which is more cleaner).
* Allows the usage of STP (Spanning Tree Protocol).
* Removed EtherIP implementation from gif(4)/tap(4).
Some input from Christos.
|
| 1.13 | 10-Dec-2005 |
elad | branches: 1.13.20; 1.13.22;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.12 | 06-Jun-2005 |
martin | branches: 1.12.2;
Since we decided "const struct mbuf *" would not do the right thing (tm),
remove ~all const from mbuf pointers.
|
| 1.11 | 31-Jan-2005 |
kim | Add RFC 3378 EtherIP support, ported from OpenBSD to NetBSD by
Hans Rosenfeld (rosenfeld at grumpf.hope-2000.org)
This change makes it possible to add gif interfaces to bridges, which
will then send and receive IP protocol 97 packets. Packets are Ethernet
frames with an EtherIP header prepended.
|
| 1.10 | 21-Apr-2004 |
itojun | branches: 1.10.4; 1.10.6;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.9 | 18-Apr-2004 |
matt | De __P()
|
| 1.8 | 11-Nov-2002 |
itojun | branches: 1.8.6;
make USE_ENCAPCHECK (in netinet*/*gif.c) to global option, GIF_ENCAPCHECK.
#ifdef out unneeded code when possible.
From: Krister Walfridsson <cato@df.lth.se>
|
| 1.7 | 16-Aug-2001 |
itojun | gif interface now uses generic software interrupt
(on archs that support it). also, make gif ALTQ-capable on outgoing.
sync with kame, comments from thorpej.
|
| 1.6 | 29-Jul-2001 |
itojun | sync gif interface code with latest kame.
IFF_RUNNING is clearified. attach/detach logic is more clearner.
the old code mistakenly set IFF_UP by itself, now the behavior is gone.
|
| 1.5 | 19-Apr-2000 |
itojun | branches: 1.5.6; 1.5.8;
introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.4 | 06-Jul-1999 |
itojun | branches: 1.4.2;
sync with KAME/NetBSD 1.4, SNAP kit 19990705.
key changes are:
- icmp6 redirect fix (dst check)
- revised ip6 multicast check for loopback i/f
- several RCS ID cleanups
|
| 1.3 | 03-Jul-1999 |
thorpej | RCS ID police.
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 28-Jun-1999 |
itojun | branches: 1.1.2;
file in_gif.h was initially added on branch kame.
|
| 1.1.2.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.1.2.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 01-Jul-1999 |
thorpej | file in_gif.h was added on branch chs-ubc2 on 1999-07-01 23:47:00 +0000
|
| 1.4.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.5.8.2 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.5.8.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.5.6.2 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.5.6.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.8.6.6 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.8.6.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.8.6.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.8.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.8.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.8.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.10.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.10.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.12.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.12.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.13.22.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.13.20.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.14.118.2 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.14.118.1 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.14.98.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.16.10.2 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by knakahara in ticket #462):
sys/net/if_gif.c: revision 1.133, 1.134, 1.137
sys/net/if_gif.h: revision 1.28-1.29
sys/netinet/in_gif.c: revision 1.90-1.91
sys/netinet/in_gif.h: revision 1.18
sys/netinet6/in6_gif.c: revision 1.88-1.89
sys/netinet6/in6_gif.h: revision 1.17
preserve gif(4) configs by psref(9) like vlan(4) and l2tp(4).
After Tx side does not use softint, gif(4) can use psref(9) for config
preservation like vlan(4) and l2tp(4).
update locking notes later.
--
update gif(4) locking notes.
--
IFF_RUNNING checking in Rx and Tx processing is unnecessary now.
Because the configs of gif (members of gif_var) are protected by psref(9).
--
remove duplicated null ckeck
|
| 1.16.10.1 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by knakahara in ticket #419):
sys/net/if_stf.c: revision 1.103
sys/net/if_stf.h: revision 1.8
sys/netinet/in_gif.c: revision 1.89
sys/netinet/in_gif.h: revision 1.17
sys/netinet/in_l2tp.c: revision 1.4
sys/netinet/ip_encap.c: revision 1.66
sys/netinet/ip_encap.h: revision 1.23
sys/netinet/ip_mroute.c: revision 1.148
sys/netinet6/in6_gif.c: revision 1.87
sys/netinet6/in6_gif.h: revision 1.16
sys/netinet6/in6_l2tp.c: revision 1.7
sys/netipsec/xform.h: revision 1.13
sys/netipsec/xform_ipip.c: revision 1.55
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.1 | 13-Nov-2006 |
dyoung | branches: 1.1.2; 1.1.6; 1.1.8;
Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.1.8.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.8.1 | 13-Nov-2006 |
yamt | file in_ifattach.h was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.1.6.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.1.6.1 | 13-Nov-2006 |
yamt | file in_ifattach.h was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.1.2.2 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.1.2.1 | 13-Nov-2006 |
ad | file in_ifattach.h was added on branch newlock2 on 2006-11-18 21:39:36 +0000
|
| 1.22 | 01-Sep-2023 |
andvar | fix typos in comments, mainly s/innner/inner/.
|
| 1.21 | 07-Dec-2022 |
knakahara | gif(4), ipsec(4) and l2tp(4) use encap_attach_addr().
|
| 1.20 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.19 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.18 | 29-Jan-2020 |
thorpej | branches: 1.18.6;
Adopt <net/if_stats.h>.
|
| 1.17 | 19-Sep-2019 |
knakahara | branches: 1.17.2;
Avoid having a rtcache directly in a percpu storage for tunnel protocols.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
|
| 1.16 | 03-Sep-2018 |
knakahara | branches: 1.16.4;
fix: l2tp(4) cannot receive packets after reset session without reset tunnel. Pointed out by k-goda@IIJ
When the following operations are done after established session, the l2tp0
cannot receive packets until done deletetunnel && tunnel "src" "dst".
|
| 1.15 | 21-Jun-2018 |
knakahara | branches: 1.15.2;
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.14 | 01-May-2018 |
maxv | Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.13 | 27-Apr-2018 |
knakahara | Fix LOCKDEBUG kernel panic when many(about 200) tunnel interfaces is created.
The tunnel interfaces are gif(4), l2tp(4), and ipsecif(4). They use mutex
itself in percpu area. When percpu_cpu_enlarge() run, the address of the
mutex in percpu area becomes different from the address which lockdebug
saved. That can cause "already initialized" false detection.
|
| 1.12 | 26-Jan-2018 |
maxv | branches: 1.12.2;
Several fixes in L2TP:
* l2tp_input(): use m_copydata, and ensure there is enough space in the
chain. Otherwise overflow.
* l2tp_tcpmss_clamp(): ensure there is enough space in the chain.
* in_l2tp_output(): don't check 'sc' against NULL, it can't be NULL.
* in_l2tp_input(): no need to call m_pullup since we use m_copydata.
Just check the space in the chain.
* in_l2tp_input(): if there is a cookie, make sure the chain has enough
space.
* in6_l2tp_input(): same changes as in_l2tp_input().
Ok knakahara@
|
| 1.11 | 25-Jan-2018 |
maxv | Style, reduce the indentation level when possible, and add a missing NULL
check after M_PREPEND.
|
| 1.10 | 22-Jan-2018 |
maxv | Fix null deref, m could be NULL if M_PREPEND fails.
|
| 1.9 | 18-Dec-2017 |
knakahara | fix mbuf leaks. pointed out and suggested by kre@n.o, thanks.
|
| 1.8 | 18-Dec-2017 |
knakahara | backout wrong fix again, sorry.
|
| 1.7 | 15-Dec-2017 |
knakahara | Fix pullup'ed mbuf leaks. The match function just requires enough mbuf length.
XXX need pullup-8
|
| 1.6 | 15-Dec-2017 |
knakahara | backout wrong fix as it causes atf net/ipsec/t_ipsec_l2tp failures.
|
| 1.5 | 11-Dec-2017 |
knakahara | fix pullup'ed mbuf leaks. pointed out by maxv@n.o, thanks.
XXX need pullup-8
|
| 1.4 | 15-Nov-2017 |
knakahara | branches: 1.4.2;
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.3 | 11-Jul-2017 |
knakahara | branches: 1.3.4;
l2tp(4): fix mbuf leak when tunnel nested over the limit
XXX need pullup -8 branch
|
| 1.2 | 30-Mar-2017 |
knakahara | branches: 1.2.4; 1.2.8;
remove duplicated validation. That is already done in l2tp_lookup_session_ref().
pointed out by s-yamaguchi@IIJ, thanks.
|
| 1.1 | 16-Feb-2017 |
knakahara | branches: 1.1.2;
add missing files.
|
| 1.1.2.3 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.1.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.1.2.1 | 16-Feb-2017 |
pgoyette | file in_l2tp.c was added on branch pgoyette-localcount on 2017-03-20 06:57:50 +0000
|
| 1.2.8.8 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1385):
sys/net/if.c 1.461
sys/net/if.h 1.277
sys/net/if_gif.c 1.149
sys/net/if_gif.h 1.33
sys/net/if_ipsec.c 1.19,1.20,1.24
sys/net/if_ipsec.h 1.5
sys/net/if_l2tp.c 1.33,1.36-1.39
sys/net/if_l2tp.h 1.7,1.8
sys/net/route.c 1.220,1.221
sys/net/route.h 1.125
sys/netinet/in_gif.c 1.95
sys/netinet/in_l2tp.c 1.17
sys/netinet/ip_input.c 1.391,1.392
sys/netinet/wqinput.c 1.6
sys/netinet6/in6_gif.c 1.94
sys/netinet6/in6_l2tp.c 1.18
sys/netinet6/ip6_forward.c 1.97
sys/netinet6/ip6_input.c 1.210,1.211
sys/netipsec/ipsec_output.c 1.82,1.83 (patched)
sys/netipsec/ipsecif.c 1.12,1.13,1.15,1.17 (patched)
sys/netipsec/key.c 1.259,1.260
ipsecif(4) support input drop packet counter.
ipsecif(4) should not increment drop counter by errors not related to if_snd. Pointed out by ozaki-r@n.o, thanks.
Remove unnecessary addresses in PF_KEY message.
MOBIKE Extensions for PF_KEY draft-schilcher-mobike-pfkey-extension-01.txt says
|
| 1.2.8.7 | 10-Sep-2018 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1018):
sys/netinet6/in6_l2tp.c: revision 1.17
sys/netinet/in_l2tp.c: revision 1.16
fix: l2tp(4) cannot receive packets after reset session without reset tunnel. Pointed out by k-goda@IIJ
When the following operations are done after established session, the l2tp0
cannot receive packets until done deletetunnel && tunnel "src" "dst".
|
| 1.2.8.6 | 13-Jul-2018 |
martin | Pull up following revision(s) via patch (requested by knakahara in ticket #905):
sys/netinet/ip_mroute.c: revision 1.160
sys/netinet6/in6_l2tp.c: revision 1.16
sys/net/if.h: revision 1.263
sys/netinet/in_l2tp.c: revision 1.15
sys/netinet/ip_icmp.c: revision 1.172
sys/netinet/igmp.c: revision 1.68
sys/netinet/ip_encap.c: revision 1.69
sys/netinet6/ip6_mroute.c: revision 1.129
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.2.8.5 | 17-May-2018 |
martin | Pull up following revision(s) (requested by knakahara in ticket #829):
sys/net/if_l2tp.c: revision 1.24
sys/net/if_ipsec.c: revision 1.13
sys/net/if_gif.h: revision 1.31
sys/netipsec/ipsecif.c: revision 1.8
sys/net/if_gif.c: revision 1.140
sys/netinet6/in6_l2tp.c: revision 1.15
sys/net/if_ipsec.h: revision 1.3
sys/netinet6/in6_gif.c: revision 1.92
sys/net/if_l2tp.h: revision 1.5
sys/netinet/in_l2tp.c: revision 1.13
sys/netinet/in_gif.c: revision 1.93
Fix LOCKDEBUG kernel panic when many(about 200) tunnel interfaces is created.
The tunnel interfaces are gif(4), l2tp(4), and ipsecif(4). They use mutex
itself in percpu area. When percpu_cpu_enlarge() run, the address of the
mutex in percpu area becomes different from the address which lockdebug
saved. That can cause "already initialized" false detection.
|
| 1.2.8.4 | 08-Mar-2018 |
martin | Pull up following revision(s) (requested by knakahara in ticket #614):
sys/net/if_l2tp.c: revision 1.20
sys/netinet6/in6_l2tp.c: revision 1.13
sys/netinet6/in6_l2tp.c: revision 1.14
sys/net/if_l2tp.h: revision 1.3
sys/net/if_l2tp.c: revision 1.13
sys/netinet/in_l2tp.c: revision 1.10
sys/net/if_l2tp.c: revision 1.18
sys/netinet/in_l2tp.c: revision 1.11
sys/net/if_l2tp.c: revision 1.19
sys/netinet/in_l2tp.c: revision 1.12
If if_attach() failed in the attach function, return. Add comments about if_initialize().
suggested by ozaki-r@n.o.
Fix null deref, m could be NULL if M_PREPEND fails.
style
Style, reduce the indentation level when possible, and add a missing NULL
check after M_PREPEND.
Several fixes in L2TP:
* l2tp_input(): use m_copydata, and ensure there is enough space in the
chain. Otherwise overflow.
* l2tp_tcpmss_clamp(): ensure there is enough space in the chain.
* in_l2tp_output(): don't check 'sc' against NULL, it can't be NULL.
* in_l2tp_input(): no need to call m_pullup since we use m_copydata.
Just check the space in the chain.
* in_l2tp_input(): if there is a cookie, make sure the chain has enough
space.
* in6_l2tp_input(): same changes as in_l2tp_input().
Ok knakahara@
Use MH_ALIGN instead, ok knakahara@.
|
| 1.2.8.3 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by knakahara in ticket #461):
sys/netinet/in_l2tp.c: revision 1.9
sys/netinet6/in6_l2tp.c: revision 1.12
fix mbuf leaks. pointed out and suggested by kre@n.o, thanks.
|
| 1.2.8.2 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by knakahara in ticket #419):
sys/net/if_stf.c: revision 1.103
sys/net/if_stf.h: revision 1.8
sys/netinet/in_gif.c: revision 1.89
sys/netinet/in_gif.h: revision 1.17
sys/netinet/in_l2tp.c: revision 1.4
sys/netinet/ip_encap.c: revision 1.66
sys/netinet/ip_encap.h: revision 1.23
sys/netinet/ip_mroute.c: revision 1.148
sys/netinet6/in6_gif.c: revision 1.87
sys/netinet6/in6_gif.h: revision 1.16
sys/netinet6/in6_l2tp.c: revision 1.7
sys/netipsec/xform.h: revision 1.13
sys/netipsec/xform_ipip.c: revision 1.55
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.2.8.1 | 12-Jul-2017 |
martin | Pull up following revision(s) (requested by knakahara in ticket #121):
sys/netinet6/in6_l2tp.c: revision 1.6
sys/netinet/in_l2tp.c: revision 1.3
l2tp(4): fix mbuf leak when tunnel nested over the limit
XXX need pullup -8 branch
|
| 1.2.4.2 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.2.4.1 | 30-Mar-2017 |
bouyer | file in_l2tp.c was added on branch bouyer-socketcan on 2017-04-21 16:54:05 +0000
|
| 1.3.4.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.3.4.1 | 11-Jul-2017 |
skrll | file in_l2tp.c was added on branch nick-nhusb on 2017-08-28 17:53:12 +0000
|
| 1.4.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.4.2.1 | 15-Nov-2017 |
jdolecek | file in_l2tp.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.12.2.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.12.2.2 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.12.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.15.2.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.15.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.15.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.16.4.1 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #238):
sys/netipsec/ipsec_output.c: revision 1.83
sys/net/route.h: revision 1.125
sys/netinet6/ip6_input.c: revision 1.210
sys/netinet6/ip6_input.c: revision 1.211
sys/net/if.c: revision 1.461
sys/net/if_gif.h: revision 1.33
sys/net/route.c: revision 1.220
sys/net/route.c: revision 1.221
sys/net/if.h: revision 1.277
sys/netinet6/ip6_forward.c: revision 1.97
sys/netinet/wqinput.c: revision 1.6
sys/net/if_ipsec.h: revision 1.5
sys/netinet6/in6_l2tp.c: revision 1.18
sys/netinet6/in6_gif.c: revision 1.94
sys/net/if_l2tp.h: revision 1.7
sys/net/if_gif.c: revision 1.149
sys/net/if_l2tp.h: revision 1.8
sys/netinet/in_gif.c: revision 1.95
sys/netinet/in_l2tp.c: revision 1.17
sys/netipsec/ipsecif.c: revision 1.17
sys/net/if_ipsec.c: revision 1.24
sys/net/if_l2tp.c: revision 1.37
sys/netinet/ip_input.c: revision 1.391
sys/net/if_l2tp.c: revision 1.38
sys/netinet/ip_input.c: revision 1.392
sys/net/if_l2tp.c: revision 1.39
Avoid having a rtcache directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
wqinput: avoid having struct wqinput_worklist directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Input handlers of wqinput normally involves sleepable operations so we must
avoid dereferencing a percpu data (struct wqinput_worklist) after executing
an input handler. Address this situation by having just a pointer to the data
in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
Add missing #include <sys/kmem.h>
-
Divide Tx context of l2tp(4) to improve performance.
It seems l2tp(4) call path is too long for instruction cache. So, dividing
l2tp(4) Tx context improves CPU use efficiency.
After this commit, l2tp(4) throughput gains 10% on my machine(Atom C3000).
-
Apply some missing changes lost on the previous commit
-
Avoid having a rtcache directly in a percpu storage for tunnel protocols.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
-
l2tp(4): avoid having struct ifqueue directly in a percpu storage.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Tx processing of l2tp(4) uses normally involves sleepable operations so we
must avoid dereferencing a percpu data (struct ifqueue) after executing Tx
processing. Address this situation by having just a pointer to the data in
a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
|
| 1.17.2.1 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.18.6.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.1 | 16-Feb-2017 |
knakahara | branches: 1.1.2; 1.1.6; 1.1.14; 1.1.18;
add missing files.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 16-Feb-2017 |
jdolecek | file in_l2tp.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.14.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.14.1 | 16-Feb-2017 |
skrll | file in_l2tp.h was added on branch nick-nhusb on 2017-08-28 17:53:12 +0000
|
| 1.1.6.2 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.1.6.1 | 16-Feb-2017 |
bouyer | file in_l2tp.h was added on branch bouyer-socketcan on 2017-04-21 16:54:05 +0000
|
| 1.1.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.1.2.1 | 16-Feb-2017 |
pgoyette | file in_l2tp.h was added on branch pgoyette-localcount on 2017-03-20 06:57:50 +0000
|
| 1.15 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.14 | 27-Mar-2020 |
jdolecek | branches: 1.14.28;
fix in4_cksum() panic "in4_cksum: mbuf 14 too short for IP header 20"
triggered by bridge_output() when passing packet originally for
interface supporting hw csum offload to destination interface
not supporting it
problem happens because bridge_output() is called after ether_output()
M_PREPEND() the ether_header into the mbuf chain, if there is not
enough space on the first mbuf of the chain, it ends up prepending
a new short mbuf with just ether_header
triggered by running UDP (IPv4) 'netio -u' benchmark with packet size 2 KB
XXX seems in6_undefer_cksum() should have similar fix, however I was
XXX not able to trigger the problem there
|
| 1.13 | 12-Dec-2018 |
rin | PR kern/53562
Add ether_sw_offload_[tr]x: handle TX/RX offload options in software.
Since this violates separation b/w L2 and L3/L4, new files are added
rather than having the routines in sys/net/if_ethersubr.c.
OK msaitoh thorpej
|
| 1.12 | 19-Sep-2018 |
rin | Fix in_undefer_cksum() and in6_undefer_cksum().
The 4th argument for in[46]_cksum() should be length of L4 header +
L4 payload. The previous revisions are wrong
- for IPv4 when hdrlen != 0
- for IPv6 always
These functions are used only in net/if_loop.c and
arch/powerpc/booke/dev/pq3etsec.c under some special circumferences.
This should be why the bugs have not been found until today.
OK maxv
|
| 1.11 | 11-Jul-2018 |
maxv | Add KASSERTs in in_undefer_cksum_tcpudp.
|
| 1.10 | 11-Jul-2018 |
maxv | Style, rename 'iph' -> 'ip', and reduce the diff between
in_undefer_cksum_tcpudp and the last part of in_undefer_cksum.
|
| 1.9 | 11-Jul-2018 |
maxv | Remove the callback, localify, and add a comment.
|
| 1.8 | 11-Jul-2018 |
maxv | Rename
ip_undefer_csum -> in_undefer_cksum
in_delayed_cksum -> in_undefer_cksum_tcpudp
The two previous names were inconsistent and misleading.
Put the two functions into in_offload.c. Add comments to explain what
we're doing.
The same could be done for IPv6.
|
| 1.7 | 26-Apr-2016 |
ozaki-r | branches: 1.7.16; 1.7.18;
Stop using rt_gwroute on packet sending paths
rt_gwroute of rtentry is a reference to a rtentry of the gateway
for a rtentry with RTF_GATEWAY. That was used by L2 (arp and ndp)
to look up L2 addresses. By separating L2 nexthop caches, we don't
need a route for the purpose and we can stop using rt_gwroute.
By doing so, we can reduce referencing and modifying rtentries,
which makes it easy to apply a lock (and/or psref) to the
routing table and rtentries.
One issue to do this is to keep RTF_REJECT behavior. It seems it
was broken when we moved rtalloc1 things from L2 output routines
(e.g., ether_output) to ip_hresolv_output, but (fortunately?)
it works unexpectedly. What we mistook are:
- RTF_REJECT was checked for any routes in L2 output routines,
but in ip_hresolv_output it is checked only when the route
is RTF_GATEWAY
- The RTF_REJECT check wasn't copied to IPv6 (nd6_output)
It seems that rt_gwroute checks hid the mistakes and it looked
work (unexpectedly) and removing rt_gwroute checks unveil the
issue. So we need to fix RTF_REJECT checks in ip_hresolv_output
and also add them to nd6_output.
One more point we have to care is returning an errno; we need
to mimic looutput behavior. Originally RTF_REJECT check was
done either in L2 output routines or in looutput. The latter is
applied when a reject route directs to a loopback interface.
However, now RTF_REJECT check is done before looutput so to keep
the original behavior we need to return an errno which looutput
chooses. Added rt_check_reject_route does such tweaks.
|
| 1.6 | 04-Jun-2015 |
ozaki-r | Pull out route lookups from L2 output routines
Route lookups for routes of RTF_GATEWAY were done in L2 output
routines such as ether_output, but they should be done in L3
i.e., before L2 output routines. This change places the lookups
between L3 output routines (say ip_output) and the L2 output
routines.
The change is based on dyoung's patch submitted in the thread:
https://mail-index.netbsd.org/tech-net/2013/02/01/msg003847.html
You can find out detailed investigations by dyoung about the
issue in there.
Note that the change introduces a workaround for MPLS. ether_output
knew that it needs to fill the ethertype of a frame as MPLS,
based on a tag of an original route (rtentry), but now we don't
pass it to ehter_output. So we have to tell that in another way.
We use mtag to do so for now, which introduces some overhead.
We should fix it somehow in the future.
Discussed on tech-kern and tech-net.
|
| 1.5 | 25-Apr-2011 |
yamt | branches: 1.5.14; 1.5.32;
ip_undefer_csum:
- don't forget ntohs.
- don't add hdrlen twice for l4 header offset.
- use M_CSUM_DATA_IPv4_IPHL instead of extracting it from ip header.
- simplify code.
- KNF.
|
| 1.4 | 14-Apr-2011 |
yamt | after ip_input.c rev.1.285 and 1.286, restore kernel_lock for if_output.
|
| 1.3 | 11-Dec-2010 |
matt | branches: 1.3.2;
Add routines to calculate a checkesum if the driver concludes that the
h/w can't do it.
|
| 1.2 | 24-Apr-2007 |
dyoung | branches: 1.2.56; 1.2.60;
Constify.
|
| 1.1 | 25-Nov-2006 |
yamt | branches: 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.14; 1.1.16;
move tso-by-software code to their own files. no functional changes.
|
| 1.1.16.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.1.14.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.1.10.1 | 07-May-2007 |
yamt | sync with head.
|
| 1.1.8.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.1.8.1 | 25-Nov-2006 |
ad | file in_offload.c was added on branch newlock2 on 2007-01-12 01:04:14 +0000
|
| 1.1.6.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.1.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.6.1 | 25-Nov-2006 |
yamt | file in_offload.c was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.1.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 25-Nov-2006 |
yamt | file in_offload.c was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.2.60.1 | 07-Jan-2011 |
matt | If using hardware checksum offload and the packet can't be h/w checksumed
(for whatever reason, some hardware is stupid) allow the driver to calculate
the checksum instead.
|
| 1.2.56.3 | 31-May-2011 |
rmind | sync with head
|
| 1.2.56.2 | 21-Apr-2011 |
rmind | sync with head
|
| 1.2.56.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.3.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.5.32.2 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.5.32.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.5.14.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.7.18.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.7.18.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.7.16.3 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.7.16.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.7.16.1 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.14.28.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.12 | 12-Dec-2018 |
rin | PR kern/53562
Add ether_sw_offload_[tr]x: handle TX/RX offload options in software.
Since this violates separation b/w L2 and L3/L4, new files are added
rather than having the routines in sys/net/if_ethersubr.c.
OK msaitoh thorpej
|
| 1.11 | 11-Jul-2018 |
maxv | Style, rename 'iph' -> 'ip', and reduce the diff between
in_undefer_cksum_tcpudp and the last part of in_undefer_cksum.
|
| 1.10 | 11-Jul-2018 |
maxv | Remove the callback, localify, and add a comment.
|
| 1.9 | 11-Jul-2018 |
maxv | Rename
ip_undefer_csum -> in_undefer_cksum
in_delayed_cksum -> in_undefer_cksum_tcpudp
The two previous names were inconsistent and misleading.
Put the two functions into in_offload.c. Add comments to explain what
we're doing.
The same could be done for IPv6.
|
| 1.8 | 25-Apr-2011 |
yamt | branches: 1.8.54; 1.8.56;
undefer csum in looutput.
looutput is used by various code (ether_output, mcast) to loopback packets.
|
| 1.7 | 11-Dec-2010 |
matt | branches: 1.7.2;
Add routines to calculate a checkesum if the driver concludes that the
h/w can't do it.
|
| 1.6 | 28-Nov-2007 |
dyoung | branches: 1.6.40; 1.6.44;
Move IN_NEED_CHECKSUM() to in_offload.h for re-use.
|
| 1.5 | 24-Apr-2007 |
dyoung | branches: 1.5.6; 1.5.8; 1.5.14;
Constify.
|
| 1.4 | 25-Nov-2006 |
yamt | branches: 1.4.4; 1.4.8; 1.4.10;
move tso-by-software code to their own files. no functional changes.
|
| 1.3 | 10-Dec-2005 |
elad | branches: 1.3.20; 1.3.22;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.2 | 10-Aug-2005 |
yamt | branches: 1.2.6;
move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.1 | 18-Apr-2005 |
yamt | branches: 1.1.2; 1.1.4;
add a function to handle M_CSUM_TSOv4 by software.
|
| 1.1.4.4 | 07-Dec-2007 |
yamt | sync with head
|
| 1.1.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.1.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.1.2.2 | 29-Apr-2005 |
kent | sync with -current
|
| 1.1.2.1 | 18-Apr-2005 |
kent | file in_offload.h was added on branch kent-audio2 on 2005-04-29 11:29:33 +0000
|
| 1.2.6.3 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.2.6.2 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.2.6.1 | 10-Aug-2005 |
skrll | file in_offload.h was added on branch ktrace-lwp on 2005-11-10 14:11:07 +0000
|
| 1.3.22.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.3.20.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.4.10.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.4.8.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.4.4.1 | 07-May-2007 |
yamt | sync with head.
|
| 1.5.14.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.5.8.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.5.6.1 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.6.44.1 | 07-Jan-2011 |
matt | If using hardware checksum offload and the packet can't be h/w checksumed
(for whatever reason, some hardware is stupid) allow the driver to calculate
the checksum instead.
|
| 1.6.40.2 | 31-May-2011 |
rmind | sync with head
|
| 1.6.40.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.7.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.8.56.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.8.54.2 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.8.54.1 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.202 | 04-Nov-2022 |
ozaki-r | ipcb: add/update the description of functions
From rmind-smpnet patches
|
| 1.201 | 04-Nov-2022 |
ozaki-r | inpcb: replace leading white spaces with tabs
|
| 1.200 | 04-Nov-2022 |
ozaki-r | inpcb: get rid of parentheses for return value
|
| 1.199 | 04-Nov-2022 |
ozaki-r | inpcb: use NULL
|
| 1.198 | 04-Nov-2022 |
ozaki-r | inpcb: use in_port_t for port numbers
|
| 1.197 | 04-Nov-2022 |
ozaki-r | inpcb: use pool_cache instead of pool
|
| 1.196 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to in6pcb_*
|
| 1.195 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.194 | 29-Oct-2022 |
ozaki-r | inpcb: fix for kernels without INET6
|
| 1.193 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.192 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.191 | 14-Oct-2022 |
ryo | Avoid error of "-Wreturn-local-addr", and simplify the logic.
However, -Wreturn-local-addr is still disabled by default by GCC_NO_RETURN_LOCAL_ADDR
in bsd.own.mk because it causes errors in other parts.
|
| 1.190 | 29-Aug-2022 |
knakahara | Add sysctl entry to control to send routing message for RTM_DYNAMIC.
Some routing daemons require such routing message to keep coherency.
If we want to let kernel send such message, set net.inet.icmp.dynamic_rt_msg=1
for IPv4, net.inet6.icmp6.dynamic_rt_msg=1 for IPv6.
Default(=0) is the same as before, that is, not send such routing message.
|
| 1.189 | 29-Jul-2022 |
knakahara | Remove obsoleted comments.
These comments are added with IFNET_LOCK by in_pcb.c:r1.180 and
in6_pcb.c:r1.162. And then, IFNET_LOCK codes are removed in
in_pcb.c:r1.183 and in6_pcb.c:r1.166, however the comments have
remained.
|
| 1.188 | 10-Jun-2022 |
knakahara | Use LIST_FOREACH macro.
|
| 1.187 | 09-Jun-2022 |
knakahara | refactor: use TAILQ_FOREACH instead of TAILQ_FOREACH_SAFE about inpt_queue.
They don't use "ninph" pointer and don't remove elements.
|
| 1.186 | 19-Oct-2021 |
roy | netinet: Allow binding the unspecified address when no addresses exist
You should always be able to bind to the unspecified address even if
no addresses have been configured on any interface.
For example, a DHCP client could be started before the loopback interface
has been fully configured.
|
| 1.185 | 08-Sep-2020 |
christos | Add IP_BINDANY, IPV6_BINDANY which can be used to bind to any address in
order to implement transparent proxies.
|
| 1.184 | 20-Aug-2020 |
riastradh | [ozaki-r] Changes to the kernel core for wireguard
|
| 1.183 | 15-May-2019 |
ozaki-r | Get rid of IFNET_LOCK for if_mcast_op to avoid a deadlock
The IFNET_LOCK was added to avoid data races on if_flags for IFF_ALLMULTI.
Unfortunatetly it caused a deadlock instead. A known scenario causing a
deadlock is to occur the following two operations concurrently: (a) a removal of
an IP adddres assigned to an interface and (b) a manipulation of multicast
groups to the interface. The resource dependency graph is like this:
softnet_lock => IFNET_LOCK => psref_target_destroy => softint => softnet_lock
Thanks to the previous commit that avoids data races on if_flags for
IFF_ALLMULTI by another approach, we can remove IFNET_LOCK and defuse the
deadlock.
PR kern/54189
|
| 1.182 | 27-Feb-2018 |
maxv | branches: 1.182.4;
Dedup: merge
ipsec4_get_policy and ipsec6_get_policy
ipsec4_delete_pcbpolicy and ipsec6_delete_pcbpolicy
The already-existing ipsec_get_policy() function is inlined in the new
one.
|
| 1.181 | 01-Jan-2018 |
christos | 1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
|
| 1.180 | 15-Dec-2017 |
ozaki-r | Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
|
| 1.179 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.178 | 25-Apr-2017 |
ozaki-r | branches: 1.178.4;
Check if solock of PCB is held when SP caches in the PCB are accessed
To this end, a back pointer from inpcbpolicy to inpcb_hdr is added.
|
| 1.177 | 20-Apr-2017 |
ozaki-r | Simplify logic of udp4_sendup and udp6_sendup
They are always passed a socket with the same protocol faimiliy
as its own: AF_INET for udp4_sendup and AF_INET6 for udp6_sendup.
|
| 1.176 | 02-Mar-2017 |
ozaki-r | Make sure imo_membership is protected by inp's lock (solock)
|
| 1.175 | 13-Feb-2017 |
ozaki-r | Replace splnet with splsoftnet
|
| 1.174 | 23-Jan-2017 |
ozaki-r | Get rid of splnet for pool(9)
We don't need it anymore.
|
| 1.173 | 11-Jan-2017 |
ozaki-r | branches: 1.173.2;
Get rid of unnecessary header inclusions
|
| 1.172 | 12-Dec-2016 |
ozaki-r | Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.171 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.170 | 29-Sep-2016 |
roy | Now that we disallow sending or receiving from invalid addresses,
allow binding to tentative addresses.
|
| 1.169 | 26-Aug-2016 |
roy | Allow bind to detached INET addresses.
|
| 1.168 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.167 | 20-Jul-2016 |
ozaki-r | Reduce scopes of variables
|
| 1.166 | 08-Jul-2016 |
ozaki-r | branches: 1.166.2;
Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.165 | 06-Jul-2016 |
ozaki-r | Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.164 | 21-Jun-2016 |
ozaki-r | Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.163 | 15-Feb-2016 |
rtr | Reduce code duplication.
Split creation of IPv4-Mapped IPv6 addresses into its own function
and use it.
No functional change intended. As posted to tech-net@
|
| 1.162 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.161 | 24-May-2015 |
rtr | remove transitional functions in{,6}_pcbconnect_m() that were used in
converting protocol user requests to accept sockaddr instead of mbufs.
remove tcp_input copy in to mbuf from sockaddr and just copy to sockaddr
to make it possible for the transitional functions to go away.
no version bump since these functions only existed for a short time and
were commented as adapters (they appeared in 7.99.15).
|
| 1.160 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.159 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.158 | 26-Apr-2015 |
rtr | return EINVAL if sin{,6}_len != sizeof(sockaddr_in{,6}) respectively in
in{,6}_pcbconnect().
checking just m->m_len isn't enough because there are various places that
assume sa_len has been properly populated.
|
| 1.157 | 24-Apr-2015 |
rtr | make accept, getsockname and getpeername syscalls use sockaddr_big and modify
pr_{accept,sockname,peername} nam parameter type from mbuf * to sockaddr *.
* retained use of mbuftypes[MT_SONAME] for now.
* bump to netbsd version 7.99.12 for parameter type change.
patch posted to tech-net@ 2015/04/19
|
| 1.156 | 03-Apr-2015 |
rtr | * change pr_bind to accept struct sockaddr * instead of struct mbuf *
* update protocol bind implementations to use/expect sockaddr *
instead of mbuf *
* introduce sockaddr_big struct for storage of addr data passed via
sys_bind; sockaddr_big is of sufficient size and alignment to
accommodate all addr data sizes received.
* modify sys_bind to allocate sockaddr_big instead of using an mbuf.
* bump kernel version to 7.99.9 for change to pr_bind() parameter type.
Patch posted to tech-net@
http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html
The choice to use a new structure sockaddr_big has been retained since
changing sockaddr_storage size would lead to unnecessary ABI change. The
use of the new structure does not preclude future work that increases
the size of sockaddr_storage and at that time sockaddr_big may be
trivially replaced.
Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
|
| 1.155 | 25-Nov-2014 |
seanb | branches: 1.155.2;
Really make SO_REUSEPORT and SO_REUSEADDR equivalent for multicast
sockets. From FreeBSD.
|
| 1.154 | 25-Nov-2014 |
seanb | Clean up any dangling ifp references in (struct in6pcb *)->in6p_v4moptions
(v4 multicast options off v4 mapped v6 socket) on interface destruction. The
code to clean this up in a true v4 socket was moved to its own function
which is now also called in the corresponding place for v6 sockets on
interface destruction.
|
| 1.153 | 10-Nov-2014 |
maxv | Do not uselessly include <sys/malloc.h>.
|
| 1.152 | 07-Sep-2014 |
rmind | in_pcbdetach: move ip_freemoptions() under softnet_lock for now (this will
be changed back once other IP paths become MP-safe). Same for IPv6 routine.
This partially reverts 1.150 of in_pcb.c and 1.127 of in6_pcb.c changes.
|
| 1.151 | 05-Aug-2014 |
rtr | branches: 1.151.2;
revert the removal of struct lwp * parameter from bind, listen and connect
user requests.
this should resolve the issue relating to nfs client hangs presented
recently by wiz on current-users@
|
| 1.150 | 03-Aug-2014 |
rmind | in_pcbdetach: not that IGMP and multicast groups are MP-safe, we can move
the ip_freemoptions() call outside the softnet_lock. Should fix PR/49065.
|
| 1.149 | 24-Jul-2014 |
rtr | split PRU_BIND and PRU_LISTEN function out of pr_generic() usrreq
switches and put into separate functions
xxx_bind(struct socket *, struct mbuf *)
xxx_listen(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_BIND with call to
pr_bind()
- replace calls to pr_generic() with req = PRU_LISTEN with call to
pr_listen()
- drop struct lwp * parameter from at_pcbsetaddr(), in_pcbbind() and
unp_bind() and always use curlwp.
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_{bind,listen}() PRUs
- l2cap_{bind,listen}() -> l2cap_{bind,listen}_pcb()
- sco_{bind,listen}() -> sco_{bind,listen}_pcb()
- rfcomm_{bind,listen}() -> rfcomm_{bind,listen}_pcb()
patch reviewed by rmind
welcome to netbsd 6.99.48
|
| 1.148 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.147 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.146 | 23-Nov-2013 |
christos | branches: 1.146.2;
convert from CIRCLEQ to TAILQ.
|
| 1.145 | 05-Jun-2013 |
christos | branches: 1.145.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.144 | 12-Apr-2013 |
christos | PR/47738: connect(2) to 239.x.y.z should return error but does not.
|
| 1.143 | 25-Jun-2012 |
christos | branches: 1.143.2;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.142 | 21-Jun-2012 |
yamt | constify, comments.
no functional changes.
|
| 1.141 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.140 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.139 | 24-Sep-2011 |
christos | branches: 1.139.2; 1.139.6;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.138 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.137 | 12-May-2009 |
elad | branches: 1.137.4; 1.137.6;
Implicit EPERM -> explicit EACCES.
Requested by ad@ and yamt@.
|
| 1.136 | 09-May-2009 |
elad | Add check for IN_MULTICAST() that was taken only to in_pcbbind_port() --
it's necessary in in_pcbbind_addr() as well.
Pointed out by Mihai Chelaru on tech-net@, thanks!
|
| 1.135 | 30-Apr-2009 |
elad | Commit changes to netinet6/in6_src.c, forgot in previous commit:
http://mail-index.netbsd.org/source-changes/2009/04/30/msg220547.html
Make in_pcbsetport() set the port number selected before passing "sin" to
kauth(9).
|
| 1.134 | 30-Apr-2009 |
elad | - Make in6_pcbbind_{addr,port}() static
- Properly authorize port binding in in_pcbsetport() and in6_pcbsetport()
- Pass struct sockaddr_in6 to in6_pcbsetport() instead of just the address,
so that we have a more complete context
- Adjust udp6_output() to craft a sockaddr_in6 as it calls in6_pcbsetport()
- Fix an issue in in_pcbbind() where we used the "dom_sa_any" pointer and
not a copy of it, pointed out by bouyer@, thanks!
Mailing list reference:
http://mail-index.netbsd.org/tech-net/2009/04/29/msg001259.html
|
| 1.133 | 23-Apr-2009 |
elad | - Make kauth(9) call logic match the one in netinet6/in6_pcb.c
- Indent a comment
|
| 1.132 | 23-Apr-2009 |
elad | Some changes to in_pcbbind():
- Extract guts to in_pcbbind_{addr,port}()
- Put the port auto-assignment logic in in_pcbsetport(), which looks very
similar to in6_pcbsetport()
- Fix a bug where "sin" was passed to kauth(9) without being set to
anything
No objections on tech-net@.
|
| 1.131 | 14-Apr-2009 |
elad | Don't set sin->sin_port and sin6->sin6_port to 0 before calling
ifa_ifwithaddr(), as we no longer do a byte compare on the entire struct.
Reviewed by and okay from dyoung@.
|
| 1.130 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.129 | 11-Oct-2008 |
pooka | branches: 1.129.2; 1.129.4; 1.129.8; 1.129.10;
Move uidinfo to its own module in kern_uidinfo.c and include in rump.
No functional change to uidinfo.
|
| 1.128 | 03-Oct-2008 |
pooka | Hallo, pool_init(). Auf wiedersehen & byebye, link set POOL_INIT().
|
| 1.127 | 04-Aug-2008 |
spz | typo fix in comment (drops the ' in drop's :)
|
| 1.126 | 04-Aug-2008 |
matt | Free the socket only after disposing of the PCB.
|
| 1.125 | 05-May-2008 |
ad | branches: 1.125.2; 1.125.6;
- Convert hashinit() to use kmem_alloc(). The hash tables can be large
and it's better to not have them in kmem_map.
- Convert a couple of minor items along the way to kmem_alloc().
- Fix some memory leaks.
|
| 1.124 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.123 | 24-Apr-2008 |
ad | branches: 1.123.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.122 | 14-Jan-2008 |
dyoung | branches: 1.122.6; 1.122.8;
Use rtcache_validate() instead of rtcache_getrt(). Shorten staircase
in in_losing().
|
| 1.121 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.120 | 16-Dec-2007 |
elad | Really fix low port allocation, by always passing a valid lwp to
in_pcbbind().
Okay dyoung@.
Note that the network code is another candidate for major cleanup... also
note that this issue is likely to be present in netinet6 code, too.
|
| 1.119 | 21-Aug-2007 |
dyoung | branches: 1.119.2; 1.119.8; 1.119.10; 1.119.14;
Use sockaddr_in_init().
|
| 1.118 | 19-Jul-2007 |
dyoung | branches: 1.118.4; 1.118.6;
Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.117 | 02-May-2007 |
dyoung | branches: 1.117.2;
Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.116 | 12-Mar-2007 |
ad | branches: 1.116.2;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.115 | 04-Mar-2007 |
christos | branches: 1.115.2;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.114 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.113 | 26-Jan-2007 |
dyoung | branches: 1.113.2;
KNF: bzero -> memset, change (struct in_ifaddr *)0 to NULL.
|
| 1.112 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.111 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.110 | 08-Dec-2006 |
joerg | When a dynamic route is deleted in in_losing and in6_losing, rtrequest
is called, but the current reference via the PCB is not removed. This
is effectively a leaked reference. Call rtfree unconditional.
|
| 1.109 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.108 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.107 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.106 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.105 | 19-Sep-2006 |
elad | Remove ugly (void *) casts from network scope authorization wrapper and
calls to it.
While here, adapt code for system scope listeners to avoid some more
casts (forgotten in previous run).
Update documentation.
|
| 1.104 | 08-Sep-2006 |
elad | branches: 1.104.2;
First take at security model abstraction.
- Add a few scopes to the kernel: system, network, and machdep.
- Add a few more actions/sub-actions (requests), and start using them as
opposed to the KAUTH_GENERIC_ISSUSER place-holders.
- Introduce a basic set of listeners that implement our "traditional"
security model, called "bsd44". This is the default (and only) model we
have at the moment.
- Update all relevant documentation.
- Add some code and docs to help folks who want to actually use this stuff:
* There's a sample overlay model, sitting on-top of "bsd44", for
fast experimenting with tweaking just a subset of an existing model.
This is pretty cool because it's *really* straightforward to do stuff
you had to use ugly hacks for until now...
* And of course, documentation describing how to do the above for quick
reference, including code samples.
All of these changes were tested for regressions using a Python-based
testsuite that will be (I hope) available soon via pkgsrc. Information
about the tests, and how to write new ones, can be found on:
http://kauth.linbsd.org/kauthwiki
NOTE FOR DEVELOPERS: *PLEASE* don't add any code that does any of the
following:
- Uses a KAUTH_GENERIC_ISSUSER kauth(9) request,
- Checks 'securelevel' directly,
- Checks a uid/gid directly.
(or if you feel you have to, contact me first)
This is still work in progress; It's far from being done, but now it'll
be a lot easier.
Relevant mailing list threads:
http://mail-index.netbsd.org/tech-security/2006/01/25/0011.html
http://mail-index.netbsd.org/tech-security/2006/03/24/0001.html
http://mail-index.netbsd.org/tech-security/2006/04/18/0000.html
http://mail-index.netbsd.org/tech-security/2006/05/15/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/01/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/25/0000.html
Many thanks to YAMAMOTO Takashi, Matt Thomas, and Christos Zoulas for help
stablizing kauth(9).
Full credit for the regression tests, making sure these changes didn't break
anything, goes to Matt Fleming and Jaime Fournier.
Happy birthday Randi! :)
|
| 1.103 | 23-Jul-2006 |
ad | branches: 1.103.4;
Use the LWP cached credentials where sane.
|
| 1.102 | 14-May-2006 |
elad | integrate kauth.
|
| 1.101 | 15-Nov-2005 |
dsl | branches: 1.101.4; 1.101.6; 1.101.8; 1.101.10; 1.101.12;
Pass the current process structure to in_pcbconnect() so that it can
pass it to in_pcbbind() so that can allocate a low numbered port
if setsockopt() has been used to set IP_PORTRANGE to IP_PORTRANGE_LOW.
While there, fail in_pcbconnect() if the in_pcbbind() fails - rather
than sending the request out from a port of zero.
This has been largely broken since the socket option was added in 1998.
|
| 1.100 | 29-May-2005 |
christos | branches: 1.100.2; 1.100.8;
- add const
- remove bogus casts
- avoid nested variables
|
| 1.99 | 07-May-2005 |
christos | PR/30154: YAMAMOTO Takashi: tcp_close locking botch
chgsbsize() as mentioned in the PR can be called from an interrupt context
via tcp_close(). Avoid calling uid_find() in chgsbsize().
- Instead of storing so_uid in struct socketvar, store *so_uidinfo
- Add a simple lock to struct uidinfo.
|
| 1.98 | 03-Feb-2005 |
perry | ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.97 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.96 | 29-Sep-2004 |
christos | branches: 1.96.4; 1.96.6;
PR/27082: Sean Boudreau: redundant assignment or NULL dereference in
in_pcbconnect()
|
| 1.95 | 25-Apr-2004 |
simonb | Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.94 | 02-Mar-2004 |
thorpej | Call ipsec_pcbconn() and ipsec_pcbdisconn() for FAST_IPSEC, too.
|
| 1.93 | 13-Jan-2004 |
itojun | avoid deref-after-free.
http://sources.zabbadoz.net/freebsd/patchset/106-ipsec-pcb-discon.diff
|
| 1.92 | 02-Jan-2004 |
itojun | whitespace
|
| 1.91 | 11-Nov-2003 |
jonathan | Change global head-of-local-IP-address list from in_ifaddr to
in_ifaddrhead. Recent changes in struct names caused a namespace
collision in fast-ipsec, which are most cleanly fixed by using
"in_ifaddrhead" as the listhead name.
|
| 1.90 | 28-Oct-2003 |
provos | use a hash table to bind to local ports; suggested by markus friedl
approved: fvdl@
|
| 1.89 | 23-Oct-2003 |
mycroft | Remove all the code to maintain ia_inpcbs. This information was only used to
close sockets on address changes, which was deemed to be a bad idea and was
summarily removed, so there is no point in wasting effort on maintaining it
any more.
|
| 1.88 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.87 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.86 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.85 | 22-Jul-2003 |
itojun | avoid code dup when check broadcast addr in bind(2)
|
| 1.84 | 21-Jul-2003 |
itojun | permit bind(2) to broadcast address, as it was permitted before.
(for instance, "ntpd -b" was broken since revision 1.82)
found report on http://pc.2ch.net/unix
|
| 1.83 | 26-Jun-2003 |
itojun | branches: 1.83.2;
check if INADDR_TO_IA gets us valid in_ifaddr or not. hopefully fix PR21964
|
| 1.82 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.81 | 16-Mar-2003 |
lukem | Enable check in in_pcbbind() to enforce sin_family == AF_INET.
If there are any "old programs which incorrectly set this" left,
they will now fail with EAFNOSUPPORT.
This make in_pcbbind() consistent with in_pcbconnect() and the other
protocol families.
As per my PR [kern/4441], which has the comment:
Steven's "TCP/IP Illustrated, Volume 2", page 730, notes that
in_pcbbind() has the check which determines if sin_family == AF_INET
commented out, but the same check in in_pcbconnect() is still active.
|
| 1.80 | 22-Oct-2002 |
simonb | "error" in in_pcbbind() was only ever set but not used, remove it.
|
| 1.79 | 11-Jun-2002 |
itojun | share policy-on-pcb for listening socket. sync w/kame
todo: share even more, avoid frequent updates of spidx
|
| 1.78 | 09-Jun-2002 |
itojun | whitespace
|
| 1.77 | 28-May-2002 |
itojun | correct in*_pcbrtentry. check cached value correctly.
|
| 1.76 | 28-May-2002 |
itojun | in in*_pcbrtentry(), check if route is still valid (RTF_UP),
and address family is still valid.
|
| 1.75 | 08-Mar-2002 |
thorpej | branches: 1.75.6;
Pool deals fairly well with physical memory shortage, but it doesn't
deal with shortages of the VM maps where the backing pages are mapped
(usually kmem_map). Try to deal with this:
* Group all information about the backend allocator for a pool in a
separate structure. The pool references this structure, rather than
the individual fields.
* Change the pool_init() API accordingly, and adjust all callers.
* Link all pools using the same backend allocator on a list.
* The backend allocator is responsible for waiting for physical memory
to become available, but will still fail if it cannot callocate KVA
space for the pages. If this happens, carefully drain all pools using
the same backend allocator, so that some KVA space can be freed.
* Change pool_reclaim() to indicate if it actually succeeded in freeing
some pages, and use that information to make draining easier and more
efficient.
* Get rid of PR_URGENT. There was only one use of it, and it could be
dealt with by the caller.
From art@openbsd.org.
|
| 1.74 | 22-Jan-2002 |
itojun | make sure to check address family on route cache. with IPv4 mapped
address we can see both AF_INET/INET6.
|
| 1.73 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.72 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.71 | 06-Aug-2001 |
itojun | branches: 1.71.4;
cache IPsec policy on in6?pcb. most of the lookup operations can be bypassed,
especially when it is a connected SOCK_STREAM in6?pcb. sync with kame.
|
| 1.70 | 25-Jul-2001 |
itojun | allocate ipsec policy buffer attached to pcb in in*_pcballoc, before
giving anyone accesses to pcb (do not reveal an inconsistent ones).
sync with kame
|
| 1.69 | 02-Jul-2001 |
itojun | branches: 1.69.2;
on interface removal, remove multicast groups joined from pcb, before
removing interface addresses. without the change, we may deref
NULL pointer in in_pcbpurgeif(). from jinmei@kame, sync with kame
|
| 1.68 | 08-Nov-2000 |
ad | branches: 1.68.2;
Update for hashinit() change.
|
| 1.67 | 25-Aug-2000 |
tron | Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
|
| 1.66 | 06-Jul-2000 |
itojun | remove unnecessary #include <netkey/key_debug.h>. from kame.
|
| 1.65 | 03-Apr-2000 |
enami | branches: 1.65.4;
- Unselect the multicast outgoing interface if it is being detached.
- Drop the multicast membership if we are joining through the interface
being detached.
|
| 1.64 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.63 | 02-Feb-2000 |
thorpej | PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.62 | 01-Feb-2000 |
thorpej | Small amount of cosmetic cleanup.
|
| 1.61 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.60 | 09-Jul-1999 |
thorpej | branches: 1.60.2; 1.60.8;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.59 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.58 | 23-Mar-1999 |
lukem | branches: 1.58.4; 1.58.6;
Ensure that you can only bind a more specific address when it is done by the
same uid or by root.
This code is from FreeBSD. (Whilst it was originally obtained from OpenBSD,
FreeBSD fixed it to work with multicast. To quote the commit message:
- Don't bother checking for conflicting sockets if we're binding to a
multicast address.
- Don't return an error if we're binding to INADDR_ANY, the conflicting
socket is bound to INADDR_ANY, and the conflicting socket has
SO_REUSEPORT set.
)
|
| 1.57 | 19-Dec-1998 |
thorpej | Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.56 | 16-Nov-1998 |
lukem | branches: 1.56.2;
if INADDR_ANY is given in in_pcbconnect(), choose the ia_addr of the first
interface, not the ia_broadaddr. should fix [standards/5645] and [kern/6425]
|
| 1.55 | 13-Nov-1998 |
lukem | simplify test in in_pcbbind() for setting wild=1; no need to check if
((so->so_proto->pr_flags & PR_CONNREQUIRED) == 0 ||
(so->so_options & SO_ACCEPTCONN) == 0)
since the latter is always true, so the former test in unnecessary.
from `TCP/IP Illustrated, Volume 2', W. Richard Stevens, p 730.
|
| 1.54 | 05-Oct-1998 |
lukem | * in_pcblookup_port(): deprecate INPLOOKUP_WILDCARD and flags in favour
of a lookup_wildcard arg; simplifies the logic a bit.
* when assigning ephemeral ports in in_pcbbind(), always call
in_pcblookup_port() with lookup_wildcard=1, so that ephemeral port
allocation on sockets with SO_REUSEADDR set won't potentially bind to a
port in use by something else (principle of least surprise).
|
| 1.53 | 30-Sep-1998 |
tls | Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.52 | 02-Aug-1998 |
thorpej | Use the pool allocator for inpcbs.
|
| 1.51 | 23-Jul-1998 |
pk | in_pcballoc(): we can't afford to wait for memory.
|
| 1.50 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.49 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.48 | 07-Feb-1998 |
chs | add flags arg to hashinit(), to pass to malloc().
|
| 1.47 | 08-Jan-1998 |
lukem | * start from the top of the given ephemeral range and work down;
results in reserved ephemeral ports starting at the top (as per
current practice), and shouldn't have a negative effect on normal
ephemeral ports...
* initialise inpt_lastlow in in_pcbinit
|
| 1.46 | 08-Jan-1998 |
lukem | add missing ; ...
|
| 1.45 | 07-Jan-1998 |
lukem | add the following, derived from FreeBSD:
* IP_PORTRANGE socket option, which controls how the ephemeral ports
are allocated. it takes the following settings:
IP_PORTRANGE_DEFAULT use anonportmin (49152) -> anonportmax (65535)
IP_PORTRANGE_HIGH as IP_PORTRANGE_DEFAULT (retained for FreeBSD
compat reasons, where these are separate)
IP_PORTRANGE_LOW use 600 -> 1023. only works if uid==0.
* in_pcb flag INP_ANONPORT. set if port was allocated ephmerally
|
| 1.44 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.43 | 05-Jan-1998 |
lukem | enhance ephemeral port allocation code:
* support sysctl net.inet.ip.anonportmin (lowest ephemeral port)
and net.inet.ip.anonportmax (highest ephemeral port).
these can't be set to >65535, < IPPORT_RESERVED (unless IPNOPRIVPORTS
is defined), and anonportmin has to be < anonportmax.
* use a cleaner way of only cycling through the available set once;
this will be useful for when a random allocation scheme is used
* define IPPORT_ANON{MIN,MAX} instead of IPPORT_USER{LOW,HIGH}
|
| 1.42 | 30-Dec-1997 |
lukem | as per the IANA assigned ports numbers document, use ports
49152..65535 for ephemeral ports (instead of 1024..5000).
closes my [kern/4440], but with correct code :)
|
| 1.41 | 27-Nov-1997 |
mrg | fix compile error when "options IPNOPROVPORTS"
|
| 1.40 | 20-Nov-1997 |
thorpej | Deal with a problem where ephemeral port shortage would case a PCB's
local address to be set, causing all further attemps to bind that PCB
to fail. From Koji Imada, PR #3857.
|
| 1.39 | 14-Oct-1997 |
matt | branches: 1.39.2;
Add support for returning maximum supported MTU when ip_output fails with
EMSGSIZE.
|
| 1.38 | 22-Sep-1997 |
thorpej | Implement in_pcbrtentry() - return the route associated with a PCB. If
one does not exist, attempt to allocate one. This is mostly pulled from
tcp_input.c.
|
| 1.37 | 23-Jul-1997 |
thorpej | branches: 1.37.2;
Pull SYN_cache_branch down into the main line.
|
| 1.36 | 10-Dec-1996 |
mycroft | branches: 1.36.8;
Return EAGAIN if binding with no specified port and the pool is empty.
|
| 1.35 | 13-Oct-1996 |
christos | backout previous kprintf changes
|
| 1.34 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.33 | 15-Sep-1996 |
mycroft | Hash unconnected PCBs.
|
| 1.32 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.31 | 05-Sep-1996 |
perry | Commit PR 2671, which adds an "IPNOPRIVPORTS" config option that turns
off the code that normally only allows root to bind low TCP
ports. Useful on firewalls and such.
|
| 1.30 | 14-Aug-1996 |
thorpej | Fix some DIAGNOSTIC printf() formats; ntohl() provides a 32-bit quantity,
and should be printed with %x, not %lx.
|
| 1.29 | 10-Jul-1996 |
cgd | print result of ntohl/htonl as a long. (makes -Wformat work on the
Alpha.)
|
| 1.28 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.27 | 26-Feb-1996 |
mrg | branches: 1.27.4;
two more local addr changes, all done differently now (idea from charles)
|
| 1.26 | 26-Feb-1996 |
mrg | if we are connecting *to* an address of any local interface, default the
local address of the socket to the same address.
|
| 1.25 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.24 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.23 | 17-Aug-1995 |
mycroft | branches: 1.23.2;
so_pcb should be a void *.
|
| 1.22 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.21 | 18-Jun-1995 |
cgd | convert pcb lists to CIRCLEQs, so that the end can be looked at more
easily, and so that the original (insque/remque) logic can be effectively
mimiced. (This fixes a bug in the previous set of list changes.)
also (since terminator is no longer null) reinstate uninitted list checks,
but mark them XXX.
|
| 1.20 | 12-Jun-1995 |
mycroft | in_pcbnotify*() don't return anything.
|
| 1.19 | 12-Jun-1995 |
mycroft | Change in_pcbnotify*() to take an errno value. Make inetctlerrmap[] an
array on ints, not u_chars.
|
| 1.18 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.17 | 04-Jun-1995 |
mycroft | Remove one more bogus cast.
|
| 1.16 | 04-Jun-1995 |
mycroft | Don't cast things unnecessarily.
|
| 1.15 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.14 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.13 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.12 | 29-Sep-1994 |
deraadt | failure to bind to a reserved port should return EACCES not EPERM.
|
| 1.11 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.10 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.9 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.8 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.7 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.6 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.5 | 11-Jun-1993 |
deraadt | The latest patch was hosed. There is some program that I used which
left extra crud at the end of the file. I blame ftpd for not doing an
ftruncate().
|
| 1.4 | 10-Jun-1993 |
deraadt | patch from Yuval Yarom, sent to me by <andrew@werple.apana.org.au>
they say: When doing an implicit bind in_pcbbind will assign used ports
if the port is bound on specific interface, and not on INADDR_ANY.
Effects of the bug range from connection drops to machine hangs.
|
| 1.3 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.23.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.27.4.2 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.27.4.1 | 10-Dec-1996 |
mycroft | From trunk:
Return EAGAIN if binding with no specified port and the pool is empty.
|
| 1.36.8.1 | 14-May-1997 |
mellon | in_pcbnotify() returns a result indicating whether or not it actually matched a pcb.
|
| 1.37.2.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.37.2.1 | 29-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.39.2.4 | 18-Jan-1999 |
cgd | pull up rev 1.56 from trunk (PR#5645 and PR#6425). (lukem)
|
| 1.39.2.3 | 01-Oct-1998 |
cgd | pull up revisions 1.49-1.50, 1.53 (via patch) from trunk. (tls)
|
| 1.39.2.2 | 28-Nov-1997 |
mellon | Pull rev 1.41 up from trunk (mrg)
|
| 1.39.2.1 | 20-Nov-1997 |
thorpej | Fix PCB binding problem caused by ephemeral port shortage.
|
| 1.56.2.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.58.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.58.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.58.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.58.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.58.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.60.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.60.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.60.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.65.4.1 | 26-Aug-2000 |
tron | Pull up from current (approved by thorpej):
Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
syssrc/sys/netinet/in.h 1.49 -> 1.50
syssrc/sys/netinet/in_pcb.c 1.66 -> 1.67
syssrc/sys/netinet/ip_input.c 1.116 -> 1.117
syssrc/sys/netinet/ip_var.h 1.41 -> 1.42
|
| 1.68.2.6 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.68.2.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.68.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.68.2.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.68.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.68.2.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.69.2.6 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.69.2.5 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.69.2.4 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.69.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.69.2.2 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.69.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.71.4.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.75.6.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.75.6.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.83.2.7 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.83.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.83.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.83.2.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.83.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.83.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.83.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.96.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.96.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.100.8.1 | 22-Nov-2005 |
yamt | sync with head.
|
| 1.100.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.100.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.100.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.100.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.100.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.101.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.101.10.3 | 06-May-2006 |
christos | - Move kauth_cred_t declaration to <sys/types.h>
- Cleanup struct ucred; forward declarations that are unused.
- Don't include <sys/kauth.h> in any header, but include it in the c files
that need it.
Approved by core.
|
| 1.101.10.2 | 10-Mar-2006 |
elad | generic_authorize() -> kauth_authorize_generic().
|
| 1.101.10.1 | 08-Mar-2006 |
elad | Adapt to kernel authorization KPI.
|
| 1.101.8.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.101.8.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.101.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.101.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.101.4.4 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.101.4.3 | 14-Mar-2006 |
rpaulo | Remove last reference to in6pcb.
|
| 1.101.4.2 | 14-Feb-2006 |
rpaulo | Remove INPCBHASH_PORT, INPCBHASH_BIND, INPCBHASH_CONNECT (moved to
in_pcb.h).
If INET6, detect whether the pcb is v4 or v6 based on the socket
family (from FreeBSD).
|
| 1.101.4.1 | 02-Feb-2006 |
rpaulo | Remove #include netinet6/in6_pcb.h.
|
| 1.103.4.3 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.103.4.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.103.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.104.2.3 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.104.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.104.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.113.2.4 | 07-May-2007 |
yamt | sync with head.
|
| 1.113.2.3 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.113.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.113.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.115.2.4 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.115.2.3 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.115.2.2 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.115.2.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.116.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.117.2.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.117.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.118.6.2 | 19-Jul-2007 |
dyoung | Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.118.6.1 | 19-Jul-2007 |
dyoung | file in_pcb.c was added on branch matt-mips64 on 2007-07-19 20:48:55 +0000
|
| 1.118.4.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.119.14.2 | 19-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.119.14.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.119.10.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.119.8.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.119.2.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.119.2.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.122.8.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.122.6.4 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.122.6.3 | 05-Oct-2008 |
mjf | Sync with HEAD.
|
| 1.122.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.122.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.123.2.3 | 16-May-2009 |
yamt | sync with head
|
| 1.123.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.123.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.125.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.125.2.2 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.125.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.129.10.1 | 10-May-2009 |
snj | branches: 1.129.10.1.2;
Apply patch (requested by sborrill in ticket #745):
Fix compilation with IPNOPRIVPORTS option.
|
| 1.129.10.1.2.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.129.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.129.4.1 | 10-May-2009 |
snj | Apply patch (requested by sborrill in ticket #745):
Fix compilation with IPNOPRIVPORTS option.
|
| 1.129.2.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.137.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.137.4.1 | 31-May-2011 |
rmind | sync with head
|
| 1.139.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.139.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.139.2.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.139.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.139.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.143.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.143.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.143.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.145.2.4 | 18-May-2014 |
rmind | sync with head
|
| 1.145.2.3 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.145.2.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.145.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.146.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.151.2.2 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.151.2.1 | 08-Sep-2014 |
msaitoh | Pull up following revision(s) (requested by rmind in ticket #80):
sys/netinet6/in6_pcb.c: revision 1.129
sys/netinet/in_pcb.c: revision 1.152
in_pcbdetach: move ip_freemoptions() under softnet_lock for now (this will
be changed back once other IP paths become MP-safe). Same for IPv6 routine.
This partially reverts 1.150 of in_pcb.c and 1.127 of in6_pcb.c changes.
|
| 1.155.2.8 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.155.2.7 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.155.2.6 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.155.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.155.2.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.155.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.155.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.155.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.166.2.6 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.166.2.5 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.166.2.4 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.166.2.3 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.166.2.2 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.166.2.1 | 26-Jul-2016 |
pgoyette | Sync with HEAD
|
| 1.173.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.178.4.3 | 18-Mar-2018 |
martin | Pull up following revision(s) (requested by tih in ticket #639):
sys/kern/uipc_socket.c: revision 1.258
sys/kern/uipc_socket.c: revision 1.259
sys/netinet/ip_input.c: revision 1.364 (via patch)
sys/netinet/ip_output.c: revision 1.289
sys/netinet/in.h: revision 1.102
sys/netinet/in_pcb.c: revision 1.181
share/man/man9/sockopt.9: revision 1.11
sys/netinet/in_pcb.h: revision 1.65
sys/sys/socketvar.h: revision 1.146
sys/kern/uipc_syscalls.c: revision 1.189
sys/netinet/ip_output.c: revision 1.290
share/man/man4/ip.4: revision 1.41
share/man/man4/ip.4: revision 1.42
sys/kern/uipc_syscalls.c: revision 1.190
pass valsize for getsockopt like we do for setsockopt
make sure that we have enough space, don't require the exact size
(Tom Ivar Helbekkmo)
1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
new sentence-new line
Remove comment now that the getsockopt code passes the size.
Add a new sockopt member to keep track of the actual size of the option
that should be returned to the caller in getsockopt(2).
(Tom Ivar Helbekkmo)
|
| 1.178.4.2 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.178.4.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.182.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.76 | 04-Nov-2022 |
ozaki-r | inpcb: use in_port_t for port numbers
|
| 1.75 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to in6pcb_*
|
| 1.74 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.73 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.72 | 28-Oct-2022 |
ozaki-r | Remove in_pcb_hdr.h
|
| 1.71 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.70 | 10-Jun-2022 |
knakahara | "inp_hash" is not used now.
|
| 1.69 | 08-Sep-2020 |
christos | Add IP_BINDANY, IPV6_BINDANY which can be used to bind to any address in
order to implement transparent proxies.
|
| 1.68 | 28-Aug-2020 |
riastradh | netinet: Include the needful so include order doesn't matter.
|
| 1.67 | 20-Aug-2020 |
riastradh | [ozaki-r] Changes to the kernel core for wireguard
|
| 1.66 | 31-May-2018 |
maxv | Remove support for non-IKE markers in the kernel. Discussed on tech-net@,
and now in PR/53334. Basically non-IKE markers come from a deprecated
draft, and our kernel code for them has never worked.
Setsockopt will now reject UDP_ENCAP_ESPINUDP_NON_IKE.
Perhaps we should also add a check in key_handle_natt_info(), to make
sure we also reject UDP_ENCAP_ESPINUDP_NON_IKE in the SADB.
|
| 1.65 | 01-Jan-2018 |
christos | branches: 1.65.2;
1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
|
| 1.64 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.63 | 02-Mar-2017 |
ozaki-r | branches: 1.63.6;
Make sure imo_membership is protected by inp's lock (solock)
|
| 1.62 | 22-Feb-2017 |
ozaki-r | Add assertions and comments for lock states of socket and pcb
|
| 1.61 | 08-Dec-2016 |
ozaki-r | branches: 1.61.2;
Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.60 | 26-Apr-2016 |
ozaki-r | branches: 1.60.2;
Sweep unnecessary route.h inclusions
|
| 1.59 | 24-May-2015 |
rtr | remove transitional functions in{,6}_pcbconnect_m() that were used in
converting protocol user requests to accept sockaddr instead of mbufs.
remove tcp_input copy in to mbuf from sockaddr and just copy to sockaddr
to make it possible for the transitional functions to go away.
no version bump since these functions only existed for a short time and
were commented as adapters (they appeared in 7.99.15).
|
| 1.58 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.57 | 24-Apr-2015 |
rtr | make accept, getsockname and getpeername syscalls use sockaddr_big and modify
pr_{accept,sockname,peername} nam parameter type from mbuf * to sockaddr *.
* retained use of mbuftypes[MT_SONAME] for now.
* bump to netbsd version 7.99.12 for parameter type change.
patch posted to tech-net@ 2015/04/19
|
| 1.56 | 03-Apr-2015 |
rtr | * change pr_bind to accept struct sockaddr * instead of struct mbuf *
* update protocol bind implementations to use/expect sockaddr *
instead of mbuf *
* introduce sockaddr_big struct for storage of addr data passed via
sys_bind; sockaddr_big is of sufficient size and alignment to
accommodate all addr data sizes received.
* modify sys_bind to allocate sockaddr_big instead of using an mbuf.
* bump kernel version to 7.99.9 for change to pr_bind() parameter type.
Patch posted to tech-net@
http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html
The choice to use a new structure sockaddr_big has been retained since
changing sockaddr_storage size would lead to unnecessary ABI change. The
use of the new structure does not preclude future work that increases
the size of sockaddr_storage and at that time sockaddr_big may be
trivially replaced.
Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
|
| 1.55 | 25-Nov-2014 |
seanb | branches: 1.55.2;
Clean up any dangling ifp references in (struct in6pcb *)->in6p_v4moptions
(v4 multicast options off v4 mapped v6 socket) on interface destruction. The
code to clean this up in a true v4 socket was moved to its own function
which is now also called in the corresponding place for v6 sockets on
interface destruction.
|
| 1.54 | 05-Aug-2014 |
rtr | revert the removal of struct lwp * parameter from bind, listen and connect
user requests.
this should resolve the issue relating to nfs client hangs presented
recently by wiz on current-users@
|
| 1.53 | 24-Jul-2014 |
rtr | split PRU_BIND and PRU_LISTEN function out of pr_generic() usrreq
switches and put into separate functions
xxx_bind(struct socket *, struct mbuf *)
xxx_listen(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_BIND with call to
pr_bind()
- replace calls to pr_generic() with req = PRU_LISTEN with call to
pr_listen()
- drop struct lwp * parameter from at_pcbsetaddr(), in_pcbbind() and
unp_bind() and always use curlwp.
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_{bind,listen}() PRUs
- l2cap_{bind,listen}() -> l2cap_{bind,listen}_pcb()
- sco_{bind,listen}() -> sco_{bind,listen}_pcb()
- rfcomm_{bind,listen}() -> rfcomm_{bind,listen}_pcb()
patch reviewed by rmind
welcome to netbsd 6.99.48
|
| 1.52 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.51 | 27-Jun-2013 |
christos | branches: 1.51.2; 1.51.6;
implement IP_PKTINFO and IP_RECVPKTINFO.
|
| 1.50 | 25-Jun-2012 |
christos | branches: 1.50.2;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.49 | 24-Sep-2011 |
christos | branches: 1.49.2;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.48 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.47 | 17-Jul-2009 |
minskim | branches: 1.47.4; 1.47.6;
Add the IP_MINTTL socket option.
The IP_MINTTL option may be used on SOCK_STREAM sockets to discard
packets with a TTL lower than the option value. This can be used to
implement the Generalized TTL Security Mechanism (GTSM) according to
RFC 3682.
OK'ed by christos@.
|
| 1.46 | 16-Jul-2009 |
minskim | Add the IP_RECVTTL option support.
If the IP_RECVTTL option is enabled on a SOCK_DGRAM socket, the
recvmsg(2) call will return the TTL of the received datagram. The
msg_control field in the msghdr structure points to a buffer that
contains a cmsghdr structure followed by the TTL value.
Modeled after FreeBSD implementation.
|
| 1.45 | 16-Dec-2007 |
elad | branches: 1.45.10; 1.45.24;
Oops. Remove kauth.h inclusion.
Pointed out by gdt@, thanks.
|
| 1.44 | 16-Dec-2007 |
elad | Really fix low port allocation, by always passing a valid lwp to
in_pcbbind().
Okay dyoung@.
Note that the network code is another candidate for major cleanup... also
note that this issue is likely to be present in netinet6 code, too.
|
| 1.43 | 19-Sep-2007 |
dyoung | branches: 1.43.8;
1) Introduce a new socket option, (SOL_SOCKET, SO_NOHEADER), that
tells a socket that it should both add a protocol header to tx'd
datagrams and remove the header from rx'd datagrams:
int onoff = 1, s = socket(...);
setsockopt(s, SOL_SOCKET, SO_NOHEADER, &onoff);
2) Add an implementation of (SOL_SOCKET, SO_NOHEADER) for raw IPv4
sockets.
3) Reorganize the protocols' pr_ctloutput implementations a bit.
Consistently return ENOPROTOOPT when an option is unsupported,
and EINVAL if a supported option's arguments are incorrect.
Reorganize the flow of code so that it's more clear how/when
options are passed down the stack until they are handled.
Shorten some pr_ctloutput staircases for readability.
4) Extract common mbuf code into subroutines, add new sockaddr
methods, and introduce a new subroutine, fsocreate(), for reuse
later; use it first in sys_socket():
struct mbuf *m_getsombuf(struct socket *so)
Create an mbuf and make its owner the socket `so'.
struct mbuf *m_intopt(struct socket *so, int val)
Create an mbuf, make its owner the socket `so', put the
int `val' into it, and set its length to sizeof(int).
int fsocreate(..., int *fd)
Create a socket, a la socreate(9), put the socket into the
given LWP's descriptor table, return the descriptor at `fd'
on success.
void *sockaddr_addr(struct sockaddr *sa, socklen_t *slenp)
const void *sockaddr_const_addr(const struct sockaddr *sa, socklen_t *slenp)
Extract a pointer to the address part of a sockaddr. Write
the length of the address part at `slenp', if `slenp' is
not NULL.
socklen_t sockaddr_getlen(const struct sockaddr *sa)
Return the length of a sockaddr. This just evaluates to
sa->sa_len. I only add this for consistency with code that
appears in a portable userland library that I am going to
import.
const struct sockaddr *sockaddr_any(const struct sockaddr *sa)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
const void *sockaddr_anyaddr(const struct sockaddr *sa, socklen_t *slenp)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
|
| 1.42 | 23-Jul-2006 |
ad | branches: 1.42.14; 1.42.28; 1.42.30;
Use the LWP cached credentials where sane.
|
| 1.41 | 10-Dec-2005 |
elad | branches: 1.41.4; 1.41.8;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.40 | 15-Nov-2005 |
dsl | Pass the current process structure to in_pcbconnect() so that it can
pass it to in_pcbbind() so that can allocate a low numbered port
if setsockopt() has been used to set IP_PORTRANGE to IP_PORTRANGE_LOW.
While there, fail in_pcbconnect() if the in_pcbbind() fails - rather
than sending the request out from a port of zero.
This has been largely broken since the socket option was added in 1998.
|
| 1.39 | 12-Feb-2005 |
manu | branches: 1.39.6; 1.39.12;
Add support for IPsec Network Address Translator traversal (NAT-T), as
described by RFC 3947 and 3948.
|
| 1.38 | 21-Apr-2004 |
itojun | branches: 1.38.4; 1.38.6;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.37 | 18-Apr-2004 |
matt | De __P()
|
| 1.36 | 23-Oct-2003 |
mycroft | Remove all the code to maintain ia_inpcbs. This information was only used to
close sockets on address changes, which was deemed to be a bad idea and was
summarily removed, so there is no point in wasting effort on maintaining it
any more.
|
| 1.35 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.34 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.33 | 15-Jun-2003 |
matt | branches: 1.33.2;
Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.32 | 02-Nov-2002 |
itojun | cleanup ipsec.h dependency. commented by perry, sync w/kame
|
| 1.31 | 09-Jun-2002 |
itojun | whitespace
|
| 1.30 | 02-Jul-2001 |
itojun | branches: 1.30.2; 1.30.14;
on interface removal, remove multicast groups joined from pcb, before
removing interface addresses. without the change, we may deref
NULL pointer in in_pcbpurgeif(). from jinmei@kame, sync with kame
|
| 1.29 | 02-Feb-2000 |
thorpej | branches: 1.29.6;
PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.28 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.27 | 01-Jul-1999 |
itojun | branches: 1.27.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.26 | 05-Oct-1998 |
lukem | branches: 1.26.8; 1.26.10;
* in_pcblookup_port(): deprecate INPLOOKUP_WILDCARD and flags in favour
of a lookup_wildcard arg; simplifies the logic a bit.
* when assigning ephemeral ports in in_pcbbind(), always call
in_pcblookup_port() with lookup_wildcard=1, so that ephemeral port
allocation on sockets with SO_REUSEADDR set won't potentially bind to a
port in use by something else (principle of least surprise).
|
| 1.25 | 18-May-1998 |
matt | Move the ppcb pointer towards the front of the structure so that it and the
pcb chain pointers can possibly be in the same cache line.
|
| 1.24 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.23 | 07-Jan-1998 |
lukem | add the following, derived from FreeBSD:
* IP_PORTRANGE socket option, which controls how the ephemeral ports
are allocated. it takes the following settings:
IP_PORTRANGE_DEFAULT use anonportmin (49152) -> anonportmax (65535)
IP_PORTRANGE_HIGH as IP_PORTRANGE_DEFAULT (retained for FreeBSD
compat reasons, where these are separate)
IP_PORTRANGE_LOW use 600 -> 1023. only works if uid==0.
* in_pcb flag INP_ANONPORT. set if port was allocated ephmerally
|
| 1.22 | 14-Oct-1997 |
matt | Add support for returning maximum supported MTU when ip_output fails with
EMSGSIZE.
|
| 1.21 | 22-Sep-1997 |
thorpej | Implement in_pcbrtentry() - return the route associated with a PCB. If
one does not exist, attempt to allocate one. This is mostly pulled from
tcp_input.c.
|
| 1.20 | 23-Jul-1997 |
thorpej | branches: 1.20.2;
Pull SYN_cache_branch down into the main line.
|
| 1.19 | 11-Jan-1997 |
thorpej | branches: 1.19.8;
Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.18 | 17-Sep-1996 |
mycroft | Overlay inp_faddr and inp_laddr into the header prototype.
|
| 1.17 | 15-Sep-1996 |
mycroft | Hash unconnected PCBs.
|
| 1.16 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.15 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.14 | 13-Feb-1996 |
christos | branches: 1.14.4;
netinet prototypes
|
| 1.13 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.12 | 18-Jun-1995 |
cgd | branches: 1.12.2;
convert pcb lists to CIRCLEQs, so that the end can be looked at more
easily, and so that the original (insque/remque) logic can be effectively
mimiced. (This fixes a bug in the previous set of list changes.)
also (since terminator is no longer null) reinstate uninitted list checks,
but mark them XXX.
|
| 1.11 | 12-Jun-1995 |
mycroft | in_pcbnotify*() don't return anything.
|
| 1.10 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.9 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.8 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.7 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.6 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 08-Dec-1993 |
hpeyerl | More multicast stuff.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.12.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.14.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.19.8.1 | 14-May-1997 |
mellon | Change return value for in_pcbnotify().
|
| 1.20.2.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.20.2.1 | 29-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.26.10.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.26.10.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.26.8.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.27.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.29.6.3 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.29.6.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.29.6.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.30.14.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.30.2.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.33.2.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.33.2.4 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.33.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.33.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.33.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.38.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.38.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.39.12.1 | 22-Nov-2005 |
yamt | sync with head.
|
| 1.39.6.3 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.39.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.39.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.41.8.1 | 11-Aug-2006 |
yamt | sync with head
|
| 1.41.4.8 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.41.4.7 | 10-Feb-2006 |
rpaulo | * move IN6_HASH() to ifdef INET6
* introduce INP_SOCKAF() from FreeBSD
* add IN{,6}PCBHASH_{CONNECT,PORT,BIND} from in6_pcb.c and in_pcb.c
|
| 1.41.4.6 | 07-Feb-2006 |
rpaulo | Add FreeBSD's locking defines (currently defined to nothing) and a
'struct lock' inside 'struct inpcb'.
|
| 1.41.4.5 | 07-Feb-2006 |
rpaulo | Fix wrong indentation.
|
| 1.41.4.4 | 05-Feb-2006 |
rpaulo | Remove 'inp_options' and 'in6p_options' define.
|
| 1.41.4.3 | 04-Feb-2006 |
rpaulo | struct mbuf can be shared between in4p_depend and in6p_depend.
|
| 1.41.4.2 | 04-Feb-2006 |
rpaulo | Add IN6P_{ATTACHED,BOUND,CONNECTED} for KAME src compatibility.
|
| 1.41.4.1 | 01-Feb-2006 |
rpaulo | Merge in6pcb with inpcb and remove inpcb_hdr since that's no longer needed.
|
| 1.42.30.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.42.28.1 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.42.14.1 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.43.8.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.45.24.1 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.45.10.2 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.45.10.1 | 18-Jul-2009 |
yamt | sync with head.
|
| 1.47.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.47.4.1 | 31-May-2011 |
rmind | sync with head
|
| 1.49.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.49.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.50.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.50.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.51.6.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.51.2.3 | 18-May-2014 |
rmind | sync with head
|
| 1.51.2.2 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.51.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.55.2.5 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.55.2.4 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.55.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.55.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.55.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.60.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.60.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.61.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.63.6.2 | 18-Mar-2018 |
martin | Pull up following revision(s) (requested by tih in ticket #639):
sys/kern/uipc_socket.c: revision 1.258
sys/kern/uipc_socket.c: revision 1.259
sys/netinet/ip_input.c: revision 1.364 (via patch)
sys/netinet/ip_output.c: revision 1.289
sys/netinet/in.h: revision 1.102
sys/netinet/in_pcb.c: revision 1.181
share/man/man9/sockopt.9: revision 1.11
sys/netinet/in_pcb.h: revision 1.65
sys/sys/socketvar.h: revision 1.146
sys/kern/uipc_syscalls.c: revision 1.189
sys/netinet/ip_output.c: revision 1.290
share/man/man4/ip.4: revision 1.41
share/man/man4/ip.4: revision 1.42
sys/kern/uipc_syscalls.c: revision 1.190
pass valsize for getsockopt like we do for setsockopt
make sure that we have enough space, don't require the exact size
(Tom Ivar Helbekkmo)
1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
new sentence-new line
Remove comment now that the getsockopt code passes the size.
Add a new sockopt member to keep track of the actual size of the option
that should be returned to the caller in getsockopt(2).
(Tom Ivar Helbekkmo)
|
| 1.63.6.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.65.2.1 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.16 | 28-Oct-2022 |
ozaki-r | Remove in_pcb_hdr.h
|
| 1.15 | 28-Aug-2020 |
riastradh | netinet: Include the needful so include order doesn't matter.
|
| 1.14 | 20-Aug-2020 |
riastradh | [ozaki-r] Changes to the kernel core for wireguard
|
| 1.13 | 02-Jun-2017 |
ozaki-r | Assert inph_locked on ipsec_pcb_skip_ipsec (was IPSEC_PCB_SKIP_IPSEC)
The assertion confirms SP caches are accessed under inph lock (solock).
|
| 1.12 | 25-Apr-2017 |
ozaki-r | Check if solock of PCB is held when SP caches in the PCB are accessed
To this end, a back pointer from inpcbpolicy to inpcb_hdr is added.
|
| 1.11 | 30-May-2014 |
christos | branches: 1.11.4; 1.11.8;
Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.10 | 23-Nov-2013 |
christos | branches: 1.10.2;
expose the pcb queue structure for convenience
|
| 1.9 | 23-Nov-2013 |
christos | convert from CIRCLEQ to TAILQ.
|
| 1.8 | 25-Jun-2012 |
christos | branches: 1.8.2; 1.8.4;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.7 | 24-Sep-2011 |
christos | branches: 1.7.2;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.6 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.5 | 04-Mar-2007 |
christos | branches: 1.5.64; 1.5.70;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.4 | 10-Dec-2005 |
elad | branches: 1.4.4; 1.4.26;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.3 | 02-Mar-2004 |
thorpej | branches: 1.3.4; 1.3.18;
Define a sotoinpcb_hdr() macro (a'la sotoinpcb()).
|
| 1.2 | 28-Oct-2003 |
provos | use a hash table to bind to local ports; suggested by markus friedl
approved: fvdl@
|
| 1.1 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.3.18.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.3.18.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.3.4.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.3.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.3.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.3.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.3.4.1 | 02-Mar-2004 |
skrll | file in_pcb_hdr.h was added on branch ktrace-lwp on 2004-08-03 10:54:36 +0000
|
| 1.4.26.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.4.4.1 | 01-Feb-2006 |
rpaulo | Merge in6pcb with inpcb and remove inpcb_hdr since that's no longer needed.
|
| 1.5.70.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.5.64.1 | 31-May-2011 |
rmind | sync with head
|
| 1.7.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.7.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.8.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.8.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.8.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.10.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.11.8.1 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.11.4.1 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1 | 02-Dec-2014 |
christos | branches: 1.1.2; 1.1.18;
add routines to print in_addr and sockaddr_in (in_print and sin_print)
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 02-Dec-2014 |
jdolecek | file in_print.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 02-Dec-2014 |
skrll | file in_print.c was added on branch nick-nhusb on 2015-04-06 15:18:23 +0000
|
| 1.131 | 03-Sep-2022 |
thorpej | Garbage-collect everything related to struct domain::dom_ifqueues
(except dom_ifqueues itself, until the next kernel version bump).
It's no longer used now that nothing uses the legacy netisr mechanism.
|
| 1.130 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.129 | 14-Aug-2018 |
maxv | Retire EtherIP, we have L2TP instead.
|
| 1.128 | 03-May-2018 |
maxv | branches: 1.128.2;
Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.127 | 15-Mar-2018 |
maxv | Add the PR_LASTHDR flag on the PFsync and CARP entries. Otherwise a
"require" IPsec policy is not enforced on them, and unauthenticated
packets will be accepted.
Tested with a require-AH configuration. Sent on tech-net@, no comment.
|
| 1.126 | 05-Feb-2018 |
maxv | branches: 1.126.2;
Declare icmperrppslim in ip_icmp.c, it shouldn't be used elsewhere.
|
| 1.125 | 27-Sep-2017 |
ozaki-r | Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
|
| 1.124 | 21-Sep-2017 |
ozaki-r | Invalidate rtcache based on a global generation counter
The change introduces a global generation counter that is incremented when any
routes have been added or deleted. When a rtcache caches a rtentry into itself,
it also stores a snapshot of the generation counter. If the snapshot equals to
the global counter, the cache is still valid, otherwise invalidated.
One drawback of the change is that all rtcaches of all protocol families are
invalidated when any routes of any protocol families are added or deleted.
If that matters, we should have separate generation counters based on
protocol families.
This change removes LIST_ENTRY from struct route, which fixes a part of
PR kern/52515.
|
| 1.123 | 14-Apr-2017 |
ozaki-r | branches: 1.123.4;
Rumpify netipsec
Note that we should modularize netipsec and reduce reverse symbol references
(referencing symbols of netipsec from net, netinet and netinet6) though,
the task needs lots of code changes. Prior to doing so, rumpifying it and
having ATF tests should be useful.
|
| 1.122 | 16-Feb-2017 |
knakahara | add l2tp(4) L2TPv3 interface.
originally implemented by IIJ SEIL team.
|
| 1.121 | 13-Feb-2017 |
ozaki-r | Protect mtudisc and redirect stuffs of icmp/icmp6 with mutex
We have to run pr_init of icmp and icmp6 prior to tcp and tcp6 ones
for mutex initialization.
|
| 1.120 | 26-Apr-2016 |
ozaki-r | branches: 1.120.2; 1.120.4;
Sweep unnecessary route.h inclusions
|
| 1.119 | 11-Apr-2016 |
ozaki-r | Sweep unncessary radix.h inclusions
|
| 1.118 | 21-Jan-2016 |
riastradh | Revert previous: ran cvs commit when I meant cvs diff. Sorry!
Hit up-arrow one too few times.
|
| 1.117 | 21-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.116 | 20-Jan-2016 |
riastradh | Eliminate struct protosw::pr_output.
You can't use this unless you know what it is a priori: the formal
prototype is variadic, and the different instances (e.g., ip_output,
route_output) have different real prototypes.
Convert the only user of it, raw_send in net/raw_cb.c, to take an
explicit callback argument. Convert the only instances of it,
route_output and key_output, to such explicit callbacks for raw_send.
Use assertions to make sure the conversion to explicit callbacks is
warranted.
Discussed on tech-net with no objections:
https://mail-index.netbsd.org/tech-net/2016/01/16/msg005484.html
|
| 1.115 | 13-Oct-2015 |
rjs | Add core networking support for SCTP.
|
| 1.114 | 31-Aug-2015 |
ozaki-r | Replace ARP cache (llinfo) with lltable/llentry
Highlights of the change are:
- Use llentry instead of llinfo to manage ARP caches
- ARP specific data are stored in the hashed list
of an interface instead of the global list (llinfo_arp)
- Fine-grain locking on llentry
- arptimer (callout) per ARP cache
- the global timer callout with the big locks can be
removed (though softnet_lock is still required for now)
- net.inet.arp.prune is now obsoleted
- it was the interval of the global timer callout
- net.inet.arp.refresh is now obsoleted
- it was a parameter that prevents expiration of active caches
- Removed to simplify the timer logic, but we may be able to
restore the feature if really needed
Proposed on tech-kern and tech-net.
|
| 1.113 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.112 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.111 | 10-Feb-2015 |
rjs | Add DCCP protocol support from KAME.
|
| 1.110 | 05-Jun-2014 |
rmind | branches: 1.110.4;
- Implement pktqueue interface for lockless IP input queue.
- Replace ipintrq and ip6intrq with the pktqueue mechanism.
- Eliminate kernel-lock from ipintr() and ip6intr().
- Some preparation work to push softnet_lock out of ipintr().
Discussed on tech-net.
|
| 1.109 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.108 | 20-Mar-2014 |
christos | branches: 1.108.2;
need compat header.
|
| 1.107 | 02-Jan-2014 |
pooka | Allow kernels compiled with INET+INET6 to be booted as IPv4-only or IPv6-only.
|
| 1.106 | 05-Jun-2013 |
christos | branches: 1.106.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.105 | 02-Mar-2013 |
christos | Under FAST_IPSEC, IPSEC_ESP is mandatory; GC it.
|
| 1.104 | 01-Mar-2013 |
joerg | Retire OSI network stack. OK core@
|
| 1.103 | 22-Mar-2012 |
drochner | branches: 1.103.2;
remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.102 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.101 | 03-May-2011 |
dyoung | branches: 1.101.4; 1.101.8;
*_drain() routines may be called with locks held, so instead of doing
any work in *_drain(), set a drain-needed flag. Do the work in the
fasttimo handler.
Contributed by Coyote Point Systems, Inc.
|
| 1.100 | 31-Mar-2011 |
dyoung | Hide the radix-trie implementation of the forwarding table so that we
will have an easier time replacing it with something different, even if
it is a second radix-trie implementation.
sys/net/route.c and sys/net/rtsock.c no longer operate directly on
radix_nodes or radix_node_heads.
Hopefully this will reduce the temptation to implement multipath or
source-based routing using grotty hacks to the grotty old radix-trie
code, too. :-)
|
| 1.99 | 16-Sep-2009 |
pooka | branches: 1.99.4; 1.99.6;
Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.98 | 14-Sep-2009 |
degroote | Import pfsync support from OpenBSD 4.2
Pfsync interface exposes change in the pf(4) over a pseudo-interface, and can
be used to synchronise different pf.
This work was part of my 2009 GSoC
No objection on tech-net@
|
| 1.97 | 28-Feb-2009 |
pooka | include opt_gateway
|
| 1.96 | 01-Feb-2009 |
pooka | branches: 1.96.2;
Init ipflow pool dynamically instead of using a linkset.
|
| 1.95 | 25-Nov-2008 |
pooka | Make dom_maxrtkey of inet/inet6domain the size of the ip_encap pack
structures. This is far from optimal, but gets rid of iffy
#ifdef INET in radix.c. The radix bonsai still needs lots of love
before loading domains dynamically is possible...
|
| 1.94 | 24-Apr-2008 |
ad | branches: 1.94.2; 1.94.8; 1.94.10;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.93 | 23-Apr-2008 |
thorpej | Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.92 | 15-Apr-2008 |
thorpej | branches: 1.92.2;
Make IGMP stats per-cpu.
|
| 1.91 | 05-Oct-2007 |
dyoung | branches: 1.91.18;
Work in progress: use a raw socket for GRE in IP encapsulation
instead of adding/subtracting our own IPv4 header.
There are many benefits: gre(4) needn't grok the outer encapsulation
header any longer, so this simplifies the gre(4) code. The IP
stack needn't grok GRE, so it is simplified, too. gre(4) will
benefit from optimizations in the socket code. Eventually, gre(4)
will gain an IPv6 encapsulation with very few new lines of code.
There is a small performance loss. A 133 MHz, 486-class AMD Elan
sinks/sources a TCP stream over GRE with about 93% the throughput
of the old code. TCP throughput on a 266 MHz, 586-class AMD Geode
is about 96% the throughput of the old code. A 175-MHz ADM5120
(MIPS) only sinks a TCP stream over GRE at about 90% of the old
code; I am still investigating that.
I produced stripped-down versions of sosend() and soreceive() for
gre(4) to use. They are guaranteed not to block, so they can be
called from a software interrupt and from a socket upcall,
respectively.
A kernel thread is no longer necessary for socket transmit/receive,
but I didn't get around to removing it, yet.
Thanks to Matt Thomas for suggesting the use of stripped-down socket
code and software interrupts, and to Andrew Doran for advice and
answers concerning software interrupts, threads, and performance.
|
| 1.90 | 19-Sep-2007 |
dyoung | branches: 1.90.2;
Don't use INADDR_ANY to initialize a const struct, because INADDR_ANY
is not necessarily const.
|
| 1.89 | 19-Sep-2007 |
dyoung | 1) Introduce a new socket option, (SOL_SOCKET, SO_NOHEADER), that
tells a socket that it should both add a protocol header to tx'd
datagrams and remove the header from rx'd datagrams:
int onoff = 1, s = socket(...);
setsockopt(s, SOL_SOCKET, SO_NOHEADER, &onoff);
2) Add an implementation of (SOL_SOCKET, SO_NOHEADER) for raw IPv4
sockets.
3) Reorganize the protocols' pr_ctloutput implementations a bit.
Consistently return ENOPROTOOPT when an option is unsupported,
and EINVAL if a supported option's arguments are incorrect.
Reorganize the flow of code so that it's more clear how/when
options are passed down the stack until they are handled.
Shorten some pr_ctloutput staircases for readability.
4) Extract common mbuf code into subroutines, add new sockaddr
methods, and introduce a new subroutine, fsocreate(), for reuse
later; use it first in sys_socket():
struct mbuf *m_getsombuf(struct socket *so)
Create an mbuf and make its owner the socket `so'.
struct mbuf *m_intopt(struct socket *so, int val)
Create an mbuf, make its owner the socket `so', put the
int `val' into it, and set its length to sizeof(int).
int fsocreate(..., int *fd)
Create a socket, a la socreate(9), put the socket into the
given LWP's descriptor table, return the descriptor at `fd'
on success.
void *sockaddr_addr(struct sockaddr *sa, socklen_t *slenp)
const void *sockaddr_const_addr(const struct sockaddr *sa, socklen_t *slenp)
Extract a pointer to the address part of a sockaddr. Write
the length of the address part at `slenp', if `slenp' is
not NULL.
socklen_t sockaddr_getlen(const struct sockaddr *sa)
Return the length of a sockaddr. This just evaluates to
sa->sa_len. I only add this for consistency with code that
appears in a portable userland library that I am going to
import.
const struct sockaddr *sockaddr_any(const struct sockaddr *sa)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
const void *sockaddr_anyaddr(const struct sockaddr *sa, socklen_t *slenp)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
|
| 1.88 | 30-Aug-2007 |
dyoung | Use malloc(9) for sockaddrs instead of pool(9), and remove dom_sa_pool
and dom_sa_len members from struct domain. Pools of fixed-size
objects are too rigid for sockaddr_dls, whose size can vary over
a wide range.
Return sockaddr_dl to its "historical" size. Now that I'm using
malloc(9) instead of pool(9) to allocate sockaddr_dl, I can create
a sockaddr_dl of any size in the kernel, so expanding sockaddr_dl
is useless.
Avoid using sizeof(struct sockaddr_dl) in the kernel.
Introduce sockaddr_dl_alloc() for allocating & initializing an
arbitrary sockaddr_dl on the heap.
Add an argument, the sockaddr length, to sockaddr_alloc(),
sockaddr_copy(), and sockaddr_dl_setaddr().
Constify: LLADDR() -> CLLADDR().
Where the kernel overwrites LLADDR(), use sockaddr_dl_setaddr(),
instead. Used properly, sockaddr_dl_setaddr() will not overrun
the end of the sockaddr.
|
| 1.87 | 13-Jun-2007 |
dyoung | branches: 1.87.2; 1.87.6; 1.87.8;
Use __arraycount().
|
| 1.86 | 06-May-2007 |
dyoung | In AppleTalk, IPv4, and IPv6 routing domains, help sockaddr_cmp()
avoid an indirect function call by comparing the family, length,
and bytes [dom->dom_sa_cmpofs, dom->dom_sa_cmpofs + dom->dom_sa_cmplen),
corresponding to the the sockaddrs' "address" members.
For ISO, actually use sockaddr_iso_cmp, for a change. Thanks to
yamt@ for pointing out my error.
|
| 1.85 | 02-May-2007 |
dyoung | Remove obsolete files netinet/in_route.[ch].
|
| 1.84 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.83 | 05-Mar-2007 |
liamjfoy | branches: 1.83.2; 1.83.4;
Move ipflow_slowtimo from ip_slowtimo and into in_proto.c
ok matt@
|
| 1.82 | 04-Mar-2007 |
liamjfoy | inet6domain -> inetdomain
thanks simon
|
| 1.81 | 04-Mar-2007 |
liamjfoy | Initialize protocol switch with structure initializers.
ok christos@
|
| 1.80 | 09-Dec-2006 |
dyoung | branches: 1.80.2;
Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.79 | 23-Nov-2006 |
rpaulo | New EtherIP driver based on tap(4) and gif(4) by Hans Rosenfeld.
Notable changes:
* Fixes PR 34268.
* Separates the code from gif(4) (which is more cleaner).
* Allows the usage of STP (Spanning Tree Protocol).
* Removed EtherIP implementation from gif(4)/tap(4).
Some input from Christos.
|
| 1.78 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.77 | 10-Oct-2006 |
dogcow | change the MOWNER_INIT define to take two args; fix extant struct mowner
decls to use it. Makes options MBUFTRACE compile again and not whinge about
missing structure declarations. (Also makes initialization consistent.)
|
| 1.76 | 07-Sep-2006 |
dogcow | branches: 1.76.2; 1.76.4;
remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.
|
| 1.75 | 30-Aug-2006 |
christos | add missing initializers
|
| 1.74 | 28-Aug-2006 |
christos | Remove excess initializer.
|
| 1.73 | 25-Aug-2006 |
matt | One step closer to loadable domains. Store pointers to a domain's soft
interrupt queues so if_detach can remove packets to removed interfaces from
them. This eliminates a lot of conditional ugly code in if.c
|
| 1.72 | 18-May-2006 |
liamjfoy | Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.71 | 11-Dec-2005 |
christos | branches: 1.71.4; 1.71.6; 1.71.8; 1.71.12;
merge ktrace-lwp.
|
| 1.70 | 19-Jul-2005 |
gdt | Add PR_PURGEIF flag for protocols to indicate that the protocol might
store a struct ifnet *, and define it for udp/tcp/rawip for INET and
INET6. When deleting a struct ifnet, invoke PRU_PURGEIF on all
protocols marked with PR_PURGEIF. Closes PR kern/29580 (mine).
|
| 1.69 | 29-Apr-2005 |
yamt | branches: 1.69.2;
move decl of inetsw to its own header to avoid array of incomplete type.
found by gcc4. reported by Adam Ciarcinski.
|
| 1.68 | 12-Feb-2005 |
manu | branches: 1.68.4;
Add support for IPsec Network Address Translator traversal (NAT-T), as
described by RFC 3947 and 3948.
|
| 1.67 | 31-Jan-2005 |
kim | Add RFC 3378 EtherIP support, ported from OpenBSD to NetBSD by
Hans Rosenfeld (rosenfeld at grumpf.hope-2000.org)
This change makes it possible to add gif interfaces to bridges, which
will then send and receive IP protocol 97 packets. Packets are Ethernet
frames with an EtherIP header prepended.
|
| 1.66 | 23-Jan-2005 |
matt | branches: 1.66.2;
Change initialzie of domains to use link sets. Switch to using STAILQ.
Add a convenience macro DOMAIN_FOREACH to interate through the domain.
|
| 1.65 | 04-Sep-2004 |
manu | branches: 1.65.4;
IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.64 | 25-Apr-2004 |
simonb | Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.63 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.62 | 04-Dec-2003 |
atatat | Dynamic sysctl.
Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(),
vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all
nodes are registered with the tree, and nodes can be added (or
removed) easily, and I/O to and from the tree is handled generically.
Since the nodes are registered with the tree, the mapping from name to
number (and back again) can now be discovered, instead of having to be
hard coded. Adding new nodes to the tree is likewise much simpler --
the new infrastructure handles almost all the work for simple types,
and just about anything else can be done with a small helper function.
All existing nodes are where they were before (numerically speaking),
so all existing consumers of sysctl information should notice no
difference.
PS - I'm sorry, but there's a distinct lack of documentation at the
moment. I'm working on sysctl(3/8/9) right now, and I promise to
watch out for buses.
|
| 1.61 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.60 | 14-Aug-2003 |
itojun | enforce ipsec policy on raw wildcard.
|
| 1.59 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.58 | 02-Nov-2002 |
itojun | branches: 1.58.6;
cleanup ipsec.h dependency. commented by perry, sync w/kame
|
| 1.57 | 25-Sep-2002 |
itojun | one too many whitespace
|
| 1.56 | 09-Jun-2002 |
itojun | whitespace
|
| 1.55 | 04-Mar-2002 |
sommerfeld | branches: 1.55.6; 1.55.8;
The "gif*" tunnelling interface does everything ipip does.
Move usage example from ipip.4 to gif.4
Excise ipip and stitch up the scars.
|
| 1.54 | 21-Dec-2001 |
itojun | use radix table for inbound tunnel lookup (would increase performance
for machines with a lot of tunnels).
update route cache for IPvX-over-IPv6 tunnel on path MTU discovery.
snyc with kame
|
| 1.53 | 21-Dec-2001 |
itojun | call rip_ctlinput on icmp4 inputs
|
| 1.52 | 21-Dec-2001 |
itojun | move protosw fragment for gif/stf to their own source code.
reduce #ifdef in stf code. sync with kame
|
| 1.51 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.50 | 30-Oct-2001 |
kml | Add in support for timing out IPv4 routes added due to redirects,
as discussed in tech-net several weeks ago. It turned out that
KAME had already added this functionality to the IPv6 stack, so
I followed their example in adding the sysctl variables
net.inet.icmp.rediraccept and net.inet.icmp.redirtimeout.
|
| 1.49 | 10-Sep-2001 |
thorpej | branches: 1.49.2;
Use a callout for the delayed ACK timer, and delete tcp_fasttimo().
Expose the delayed ACK timer as net.inet.tcp.delack_ticks.
|
| 1.48 | 21-Mar-2001 |
thorpej | branches: 1.48.2; 1.48.4;
Add a protosw flag, PR_ABRTACPTDIS (Abort on Accept of Disconnected
Socket), and add it to the protocols that use that behavior (all
PR_LISTEN protocols except for PF_LOCAL stream sockets).
|
| 1.47 | 01-Mar-2001 |
itojun | branches: 1.47.2;
make sure to enforce inbound ipsec policy checking, for any protocols on top
of ip (check it when final header is visited). sync with kame.
XXX kame team will need to re-check policy engine code
|
| 1.46 | 21-Feb-2001 |
itojun | need PR_ADDR|PR_ATOMIC for IPPROTO_EON. fix typo. from chopps, sync with kame
|
| 1.45 | 20-Feb-2001 |
itojun | ISO over IPv4/v6 by EON encapsulation. from chopps, sync with kame.
|
| 1.44 | 18-Oct-2000 |
thorpej | Restructure the Path MTU Discovery code somewhat to avoid
entering rtentry's for hosts we're not actually communicating
with.
Do this by invoking the ctlinput for the protocol, which is
responsible for validating the ICMP message:
* TCP -- Lookup the connection based on the address/port
pairs in the ICMP message.
* AH/ESP -- Lookup the SA based on the SPI in the ICMP message.
If validation succeeds, ctlinput is responsible for calling
icmp_mtudisc(). icmp_mtudisc() then invokes callbacks registered
by protocols (such as TCP) which want to take some sort of special
action when a path's MTU changes. For TCP, this is where we now
refresh cached routes and re-enter slow-start.
As a side-effect, this fixes the problem where TCP would not be
notified when a path's MTU changed if AH/ESP were being used.
XXX Note, this is only a fix for the IPv4 case. For the IPv6
XXX case, we need to wait for the KAME folks.
Reviewed by sommerfeld@netbsd.org and itojun@netbsd.org.
|
| 1.43 | 18-Oct-2000 |
itojun | move tcp syn cache parameters from in_proto.c to tcp_subr.c.
it makes more sense and helps INET6-only (INET-less) build.
|
| 1.42 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.41 | 27-Jul-2000 |
itojun | implement net.inet.tcp.rstppslimit to limit TCP RSTs by packet-per-second
basis. default: 100pps
set default value for net.inet.tcp.rstratelimit to 0 (disabled),
NOTE: it does not work right for smaller-than-1/hz interval. maybe we should
nuke it, or make it impossible to set smaller-than-1/hz value.
|
| 1.40 | 10-Jul-2000 |
itojun | implement net.inet.icmp.errppslimit.
make default value for net.inet.icmp.erratelimit to 0, as < 10ms value
does not do the right thing.
|
| 1.39 | 19-Apr-2000 |
itojun | branches: 1.39.4;
introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.38 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.37 | 15-Feb-2000 |
thorpej | Add support for rate-limiting RSTs sent in response to no socket for
an incoming packet. Default minimum interval is 10ms. The interval
is changeable via the "net.inet.tcp.rstratelimit" sysctl variable.
|
| 1.36 | 15-Feb-2000 |
thorpej | Add ICMP error rate limiting, based on the same for ICMP6.
Note, we're reusing the previously unused slot for "MTU discovery" (which
was moved to the "net.inet.ip" branch of the sysctl tree quite some time
ago).
|
| 1.35 | 10-Feb-2000 |
itojun | fix ip4 protosw.
gif interface and gre interface should be able to coexist.
|
| 1.34 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.33 | 09-Jul-1999 |
thorpej | branches: 1.33.2; 1.33.8;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.32 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.31 | 01-Jul-1999 |
darrenr | add PR_LISTEN to protocols which support listen(2)
|
| 1.30 | 29-Apr-1999 |
thorpej | Implement retransmit logic for the SYN cache engine. Fixes a rare condition
where one side can think a connection exists, where the other side thinks
the connection was never established.
The original problem was first reported by Ty Sarna in PR #5909. The
original fix I made to the code didn't cover all cases. The problem this
fix addresses was reported by Christoph Badura via private e-mail.
Many thanks to Bill Sommerfeld for helping me to test this code, and
for finding a subtle bug.
|
| 1.29 | 14-Jan-1999 |
thorpej | branches: 1.29.2;
Domains are associated with protocol families, not address families.
|
| 1.28 | 11-Jan-1999 |
thorpej | Adjust for the new IP-IP input path.
|
| 1.27 | 22-Dec-1998 |
thorpej | ipip_input() -> mrt_ipip_input().
|
| 1.26 | 30-Sep-1998 |
hwr | Start supporting IPPROTO_MOBILE (55) encapsulation. This is yet
another tunneling protocol used by the Mobile-IP people. See RFC 2004
for this.
|
| 1.25 | 13-Sep-1998 |
hwr | Add a gre tunnel pseudo network device. Gre = generic route encapsulation.
This device shows up like any other network interface and can be used to
tunnel L3 protocols as e.g. IP over IP.
|
| 1.24 | 15-Jul-1998 |
thorpej | Garbage collect `imp' and `hy'. We don't have the rest of the code, and
it's not like anyone is ever going to be using either of them.
|
| 1.23 | 05-Jul-1998 |
jonathan | defopt NS, NSIP.
|
| 1.22 | 05-Jul-1998 |
jonathan | defopt ISO TPIP.
|
| 1.21 | 05-Jul-1998 |
jonathan | defopt EON.
|
| 1.20 | 07-May-1998 |
thorpej | Rework the syn cache code somewhat:
- Don't use home-grown queue manipulation. Use <sys/queue.h> instead. The
data structures are a little larger, but we are otherwise wasting the
memory chunk anyway (we're already a 64-byte malloc bucket).
- Fix a bug in the cache-is-full case: if the oldest element removed from
the first non-empty bucket was the only element in the bucket, the
bucket wouldn't be removed from the bucket cache, causing queue corruption
later.
- Optimize the syn cache timers by using PRT timers rather than home-grown
decrement-and-propagate timers.
This code is now a fair bit smaller, and significantly easier to read
and understand.
|
| 1.19 | 12-Jan-1998 |
scottr | Use option header file for MROUTING
|
| 1.18 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.17 | 23-Jul-1997 |
thorpej | Pull SYN_cache_branch down into the main line.
|
| 1.16 | 10-Oct-1996 |
christos | branches: 1.16.8;
- fix NSIP; it referenced non-existing functions.
|
| 1.15 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.14 | 18-Feb-1996 |
christos | Fix PR/2095 options MROUTING did not compile.
|
| 1.13 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.12 | 30-Sep-1995 |
thorpej | Implement tcp_sysctl(). Add a sysctl option to enable/disable RFC1323
extensions to TCP. From John Kohl <jtk@kolvir.blrc.ma.us>.
|
| 1.11 | 31-May-1995 |
mycroft | Integrate multicast 3.5 distribution, with several bugs fixed and general
cleanup. This is a (working) snapshot of work in progress.
|
| 1.10 | 31-May-1995 |
mycroft | Implement IGMP v2. Based on the Multicast 3.5 distribution.
|
| 1.9 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.8 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.7 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.6 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 10-Apr-1993 |
glass | fixed to be compliant, subservient, and to take advantage of the newly
hacked config(8)
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.16.8.1 | 14-May-1997 |
mellon | Add syn_cache variables
|
| 1.29.2.1 | 29-Apr-1999 |
perry | branches: 1.29.2.1.2; 1.29.2.1.4;
pullup 1.29->1.30 (thorpej)
|
| 1.29.2.1.4.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.29.2.1.4.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.29.2.1.4.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.29.2.1.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.29.2.1.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.29.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.33.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.33.2.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.33.2.2 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.33.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.39.4.4 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #143)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.39.4.3 | 09-Sep-2003 |
msaitoh | Pull up rev. 1.60 via patch (requested by itojun in ticket #68):
enforce ipsec policy on raw wildcard.
|
| 1.39.4.2 | 11-Mar-2001 |
he | Pull up revision 1.47 (via patch, requested by itojun):
Ensure that we enforce inbound IPsec policy on all IP protocols,
not just TCP, UDP and ICMP.
|
| 1.39.4.1 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.47.2.8 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.47.2.7 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.47.2.6 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.47.2.5 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.47.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.47.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.47.2.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.47.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.48.4.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.48.2.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.48.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.48.2.3 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.48.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.48.2.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.49.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.55.8.1 | 04-Oct-2003 |
tron | Pull up revision 1.60 (requested by itojun in ticket #1409):
enforce ipsec policy on raw wildcard.
|
| 1.55.6.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.58.6.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.58.6.6 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.58.6.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.58.6.4 | 24-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.58.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.58.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.58.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.65.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.66.2.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.68.4.1 | 15-Aug-2005 |
tron | Pull up revision 1.70 (requested by gdt in ticket #661):
Add PR_PURGEIF flag for protocols to indicate that the protocol might
store a struct ifnet *, and define it for udp/tcp/rawip for INET and
INET6. When deleting a struct ifnet, invoke PRU_PURGEIF on all
protocols marked with PR_PURGEIF. Closes PR kern/29580 (mine).
|
| 1.69.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.69.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.69.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.69.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.71.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.71.8.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.71.8.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.71.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.71.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.71.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.76.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.76.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.76.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.76.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.80.2.2 | 07-May-2007 |
yamt | sync with head.
|
| 1.80.2.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.83.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.83.2.3 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.83.2.2 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.83.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.87.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.87.6.3 | 07-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.87.6.2 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.87.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.87.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.90.2.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.91.18.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.91.18.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.92.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.94.10.2 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.94.10.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.94.8.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.94.2.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.94.2.2 | 16-Sep-2009 |
yamt | sync with head
|
| 1.94.2.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.96.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.99.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.99.4.2 | 31-May-2011 |
rmind | sync with head
|
| 1.99.4.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.101.8.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.101.8.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.101.4.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.101.4.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.103.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.103.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.103.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.106.2.3 | 18-May-2014 |
rmind | sync with head
|
| 1.106.2.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.106.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.108.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.110.4.8 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.110.4.7 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.110.4.6 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.110.4.5 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.110.4.4 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.110.4.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.110.4.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.110.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.120.4.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.120.2.2 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.120.2.1 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.123.4.3 | 31-Mar-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #676):
sys/netinet/in_proto.c: revision 1.127
sys/netinet6/in6_proto.c: revision 1.122
Add the PR_LASTHDR flag on the PFsync and CARP entries. Otherwise a
"require" IPsec policy is not enforced on them, and unauthenticated
packets will be accepted.
Tested with a require-AH configuration. Sent on tech-net@, no comment.
|
| 1.123.4.2 | 24-Oct-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #305):
distrib/sets/lists/tests/mi: revision 1.762
sys/net/route.c: revision 1.198-1.201
sys/net/route.h: revision 1.114
sys/netatalk/at_proto.c: revision 1.22
sys/netinet/in_proto.c: revision 1.124
sys/netinet6/in6_proto.c: revision 1.118
sys/netmpls/mpls_proto.c: revision 1.31
sys/netnatm/natm_proto.c: revision 1.18
sys/rump/net/lib/libsockin/sockin.c: revision 1.65
sys/sys/domain.h: revision 1.33
tests/net/route/Makefile: revision 1.6
tests/net/route/t_rtcache.sh: revision 1.1
Add tests of rtcache invalidation
Remove unnecessary NULL check of rt_ifp
It's always non-NULL.
Invalidate rtcache based on a global generation counter
The change introduces a global generation counter that is incremented when any
routes have been added or deleted. When a rtcache caches a rtentry into itself,
it also stores a snapshot of the generation counter. If the snapshot equals to
the global counter, the cache is still valid, otherwise invalidated.
One drawback of the change is that all rtcaches of all protocol families are
invalidated when any routes of any protocol families are added or deleted.
If that matters, we should have separate generation counters based on
protocol families.
This change removes LIST_ENTRY from struct route, which fixes a part of
PR kern/52515.
Remove the global lock for rtcache
Thanks to removal of LIST_ENTRY of struct route, rtcaches are accessed only by
their users. And in existing usages a rtcache is guranteed to be not accessed
simultaneously. So the rtcache framework doesn't need any exclusion controls
in itself.
Synchronize on rtcache_generation with rtlock
It's racy if NET_MPSAFE is enabled.
Pointed out by joerg@
|
| 1.123.4.1 | 21-Oct-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #300):
crypto/dist/ipsec-tools/src/setkey/parse.y: 1.19
crypto/dist/ipsec-tools/src/setkey/token.l: 1.20
distrib/sets/lists/tests/mi: 1.754, 1.757, 1.759
doc/TODO.smpnet: 1.12-1.13
sys/net/pfkeyv2.h: 1.32
sys/net/raw_cb.c: 1.23-1.24, 1.28
sys/net/raw_cb.h: 1.28
sys/net/raw_usrreq.c: 1.57-1.58
sys/net/rtsock.c: 1.228-1.229
sys/netinet/in_proto.c: 1.125
sys/netinet/ip_input.c: 1.359-1.361
sys/netinet/tcp_input.c: 1.359-1.360
sys/netinet/tcp_output.c: 1.197
sys/netinet/tcp_var.h: 1.178
sys/netinet6/icmp6.c: 1.213
sys/netinet6/in6_proto.c: 1.119
sys/netinet6/ip6_forward.c: 1.88
sys/netinet6/ip6_input.c: 1.181-1.182
sys/netinet6/ip6_output.c: 1.193
sys/netinet6/ip6protosw.h: 1.26
sys/netipsec/ipsec.c: 1.100-1.122
sys/netipsec/ipsec.h: 1.51-1.61
sys/netipsec/ipsec6.h: 1.18-1.20
sys/netipsec/ipsec_input.c: 1.44-1.51
sys/netipsec/ipsec_netbsd.c: 1.41-1.45
sys/netipsec/ipsec_output.c: 1.49-1.64
sys/netipsec/ipsec_private.h: 1.5
sys/netipsec/key.c: 1.164-1.234
sys/netipsec/key.h: 1.20-1.32
sys/netipsec/key_debug.c: 1.18-1.21
sys/netipsec/key_debug.h: 1.9
sys/netipsec/keydb.h: 1.16-1.20
sys/netipsec/keysock.c: 1.59-1.62
sys/netipsec/keysock.h: 1.10
sys/netipsec/xform.h: 1.9-1.12
sys/netipsec/xform_ah.c: 1.55-1.74
sys/netipsec/xform_esp.c: 1.56-1.72
sys/netipsec/xform_ipcomp.c: 1.39-1.53
sys/netipsec/xform_ipip.c: 1.50-1.54
sys/netipsec/xform_tcp.c: 1.12-1.16
sys/rump/librump/rumpkern/Makefile.rumpkern: 1.170
sys/rump/librump/rumpnet/net_stub.c: 1.27
sys/sys/protosw.h: 1.67-1.68
tests/net/carp/t_basic.sh: 1.7
tests/net/if_gif/t_gif.sh: 1.11
tests/net/if_l2tp/t_l2tp.sh: 1.3
tests/net/ipsec/Makefile: 1.7-1.9
tests/net/ipsec/algorithms.sh: 1.5
tests/net/ipsec/common.sh: 1.4-1.6
tests/net/ipsec/t_ipsec_ah_keys.sh: 1.2
tests/net/ipsec/t_ipsec_esp_keys.sh: 1.2
tests/net/ipsec/t_ipsec_gif.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_l2tp.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_misc.sh: 1.8-1.18
tests/net/ipsec/t_ipsec_sockopt.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tcp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_transport.sh: 1.5-1.6
tests/net/ipsec/t_ipsec_tunnel.sh: 1.9
tests/net/ipsec/t_ipsec_tunnel_ipcomp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tunnel_odd.sh: 1.3
tests/net/mcast/t_mcast.sh: 1.6
tests/net/net/t_ipaddress.sh: 1.11
tests/net/net_common.sh: 1.20
tests/net/npf/t_npf.sh: 1.3
tests/net/route/t_flags.sh: 1.20
tests/net/route/t_flags6.sh: 1.16
usr.bin/netstat/fast_ipsec.c: 1.22
Do m_pullup before mtod
It may fix panicks of some tests on anita/sparc and anita/GuruPlug.
---
KNF
---
Enable DEBUG for babylon5
---
Apply C99-style struct initialization to xformsw
---
Tweak outputs of netstat -s for IPsec
- Get rid of "Fast"
- Use ipsec and ipsec6 for titles to clarify protocol
- Indent outputs of sub protocols
Original outputs were organized like this:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
New outputs are organized like this:
ipsec:
ah:
esp:
ipip:
ipcomp:
ipsec6:
ah:
esp:
ipip:
ipcomp:
---
Add test cases for IPComp
---
Simplify IPSEC_OSTAT macro (NFC)
---
KNF; replace leading whitespaces with hard tabs
---
Introduce and use SADB_SASTATE_USABLE_P
---
KNF
---
Add update command for testing
Updating an SA (SADB_UPDATE) requires that a process issuing
SADB_UPDATE is the same as a process issued SADB_ADD (or SADB_GETSPI).
This means that update command must be used with add command in a
configuration of setkey. This usage is normally meaningless but
useful for testing (and debugging) purposes.
---
Add test cases for updating SA/SP
The tests require newly-added udpate command of setkey.
---
PR/52346: Frank Kardel: Fix checksumming for NAT-T
See XXX for improvements.
---
Remove codes for PACKET_TAG_IPSEC_IN_CRYPTO_DONE
It seems that PACKET_TAG_IPSEC_IN_CRYPTO_DONE is for network adapters
that have IPsec accelerators; a driver sets the mtag to a packet
when its device has already encrypted the packet.
Unfortunately no driver implements such offload features for long
years and seems unlikely to implement them soon. (Note that neither
FreeBSD nor Linux doesn't have such drivers.) Let's remove related
(unused) codes and simplify the IPsec code.
---
Fix usages of sadb_msg_errno
---
Avoid updating sav directly
On SADB_UPDATE a target sav was updated directly, which was unsafe.
Instead allocate another sav, copy variables of the old sav to
the new one and replace the old one with the new one.
---
Simplify; we can assume sav->tdb_xform cannot be NULL while it's valid
---
Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
---
Use explicit_memset to surely zero-clear key_auth and key_enc
---
Make sure to clear keys on error paths of key_setsaval
---
Add missing KEY_FREESAV
---
Make sure a sav is inserted to a sah list after its initialization completes
---
Remove unnecessary zero-clearing codes from key_setsaval
key_setsaval is now used only for a newly-allocated sav. (It was
used to reset variables of an existing sav.)
---
Correct wrong assumption of sav->refcnt in key_delsah
A sav in a list is basically not to be sav->refcnt == 0. And also
KEY_FREESAV assumes sav->refcnt > 0.
---
Let key_getsavbyspi take a reference of a returning sav
---
Use time_mono_to_wall (NFC)
---
Separate sending message routine (NFC)
---
Simplify; remove unnecessary zero-clears
key_freesaval is used only when a target sav is being destroyed.
---
Omit NULL checks for sav->lft_c
sav->lft_c can be NULL only when initializing or destroying sav.
---
Omit unnecessary NULL checks for sav->sah
---
Omit unnecessary check of sav->state
key_allocsa_policy picks a sav of either MATURE or DYING so we
don't need to check its state again.
---
Simplify; omit unnecessary saidx passing
- ipsec_nextisr returns a saidx but no caller uses it
- key_checkrequest is passed a saidx but it can be gotton by
another argument (isr)
---
Fix splx isn't called on some error paths
---
Fix header size calculation of esp where sav is NULL
---
Fix header size calculation of ah in the case sav is NULL
This fix was also needed for esp.
---
Pass sav directly to opencrypto callback
In a callback, use a passed sav as-is by default and look up a sav
only if the passed sav is dead.
---
Avoid examining freshness of sav on packet processing
If a sav list is sorted (by lft_c->sadb_lifetime_addtime) in advance,
we don't need to examine each sav and also don't need to delete one
on the fly and send up a message. Fortunately every sav lists are sorted
as we need.
Added key_validate_savlist validates that each sav list is surely sorted
(run only if DEBUG because it's not cheap).
---
Add test cases for SAs with different SPIs
---
Prepare to stop using isr->sav
isr is a shared resource and using isr->sav as a temporal storage
for each packet processing is racy. And also having a reference from
isr to sav makes the lifetime of sav non-deterministic; such a reference
is removed when a packet is processed and isr->sav is overwritten by
new one. Let's have a sav locally for each packet processing instead of
using shared isr->sav.
However this change doesn't stop using isr->sav yet because there are
some users of isr->sav. isr->sav will be removed after the users find
a way to not use isr->sav.
---
Fix wrong argument handling
---
fix printf format.
---
Don't validate sav lists of LARVAL or DEAD states
We don't sort the lists so the validation will always fail.
Fix PR kern/52405
---
Make sure to sort the list when changing the state by key_sa_chgstate
---
Rename key_allocsa_policy to key_lookup_sa_bysaidx
---
Separate test files
---
Calculate ah_max_authsize on initialization as well as esp_max_ivlen
---
Remove m_tag_find(PACKET_TAG_IPSEC_PENDING_TDB) because nobody sets the tag
---
Restore a comment removed in previous
The comment is valid for the below code.
---
Make tests more stable
sleep command seems to wait longer than expected on anita so
use polling to wait for a state change.
---
Add tests that explicitly delete SAs instead of waiting for expirations
---
Remove invalid M_AUTHIPDGM check on ESP isr->sav
M_AUTHIPDGM flag is set to a mbuf in ah_input_cb. An sav of ESP can
have AH authentication as sav->tdb_authalgxform. However, in that
case esp_input and esp_input_cb are used to do ESP decryption and
AH authentication and M_AUTHIPDGM never be set to a mbuf. So
checking M_AUTHIPDGM of a mbuf on isr->sav of ESP is meaningless.
---
Look up sav instead of relying on unstable sp->req->sav
This code is executed only in an error path so an additional lookup
doesn't matter.
---
Correct a comment
---
Don't release sav if calling crypto_dispatch again
---
Remove extra KEY_FREESAV from ipsec_process_done
It should be done by the caller.
---
Don't bother the case of crp->crp_buf == NULL in callbacks
---
Hold a reference to an SP during opencrypto processing
An SP has a list of isr (ipsecrequest) that represents a sequence
of IPsec encryption/authentication processing. One isr corresponds
to one opencrypto processing. The lifetime of an isr follows its SP.
We pass an isr to a callback function of opencrypto to continue
to a next encryption/authentication processing. However nobody
guaranteed that the isr wasn't freed, i.e., its SP wasn't destroyed.
In order to avoid such unexpected destruction of isr, hold a reference
to its SP during opencrypto processing.
---
Don't make SAs expired on tests that delete SAs explicitly
---
Fix a debug message
---
Dedup error paths (NFC)
---
Use pool to allocate tdb_crypto
For ESP and AH, we need to allocate an extra variable space in addition
to struct tdb_crypto. The fixed size of pool items may be larger than
an actual requisite size of a buffer, but still the performance
improvement by replacing malloc with pool wins.
---
Don't use unstable isr->sav for header size calculations
We may need to optimize to not look up sav here for users that
don't need to know an exact size of headers (e.g., TCP segmemt size
caclulation).
---
Don't use sp->req->sav when handling NAT-T ESP fragmentation
In order to do this we need to look up a sav however an additional
look-up degrades performance. A sav is later looked up in
ipsec4_process_packet so delay the fragmentation check until then
to avoid an extra look-up.
---
Don't use key_lookup_sp that depends on unstable sp->req->sav
It provided a fast look-up of SP. We will provide an alternative
method in the future (after basic MP-ification finishes).
---
Stop setting isr->sav on looking up sav in key_checkrequest
---
Remove ipsecrequest#sav
---
Stop setting mtag of PACKET_TAG_IPSEC_IN_DONE because there is no users anymore
---
Skip ipsec_spi_*_*_preferred_new_timeout when running on qemu
Probably due to PR 43997
---
Add localcount to rump kernels
---
Remove unused macro
---
Fix key_getcomb_setlifetime
The fix adjusts a soft limit to be 80% of a corresponding hard limit.
I'm not sure the fix is really correct though, at least the original
code is wrong. A passed comb is zero-cleared before calling
key_getcomb_setlifetime, so
comb->sadb_comb_soft_addtime = comb->sadb_comb_soft_addtime * 80 / 100;
is meaningless.
---
Provide and apply key_sp_refcnt (NFC)
It simplifies further changes.
---
Fix indentation
Pointed out by knakahara@
---
Use pslist(9) for sptree
---
Don't acquire global locks for IPsec if NET_MPSAFE
Note that the change is just to make testing easy and IPsec isn't MP-safe yet.
---
Let PF_KEY socks hold their own lock instead of softnet_lock
Operations on SAD and SPD are executed via PF_KEY socks. The operations
include deletions of SAs and SPs that will use synchronization mechanisms
such as pserialize_perform to wait for references to SAs and SPs to be
released. It is known that using such mechanisms with holding softnet_lock
causes a dead lock. We should avoid the situation.
---
Make IPsec SPD MP-safe
We use localcount(9), not psref(9), to make the sptree and secpolicy (SP)
entries MP-safe because SPs need to be referenced over opencrypto
processing that executes a callback in a different context.
SPs on sockets aren't managed by the sptree and can be destroyed in softint.
localcount_drain cannot be used in softint so we delay the destruction of
such SPs to a thread context. To do so, a list to manage such SPs is added
(key_socksplist) and key_timehandler_spd deletes dead SPs in the list.
For more details please read the locking notes in key.c.
Proposed on tech-kern@ and tech-net@
---
Fix updating ipsec_used
- key_update_used wasn't called in key_api_spddelete2 and key_api_spdflush
- key_update_used wasn't called if an SP had been added/deleted but
a reply to userland failed
---
Fix updating ipsec_used; turn on when SPs on sockets are added
---
Add missing IPsec policy checks to icmp6_rip6_input
icmp6_rip6_input is quite similar to rip6_input and the same checks exist
in rip6_input.
---
Add test cases for setsockopt(IP_IPSEC_POLICY)
---
Don't use KEY_NEWSP for dummy SP entries
By the change KEY_NEWSP is now not called from softint anymore
and we can use kmem_zalloc with KM_SLEEP for KEY_NEWSP.
---
Comment out unused functions
---
Add test cases that there are SPs but no relevant SAs
---
Don't allow sav->lft_c to be NULL
lft_c of an sav that was created by SADB_GETSPI could be NULL.
---
Clean up clunky eval strings
- Remove unnecessary \ at EOL
- This allows to omit ; too
- Remove unnecessary quotes for arguments of atf_set
- Don't expand $DEBUG in eval
- We expect it's expanded on execution
Suggested by kre@
---
Remove unnecessary KEY_FREESAV in an error path
sav should be freed (unreferenced) by the caller.
---
Use pslist(9) for sahtree
---
Use pslist(9) for sah->savtree
---
Rename local variable newsah to sah
It may not be new.
---
MP-ify SAD slightly
- Introduce key_sa_mtx and use it for some list operations
- Use pserialize for some list iterations
---
Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
---
Remove out-of-date log output
Pointed out by riastradh@
---
Use KDASSERT instead of KASSERT for mutex_ownable
Because mutex_ownable is too heavy to run in a fast path
even for DIAGNOSTIC + LOCKDEBUG.
Suggested by riastradh@
---
Assemble global lists and related locks into cache lines (NFCI)
Also rename variable names from *tree to *list because they are
just lists, not trees.
Suggested by riastradh@
---
Move locking notes
---
Update the locking notes
- Add locking order
- Add locking notes for misc lists such as reglist
- Mention pserialize, key_sp_ref and key_sp_unref on SP operations
Requested by riastradh@
---
Describe constraints of key_sp_ref and key_sp_unref
Requested by riastradh@
---
Hold key_sad.lock on SAVLIST_WRITER_INSERT_TAIL
---
Add __read_mostly to key_psz
Suggested by riastradh@
---
Tweak wording (pserialize critical section => pserialize read section)
Suggested by riastradh@
---
Add missing mutex_exit
---
Fix setkey -D -P outputs
The outputs were tweaked (by me), but I forgot updating libipsec
in my local ATF environment...
---
MP-ify SAD (key_sad.sahlist and sah entries)
localcount(9) is used to protect key_sad.sahlist and sah entries
as well as SPD (and will be used for SAD sav).
Please read the locking notes of SAD for more details.
---
Introduce key_sa_refcnt and replace sav->refcnt with it (NFC)
---
Destroy sav only in the loop for DEAD sav
---
Fix KASSERT(solocked(sb->sb_so)) failure in sbappendaddr that is called eventually from key_sendup_mbuf
If key_sendup_mbuf isn't passed a socket, the assertion fails.
Originally in this case sb->sb_so was softnet_lock and callers
held softnet_lock so the assertion was magically satisfied.
Now sb->sb_so is key_so_mtx and also softnet_lock isn't always
held by callers so the assertion can fail.
Fix it by holding key_so_mtx if key_sendup_mbuf isn't passed a socket.
Reported by knakahara@
Tested by knakahara@ and ozaki-r@
---
Fix locking notes of SAD
---
Fix deadlock between key_sendup_mbuf called from key_acquire and localcount_drain
If we call key_sendup_mbuf from key_acquire that is called on packet
processing, a deadlock can happen like this:
- At key_acquire, a reference to an SP (and an SA) is held
- key_sendup_mbuf will try to take key_so_mtx
- Some other thread may try to localcount_drain to the SP with
holding key_so_mtx in say key_api_spdflush
- In this case localcount_drain never return because key_sendup_mbuf
that has stuck on key_so_mtx never release a reference to the SP
Fix the deadlock by deferring key_sendup_mbuf to the timer
(key_timehandler).
---
Fix that prev isn't cleared on retry
---
Limit the number of mbufs queued for deferred key_sendup_mbuf
It's easy to be queued hundreds of mbufs on the list under heavy
network load.
---
MP-ify SAD (savlist)
localcount(9) is used to protect savlist of sah. The basic design is
similar to MP-ifications of SPD and SAD sahlist. Please read the
locking notes of SAD for more details.
---
Simplify ipsec_reinject_ipstack (NFC)
---
Add per-CPU rtcache to ipsec_reinject_ipstack
It reduces route lookups and also reduces rtcache lock contentions
when NET_MPSAFE is enabled.
---
Use pool_cache(9) instead of pool(9) for tdb_crypto objects
The change improves network throughput especially on multi-core systems.
---
Update
ipsec(4), opencrypto(9) and vlan(4) are now MP-safe.
---
Write known issues on scalability
---
Share a global dummy SP between PCBs
It's never be changed so it can be pre-allocated and shared safely between PCBs.
---
Fix race condition on the rawcb list shared by rtsock and keysock
keysock now protects itself by its own mutex, which means that
the rawcb list is protected by two different mutexes (keysock's one
and softnet_lock for rtsock), of course it's useless.
Fix the situation by having a discrete rawcb list for each.
---
Use a dedicated mutex for rt_rawcb instead of softnet_lock if NET_MPSAFE
---
fix localcount leak in sav. fixed by ozaki-r@n.o.
I commit on behalf of him.
---
remove unnecessary comment.
---
Fix deadlock between pserialize_perform and localcount_drain
A typical ussage of localcount_drain looks like this:
mutex_enter(&mtx);
item = remove_from_list();
pserialize_perform(psz);
localcount_drain(&item->localcount, &cv, &mtx);
mutex_exit(&mtx);
This sequence can cause a deadlock which happens for example on the following
situation:
- Thread A calls localcount_drain which calls xc_broadcast after releasing
a specified mutex
- Thread B enters the sequence and calls pserialize_perform with holding
the mutex while pserialize_perform also calls xc_broadcast
- Thread C (xc_thread) that calls an xcall callback of localcount_drain tries
to hold the mutex
xc_broadcast of thread B doesn't start until xc_broadcast of thread A
finishes, which is a feature of xcall(9). This means that pserialize_perform
never complete until xc_broadcast of thread A finishes. On the other hand,
thread C that is a callee of xc_broadcast of thread A sticks on the mutex.
Finally the threads block each other (A blocks B, B blocks C and C blocks A).
A possible fix is to serialize executions of the above sequence by another
mutex, but adding another mutex makes the code complex, so fix the deadlock
by another way; the fix is to release the mutex before pserialize_perform
and instead use a condvar to prevent pserialize_perform from being called
simultaneously.
Note that the deadlock has happened only if NET_MPSAFE is enabled.
---
Add missing ifdef NET_MPSAFE
---
Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
---
- sanitize key debugging so that we don't print extra newlines or unassociated
debugging messages.
- remove unused functions and make internal ones static
- print information in one line per message
---
humanize printing of ip addresses
---
cast reduction, NFC.
---
Fix typo in comment
---
Pull out ipsec_fill_saidx_bymbuf (NFC)
---
Don't abuse key_checkrequest just for looking up sav
It does more than expected for example key_acquire.
---
Fix SP is broken on transport mode
isr->saidx was modified accidentally in ipsec_nextisr.
Reported by christos@
Helped investigations by christos@ and knakahara@
---
Constify isr at many places (NFC)
---
Include socketvar.h for softnet_lock
---
Fix buffer length for ipsec_logsastr
|
| 1.126.2.4 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.126.2.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.126.2.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.126.2.1 | 15-Mar-2018 |
pgoyette | Synch with HEAD
|
| 1.128.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.4 | 21-Jan-2016 |
riastradh | Revert previous: ran cvs commit when I meant cvs diff. Sorry!
Hit up-arrow one too few times.
|
| 1.3 | 21-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.2 | 10-Dec-2005 |
elad | branches: 1.2.140;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.1 | 29-Apr-2005 |
yamt | branches: 1.1.2; 1.1.4; 1.1.10;
move decl of inetsw to its own header to avoid array of incomplete type.
found by gcc4. reported by Adam Ciarcinski.
|
| 1.1.10.3 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.1.10.2 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.1.10.1 | 29-Apr-2005 |
skrll | file in_proto.h was added on branch ktrace-lwp on 2005-11-10 14:11:07 +0000
|
| 1.1.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.1.2.2 | 29-Apr-2005 |
kent | sync with -current
|
| 1.1.2.1 | 29-Apr-2005 |
kent | file in_proto.h was added on branch kent-audio2 on 2005-04-29 11:29:33 +0000
|
| 1.2.140.1 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.8 | 02-May-2007 |
dyoung | Remove obsolete files netinet/in_route.[ch].
|
| 1.7 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.6 | 22-Apr-2007 |
dyoung | In in_rtflushall(), clear the route caches using rtcache_clear()
instead of rtcache_free(). It is not desirable to clear the cached
destination as well as the route, however, rtcache_free() will
eventually release all resources held by the cache, including the
destination.
Add some additional diagnostic assertions.
|
| 1.5 | 18-Apr-2007 |
dyoung | Add optimization hint for compiler. In a debug printf,
s/freeing/flushing/.
|
| 1.4 | 17-Feb-2007 |
dyoung | branches: 1.4.4; 1.4.6;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.3 | 17-Feb-2007 |
dyoung | branches: 1.3.2;
s/in_rtflush/in_rtcache/g
|
| 1.2 | 05-Jan-2007 |
joerg | branches: 1.2.2;
Use rtcache_free for consistency.
|
| 1.1 | 09-Dec-2006 |
dyoung | branches: 1.1.2; 1.1.4;
Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.1.4.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.1.4.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.1.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 09-Dec-2006 |
yamt | file in_route.c was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.1.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.1.2.1 | 09-Dec-2006 |
yamt | file in_route.c was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.2.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.2.2.1 | 05-Jan-2007 |
ad | file in_route.c was added on branch newlock2 on 2007-01-12 01:04:14 +0000
|
| 1.3.2.3 | 07-May-2007 |
yamt | sync with head.
|
| 1.3.2.2 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.3.2.1 | 17-Feb-2007 |
yamt | file in_route.c was added on branch yamt-idlelwp on 2007-02-27 16:54:53 +0000
|
| 1.4.6.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.4.4.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.2 | 02-May-2007 |
dyoung | Remove obsolete files netinet/in_route.[ch].
|
| 1.1 | 09-Dec-2006 |
dyoung | branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.12; 1.1.14;
Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.1.14.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.1.12.2 | 09-Jun-2007 |
ad | Sync with head.
|
| 1.1.12.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.1.8.1 | 07-May-2007 |
yamt | sync with head.
|
| 1.1.6.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.1.6.1 | 09-Dec-2006 |
ad | file in_route.h was added on branch newlock2 on 2007-01-12 01:04:14 +0000
|
| 1.1.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.1.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 09-Dec-2006 |
yamt | file in_route.h was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.1.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.1.2.1 | 09-Dec-2006 |
yamt | file in_route.h was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.17 | 07-Jul-2016 |
ozaki-r | Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.16 | 21-Sep-2015 |
skrll | Make this compile again
|
| 1.15 | 31-Aug-2015 |
ozaki-r | Hook up lltable/llentry with the kernel (and rumpkernel)
It is built and initialized on bootup, but there is no user for now.
Most codes in in.c are imported from FreeBSD as well as lltable/llentry.
|
| 1.14 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.13 | 08-Jun-2015 |
roy | Don't set errno. Thanks to skrll@
|
| 1.12 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.11 | 25-Feb-2014 |
pooka | branches: 1.11.6;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.10 | 21-Oct-2013 |
christos | fix type of sysctl, from nisimura@
|
| 1.9 | 02-Jun-2012 |
dsl | branches: 1.9.2; 1.9.4;
Add some pre-processor magic to verify that the type of the data item
passed to sysctl_createv() actually matches the declared type for
the item itself.
In the places where the caller specifies a function and a structure
address (typically the 'softc') an explicit (void *) cast is now needed.
Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c
sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting
AcpiGbl_EnableAmlDebugObject.
(mostly passing the address of a uint64_t when typed as CTLTYPE_INT).
I've test built quite a few kernels, but there may be some unfixed MD
fallout. Most likely passing &char[] to char *.
Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
|
| 1.8 | 19-Oct-2009 |
rmind | branches: 1.8.12;
Drop 3rd and 4th clauses from David Young's license.
Reviewed and approved by dyoung@ (copyright holder).
|
| 1.7 | 30-Aug-2009 |
dyoung | Stop the admin from creating nodes under net.inet.ip.interfaces or
net.inet.ip.interfaces.<ifname>.
|
| 1.6 | 04-Dec-2007 |
dyoung | branches: 1.6.16; 1.6.26; 1.6.34;
Use IFADDR_FOREACH().
|
| 1.5 | 22-Feb-2007 |
dyoung | branches: 1.5.16; 1.5.18; 1.5.24; 1.5.26;
Reverse sense of preference numbers: prefer source addresses with
higher preference numbers. Thanks to Mihai Chelaru for pointing
out my mistake.
|
| 1.4 | 22-Feb-2007 |
dyoung | Add net.inet.ip.selectsrc.default even if GETIFA_DEBUG is not
#define'd.
|
| 1.3 | 16-Nov-2006 |
christos | branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8; 1.3.10;
__unused removal on arguments; approved by core.
|
| 1.2 | 13-Nov-2006 |
dyoung | Plug memory leak.
|
| 1.1 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.3.10.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.3.8.4 | 07-Dec-2007 |
yamt | sync with head
|
| 1.3.8.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.3.8.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.3.8.1 | 16-Nov-2006 |
yamt | file in_selsrc.c was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.3.6.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.3.6.1 | 16-Nov-2006 |
yamt | file in_selsrc.c was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.3.4.1 | 04-Mar-2007 |
bouyer | Pull up following revision(s) (requested by dyoung in ticket #481):
sys/netinet/in_selsrc.c: revision 1.4
sys/netinet/in_selsrc.c: revision 1.5
share/man/man9/in_getifa.9: revision 1.3
Add net.inet.ip.selectsrc.default even if GETIFA_DEBUG is not
Reverse sense of preference numbers: prefer source addresses with
higher preference numbers. Thanks to Mihai Chelaru for pointing
out my mistake.
|
| 1.3.2.2 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.3.2.1 | 16-Nov-2006 |
ad | file in_selsrc.c was added on branch newlock2 on 2006-11-18 21:39:36 +0000
|
| 1.5.26.1 | 08-Dec-2007 |
ad | Sync with head.
|
| 1.5.24.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.5.18.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.5.16.1 | 09-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.6.34.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.6.26.1 | 26-Sep-2009 |
snj | Pull up following revision(s) (requested by dyoung in ticket #937):
sys/netinet/in_selsrc.c: revision 1.7
Stop the admin from creating nodes under net.inet.ip.interfaces or
net.inet.ip.interfaces.<ifname>.
|
| 1.6.16.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.6.16.1 | 16-Sep-2009 |
yamt | sync with head
|
| 1.8.12.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.8.12.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.9.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.9.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.9.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.11.6.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.11.6.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.11.6.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.2 | 31-Aug-2015 |
ozaki-r | Hook up lltable/llentry with the kernel (and rumpkernel)
It is built and initialized on bootup, but there is no user for now.
Most codes in in.c are imported from FreeBSD as well as lltable/llentry.
|
| 1.1 | 13-Nov-2006 |
dyoung | branches: 1.1.2; 1.1.6; 1.1.8; 1.1.104; 1.1.124;
Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.1.124.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.1.104.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.8.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.8.1 | 13-Nov-2006 |
yamt | file in_selsrc.h was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.1.6.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.1.6.1 | 13-Nov-2006 |
yamt | file in_selsrc.h was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.1.2.2 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.1.2.1 | 13-Nov-2006 |
ad | file in_selsrc.h was added on branch newlock2 on 2006-11-18 21:39:36 +0000
|
| 1.14 | 28-Aug-2020 |
riastradh | netinet: Include the needful so include order doesn't matter.
|
| 1.13 | 10-Dec-2005 |
elad | Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.12 | 18-Apr-2004 |
matt | branches: 1.12.12;
De __P()
|
| 1.11 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.10 | 10-Feb-1998 |
perry | branches: 1.10.48;
add/cleanup multiple inclusion protection.
|
| 1.9 | 07-Jul-1997 |
phil | Protect against double inclusion. PR 3524.
|
| 1.8 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.7 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.6 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.5 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.4 | 08-Jan-1994 |
mycroft | More prototypes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.10.48.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.10.48.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.10.48.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.10.48.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.12.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.105 | 11-Jun-2025 |
ozaki-r | in: get rid of unused argument from ip_newid() and ip_newid_range()
|
| 1.104 | 05-Jun-2025 |
ozaki-r | Apply if_first_addr() and if_first_addr_psref()
|
| 1.103 | 19-Nov-2022 |
yamt | branches: 1.103.2; 1.103.8;
Make arp have its own mowner
This helped me to debug mbuf leaks in arp.
(if_arp.c rev. 1.298)
|
| 1.102 | 08-Mar-2021 |
christos | mv <sys/cprng.h> include to the kernel portion
|
| 1.101 | 08-Mar-2021 |
christos | reinstate a simple version of ip_randomid()
|
| 1.100 | 08-Mar-2021 |
christos | remove now unused pseudo-random ip id code.
|
| 1.99 | 08-Mar-2021 |
christos | Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
|
| 1.98 | 11-Sep-2020 |
roy | branches: 1.98.2;
inet: Add SIOCGNBRINFO to retrieve neighbor state about an address
|
| 1.97 | 29-Nov-2018 |
ozaki-r | branches: 1.97.4;
Introduce and use ip_dad_enabled() and ip6_dad_enabled() functions
|
| 1.96 | 19-Apr-2018 |
christos | branches: 1.96.2;
s/static inline/static __inline/g for consistency.
|
| 1.95 | 12-May-2017 |
ryo | branches: 1.95.2; 1.95.8;
replace in_fmtaddr() by IN_PRINT(), and delete function in_fmtaddr()
|
| 1.94 | 16-Jan-2017 |
christos | branches: 1.94.4;
really, use.
|
| 1.93 | 16-Jan-2017 |
christos | rename arplog -> ARPLOG to make it clear that it is a macro and tuck-in the
buffer used for address formatting.
|
| 1.92 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.91 | 02-Jan-2017 |
christos | branches: 1.91.2;
- You can't just call the pfil hook to remove an address before an address
is removed! Hold a reference instead, remove it, and then free it.
- GC iatoifa()
|
| 1.90 | 06-Dec-2016 |
knakahara | add API to manipulate ifa->ia_hash and ia_hash_pslist_entry, and fix ia_hash_pslist_entry race by using them.
in_ifaddr_lock is required before writing ifa->ia_hash and
ia_hash_pslist_entry to serialize writer processings.
reviewed by ozaki-r@n.o.
|
| 1.89 | 18-Nov-2016 |
knakahara | We must use PSLIST_ENTRY_DESTROY after PSLIST_WRITER_REMOVE and waiting all readers done.
And then, if we want to re-insert the removed pslist element, we need to
call PSLIST_ENTERY_INIT again.
advised by riastradh@n.o and reviewed by ozaki-r@n.o, thanks.
|
| 1.88 | 11-Oct-2016 |
roy | Implement RFC 5227 2.4 Ongoing Conflict Detection and Address Defence.
If ip_dad_count is 0, then the conflict is just logged and the address
is not marked as duplicated.
|
| 1.87 | 29-Sep-2016 |
roy | in_ifscrub is no longer needed.
|
| 1.86 | 29-Sep-2016 |
roy | Set dstaddr in in_ifinit so that sppp consumers announce the correct
dstaddr in routing messages.
|
| 1.85 | 18-Sep-2016 |
christos | Dealing with arplog is a bit more complicated...
|
| 1.84 | 17-Sep-2016 |
christos | protect arplog with INET
|
| 1.83 | 16-Sep-2016 |
roy | Drop hostIsNew from in_ifinit, let the function work out if the address
has changed.
Sync address flag setup with the IPv6 counterpart.
When scrubbing the address, or setting up the address fails, restore the
old address flags as well as the old address.
|
| 1.82 | 15-Sep-2016 |
roy | Allow arplog to be used outside of if_arp.c
|
| 1.81 | 13-Sep-2016 |
christos | remove trailing spaces. userland does not catch this?
|
| 1.80 | 13-Sep-2016 |
christos | add bits for address flags
|
| 1.79 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.78 | 08-Jul-2016 |
ozaki-r | branches: 1.78.2;
Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.77 | 08-Jul-2016 |
ozaki-r | Kill remaining use of the old lists of IP addresses
|
| 1.76 | 06-Jul-2016 |
ozaki-r | Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.75 | 06-Jul-2016 |
ozaki-r | Add and use pslist(9)-based hashtable for IPv4 addresses
Note that we leave the old hashtable to keep vmstat -H working.
|
| 1.74 | 31-Aug-2015 |
ozaki-r | Replace ARP cache (llinfo) with lltable/llentry
Highlights of the change are:
- Use llentry instead of llinfo to manage ARP caches
- ARP specific data are stored in the hashed list
of an interface instead of the global list (llinfo_arp)
- Fine-grain locking on llentry
- arptimer (callout) per ARP cache
- the global timer callout with the big locks can be
removed (though softnet_lock is still required for now)
- net.inet.arp.prune is now obsoleted
- it was the interval of the global timer callout
- net.inet.arp.refresh is now obsoleted
- it was a parameter that prevents expiration of active caches
- Removed to simplify the timer logic, but we may be able to
restore the feature if really needed
Proposed on tech-kern and tech-net.
|
| 1.73 | 31-Aug-2015 |
ozaki-r | Hook up lltable/llentry with the kernel (and rumpkernel)
It is built and initialized on bootup, but there is no user for now.
Most codes in in.c are imported from FreeBSD as well as lltable/llentry.
|
| 1.72 | 16-May-2015 |
roy | Separate ARP handling DAD from inet.
This is done by signalling the intent to try tentative addresses
and then clearing the intent once the address is setup.
When the ARP handler is installed (arp_ifinit) then it adds
dad start and stop functions to the address which are used instead
of calling ARP directly.
|
| 1.71 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.70 | 01-Jul-2014 |
rtr | branches: 1.70.4;
fix parameter types in pr_ioctl, called xx_control() functions and remove
abuse of pointer to struct mbuf type.
param2 changed to u_long type and uses parameter name 'cmd' (ioctl command)
param3 changed to void * type and uses parameter name 'data'
param4 changed to struct ifnet * and uses parameter name 'ifp'
param5 has been removed (formerly struct lwp *) and uses of 'l' have been
replaced with curlwp from curproc(9).
callers have had (now unnecessary) casts to struct mbuf * removed, called
code has had (now unnecessary) casts to u_long, void * and struct ifnet *
respectively removed.
reviewed by rmind@
|
| 1.69 | 05-Jun-2014 |
rmind | - Implement pktqueue interface for lockless IP input queue.
- Replace ipintrq and ip6intrq with the pktqueue mechanism.
- Eliminate kernel-lock from ipintr() and ip6intr().
- Some preparation work to push softnet_lock out of ipintr().
Discussed on tech-net.
|
| 1.68 | 29-May-2014 |
rmind | Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.67 | 23-May-2014 |
rmind | Make ip_input() static, there is no need to expose it.
|
| 1.66 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.65 | 05-Nov-2010 |
rmind | branches: 1.65.18; 1.65.22; 1.65.32;
ip_randomid: make mechanism MP-safe and more modular.
OK matt@
|
| 1.64 | 19-Jul-2010 |
rmind | Abstract IP reassembly into single generic routine - ip_reass_packet().
Make struct ipq private and struct ipqent not visible to userland.
Push ip_len adjustment into reassembly layer.
OK matt@
|
| 1.63 | 13-Jul-2010 |
rmind | Split-off IPv4 re-assembly mechanism into a separate module. Abstract
into ip_reass_init(), ip_reass_lookup(), etc (note: abstraction is not
yet complete). No functional changes to the actual mechanism.
OK matt@
|
| 1.62 | 28-Apr-2008 |
martin | branches: 1.62.20; 1.62.22;
Remove clause 3 and 4 from TNF licenses
|
| 1.61 | 06-Feb-2008 |
matt | branches: 1.61.6; 1.61.8; 1.61.10;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.60 | 05-Dec-2007 |
dyoung | Extract common code, creating a subroutine if_purgeaddrs(ifp,
family, purgeaddr) which applies function `purgeaddr' to each
address on `ifp' belonging to `family'.
|
| 1.59 | 01-Sep-2007 |
dyoung | branches: 1.59.6; 1.59.8;
Use ifreq_setaddr(), ifreq_getaddr(), sockaddr_in_init(), and
sockaddr_copy(). Constify. Compare pointers with NULL, not 0.
Don't "test truth" of pointers, but compare with NULL.
|
| 1.58 | 04-Mar-2007 |
christos | branches: 1.58.2; 1.58.10; 1.58.14; 1.58.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.57 | 23-Jul-2006 |
ad | branches: 1.57.10;
Use the LWP cached credentials where sane.
|
| 1.56 | 10-Dec-2005 |
elad | branches: 1.56.4; 1.56.8;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.55 | 09-Mar-2005 |
atatat | branches: 1.55.4;
Add the following nodes to the sysctl tree:
net.local.stream.pcblist
net.local.dgram.pcblist
net.inet.tcp.pcblist
net.inet.udp.pcblist
net.inet.raw.pcblist
net.inet6.tcp6.pcblist
net.inet6.udp6.pcblist
net.inet6.raw6.pcblist
which allow retrieval of the pcbs in use for those protocols. The
struct involved is 32/64 bit clean and incorporates parts of struct
inpcb, struct unpcb, a bit of struct tcpcb, and two socket addresses.
|
| 1.54 | 24-Jan-2005 |
matt | branches: 1.54.2;
Add IFNET_FOREACH and IFADDR_FOREACH macros and start using them.
|
| 1.53 | 21-Apr-2004 |
itojun | branches: 1.53.4;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.52 | 18-Apr-2004 |
matt | De __P()
|
| 1.51 | 11-Nov-2003 |
jonathan | Change global head-of-local-IP-address list from in_ifaddr to
in_ifaddrhead. Recent changes in struct names caused a namespace
collision in fast-ipsec, which are most cleanly fixed by using
"in_ifaddrhead" as the listhead name.
|
| 1.50 | 23-Oct-2003 |
mycroft | Remove all the code to maintain ia_inpcbs. This information was only used to
close sockets on address changes, which was deemed to be a bad idea and was
summarily removed, so there is no point in wasting effort on maintaining it
any more.
|
| 1.49 | 18-Aug-2003 |
itojun | since we cope with packets with addess on !IFF_UP interface in ip_input()
properly, IFF_UP check in INADDR_TO_IA is obsolete (or too much).
|
| 1.48 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.47 | 14-Jul-2003 |
itojun | correct igmp. from love
|
| 1.46 | 26-Jun-2003 |
itojun | branches: 1.46.2;
tabify
|
| 1.45 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.44 | 12-May-2002 |
matt | Eliminate commons.
|
| 1.43 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.42 | 04-Nov-2001 |
matt | Change a few variable/tables to const since they are read-only.
|
| 1.41 | 08-Oct-2000 |
enami | branches: 1.41.2; 1.41.4; 1.41.8;
- Keep track of allhost multicast address record we joined into
each in_ifaddr and delete it when an address is purged.
- Don't simply try to delete a multicast address record listed in the
ia_multiaddrs. It results a dangling pointer. Let who holds a
reference to it to delete it.
|
| 1.40 | 08-Oct-2000 |
itojun | implement multicast kludge table for IPv4.
- when all the interface address is removed from an interface, and there's
multicast groups still left joined, keep it in kludge table.
- when an interface address is added again, recover multicast groups from
kludge table.
this will avoid problem with dangling in_ifaddr on pcmcia card removal,
due to the link from multicast group info (in_multi).
the code is basically from sys/netinet6/in6.c (jinmei@kame).
pointed out by: Shiva Shenoy <shiva_s@yahoo.com>
|
| 1.39 | 30-Mar-2000 |
augustss | branches: 1.39.4;
Remove register declarations.
|
| 1.38 | 30-Mar-2000 |
simonb | Delete redundant decl of in_socktrim() - it's in <netinet/in.h>.
|
| 1.37 | 02-Feb-2000 |
thorpej | PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.36 | 01-Feb-2000 |
thorpej | First-draft if_detach() implementation, originally from Bill Studnemund,
although this version has been changed somewhat:
- reference counting on ifaddrs isn't as complete as Bill's original
work was. This is hard to get right, and we should attack one
protocol at a time.
- This doesn't do reference counting or dynamic allocation of ifnets yet.
- This version introduces a new PRU -- PRU_PURGEADDR, which is used to
purge an ifaddr from a protocol. The old method Bill used didn't work
on all protocols, and it only worked on some because it was Very Lucky.
This mostly works ... i.e. works for my USB Ethernet, except for a dangling
ifaddr reference left by the IPv6 code; have not yet tracked this down.
|
| 1.35 | 01-Jul-1999 |
itojun | branches: 1.35.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.34 | 16-May-1999 |
thorpej | Sigh, fix some broken logic in the last change to INADDR_TO_IA(), and make
the macro a little more obvious. Should fix kern/7589, from Jens A Nilsson.
|
| 1.33 | 03-May-1999 |
thorpej | In INADDR_TO_IA(), skip interfaces which are not up. Revert previous change
to ip_input.c to check the interface status after INADDR_TO_IA().
Fix cooked up by Heiko Rupp and myself.
Fixes PR 7480.
|
| 1.32 | 19-Dec-1998 |
thorpej | branches: 1.32.2;
Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.31 | 30-Sep-1998 |
tls | branches: 1.31.4;
Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.30 | 14-Aug-1998 |
scottr | Fix the NEXT_IA_WITH_SAME_ADDR macro introduced in 1.27: it was finding
the first in_ifaddr structure with a different internet address! Reverse
the sense of the test. Spotted by and fix from Eric Haszlakiewicz.
|
| 1.29 | 29-Jul-1998 |
tls | change IN_IFADDR_HASH_SIZE to 509, which actually uses no more space than 293 due to rounding up to nearest power of two in hashinit.
|
| 1.28 | 16-Jul-1998 |
tls | Put original hash function back. It wastes a little bit of space, but is much more even -- think of the case of a web service provider, some of whose customers end up getting 'inferior service' because they're on addresses that happen to be out at the end of a hash chain. With webservers with thousands of addresses, this is a real issue. If the wasted space is a big deal, we could pick a prime number that's slightly _less_ than a power of two...
|
| 1.27 | 02-Jul-1998 |
is | The rewrite of if_arp.c to work with the hashed interface address lists
(1.44) missed a test for the right interface, making some machines answer
to some bogus arp requests (like for WHO-HAS 127.0.0.1).
The quick patch in 1.46-1.47 does not work for so-called "unnumbered"
interfaces, that is, (point-to-point) interfaces that share their local
address with another (e.g., the Ethernet) interface.
We add a macro to in_var.h, to step (in the current implementation) through
the hash chain and fine more entries with the same address, and use that
in if_arp.c to find one which belongs to our interface.
|
| 1.26 | 01-Jun-1998 |
thorpej | Eek, we were wasting almost half of the in_ifaddr hash space by modulo'ing
with IN_IFADDR_HASH_SIZE. Instead, AND with the hash mask computed by
hashinit().
|
| 1.25 | 29-May-1998 |
matt | Change arp so its console log messages print out IP addresses in
dotted quad format instead of hex.
|
| 1.24 | 04-May-1998 |
matt | Default IP flow to being enabled. Add a sysctl to control the maximum
number of flows (net.inet.ip.maxflows). If set to 0, will disable fast
path forwarding.
|
| 1.23 | 29-Apr-1998 |
matt | Add support for "fast" forwarding. Add hooks in if_ethersubr.c and
if_fddisubr.c to fastpath IP forwarding. If ip_forward successfully
forwards a packet, it will create a cache (ipflow) entry. ether_input
and fddi_input will first call ipflow_fastforward with the received
packet and if the packet passes enough tests, it will be forwarded (the
ttl is decremented and the cksum is adjusted incrementally).
|
| 1.22 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.21 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.20 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.19 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.18 | 23-Jul-1997 |
thorpej | branches: 1.18.6;
Pull SYN_cache_branch down into the main line.
|
| 1.17 | 22-May-1996 |
mycroft | branches: 1.17.8;
Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.16 | 13-Feb-1996 |
christos | branches: 1.16.4;
netinet prototypes
|
| 1.15 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.14 | 04-Jun-1995 |
mycroft | Don't cast things unnecessarily.
|
| 1.13 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.12 | 31-May-1995 |
mycroft | Implement IGMP v2. Based on the Multicast 3.5 distribution.
|
| 1.11 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.10 | 29-Mar-1995 |
briggs | KERNEL -> _KERNEL
|
| 1.9 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.8 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.7 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.6 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 08-Dec-1993 |
hpeyerl | more Multicast stuff.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.16.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.17.8.1 | 14-May-1997 |
mellon | Add prototype for in_setmaxmtu()
|
| 1.18.6.1 | 01-Oct-1998 |
cgd | pull up revisions 1.21-1.22, 1.27, 1.29-1.30, 1.31 (via patch) from
trunk. (tls)
|
| 1.31.4.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.32.2.2 | 07-Jun-1999 |
perry | pullup 1.33->1.34 (thorpej): fix INADDR_TO_IA()
|
| 1.32.2.1 | 03-May-1999 |
perry | branches: 1.32.2.1.2; 1.32.2.1.4;
pullup 1.32->1.33 (thorpej)
|
| 1.32.2.1.4.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.32.2.1.4.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.32.2.1.4.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.32.2.1.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.32.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.35.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.39.4.1 | 17-Oct-2000 |
tv | Pullup 1.41 [enami]:
- Keep track of allhost multicast address record we joined into
each in_ifaddr and delete it when an address is purged.
- Don't simply try to delete a multicast address record listed in the
ia_multiaddrs. It results a dangling pointer. Let who holds a
reference to it to delete it.
Also 1.40 [itojun, req by enami]:
implement multicast kludge table for IPv4.
- when all the interface address is removed from an interface, and there's
multicast groups still left joined, keep it in kludge table.
- when an interface address is added again, recover multicast groups from
kludge table.
this will avoid problem with dangling in_ifaddr on pcmcia card removal,
due to the link from multicast group info (in_multi).
|
| 1.41.8.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.41.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.41.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.41.2.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.41.2.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.46.2.6 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.46.2.5 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.46.2.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.46.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.46.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.46.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.53.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.54.2.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.55.4.5 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.55.4.4 | 07-Dec-2007 |
yamt | sync with head
|
| 1.55.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.55.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.55.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.56.8.1 | 11-Aug-2006 |
yamt | sync with head
|
| 1.56.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.57.10.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.58.16.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.58.16.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.58.16.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.58.14.2 | 09-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.58.14.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.58.10.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.58.2.1 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.59.8.1 | 08-Dec-2007 |
ad | Sync with head.
|
| 1.59.6.2 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.59.6.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.61.10.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.61.8.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.61.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.62.22.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.62.20.1 | 06-Nov-2010 |
uebayasi | Sync with HEAD.
|
| 1.65.32.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.65.22.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.65.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.65.18.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.70.4.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.70.4.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.70.4.5 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.70.4.4 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.70.4.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.70.4.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.70.4.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.78.2.4 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.78.2.3 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.78.2.2 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.78.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.91.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.94.4.1 | 19-May-2017 |
pgoyette | Resolve conflicts from previous merge (all resulting from $NetBSD
keywork expansion)
|
| 1.95.8.2 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.95.8.1 | 22-Apr-2018 |
pgoyette | Sync with HEAD
|
| 1.95.2.1 | 09-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1662):
sys/netinet/tcp_subr.c: revision 1.286
sys/netinet/tcp_timer.c: revision 1.96
sys/netinet/in_var.h: revision 1.102
sys/netinet/in_var.h: revision 1.99
Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
Add some randomness to the iss offset
Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
mv <sys/cprng.h> include to the kernel portion
|
| 1.96.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.97.4.1 | 09-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1229):
sys/netinet/tcp_subr.c: revision 1.286
sys/netinet/tcp_timer.c: revision 1.96
sys/netinet/in_var.h: revision 1.102
sys/netinet/in_var.h: revision 1.99
Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
Add some randomness to the iss offset
Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
mv <sys/cprng.h> include to the kernel portion
|
| 1.98.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.103.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.103.2.2 | 01-Oct-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1164):
sys/net/link_proto.c: revision 1.41
sys/netinet6/in6.c: revision 1.293
sys/net/if.h: revision 1.307
sys/netinet/ip_icmp.c: revision 1.180
sys/dev/vmt/vmt_subr.c: revision 1.11
sys/netinet6/in6_var.h: revision 1.105
sys/netinet6/in6_var.h: revision 1.106
sys/net/if.c: revision 1.532
sys/net/if.c: revision 1.533
sys/netinet6/mld6.c: revision 1.102
sys/netinet/in_var.h: revision 1.104
sys/net/if_spppsubr.c: revision 1.270
sys/net/if_spppsubr.c: revision 1.271
sys/netinet6/nd6.c: revision 1.284
if: introduce if_first_addr() (and psref variant)
It returns a first address (ifa) of a given family on a given interface.
It will replace a bunch of open codes and make their intent clear.
Apply if_first_addr() and if_first_addr_psref()
in6: introduce in6ifa_first_lladdr() (and psref variant)
It returns a first link-local address (ifa) on a given interface.
Apply in6ifa_first_lladdr() and in6ifa_first_lladdr_psref()
|
| 1.103.2.1 | 29-Jul-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1140):
sys/netinet/ip_output.c: revision 1.330
sys/netinet/sctp_output.c: revision 1.39
sys/netinet/ip_mroute.c: revision 1.166
sys/netipsec/ipsecif.c: revision 1.24
sys/netipsec/xform_ipip.c: revision 1.80
sys/netinet/ip_output.c: revision 1.327
sys/netinet/ip_output.c: revision 1.328
sys/netinet/ip_input.c: revision 1.406
sys/netinet/ip_output.c: revision 1.329
sys/netinet/in_var.h: revision 1.105
in: get rid of unused argument from ip_newid() and ip_newid_range()
in: take a reference of ifp on IP_ROUTETOIF
The ifp could be released after ia4_release(ia).
in: narrow the scope of ifa in ip_output (NFC)
sctp: follow the recent change of ip_newid()
in: avoid racy ifa_acquire(rt->rt_ifa) in ip_output()
If a rtentry is being destroyed asynchronously, ifa referenced by rt_ifa
can be destructed and taking ifa_acquire(rt->rt_ifa) aborts with a
KASSERT failure. Fortunately, the ifa is not actually freed because of
a reference by rt_ifa, it can be available (except some functions like
psref) so as long the rtentry is held.
PR kern/59527
in: avoid racy ia4_acquire(ifatoia(rt->rt_ifa) in ip_rtaddr()
Same as the case of ip_output(), it's racy and should be avoided.
PR kern/59527
|
| 1.39 | 17-Apr-2022 |
andvar | fix various typos in comments.
|
| 1.38 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.37 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.36 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.35 | 28-Aug-2020 |
riastradh | branches: 1.35.2;
netinet: Include the needful so include order doesn't matter.
|
| 1.34 | 02-Nov-2012 |
christos | fix typo
|
| 1.33 | 02-Nov-2012 |
christos | make this standalone, like every others (except OpenBSD)
|
| 1.32 | 24-Jul-2011 |
christos | branches: 1.32.2; 1.32.8; 1.32.12; 1.32.14;
Fill in missing IPTOS defines (from Linux/OpenBSD)
|
| 1.31 | 25-Dec-2007 |
perry | Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.30 | 21-Dec-2007 |
matt | Add fix for ip_id information leakage. Since the leakage information is
primarily used with TCP SYN and RST packets and such packets are less than
the smallest sized packet that an IP stack is allowed to fragment, we simply
set ip_id to 0 for all packets 68 bytes or less.
|
| 1.29 | 17-Dec-2006 |
christos | branches: 1.29.20; 1.29.26; 1.29.28; 1.29.32;
According to ANSI c the only portably defined bitfields are unsigned int ones.
|
| 1.28 | 05-Sep-2006 |
rpaulo | branches: 1.28.2; 1.28.4;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.27 | 10-Dec-2005 |
elad | branches: 1.27.4; 1.27.8;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.26 | 25-Apr-2004 |
jonathan | branches: 1.26.12;
Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.25 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.24 | 01-Apr-2003 |
dogcow | branches: 1.24.2;
bring into conformance with RFC 3514
|
| 1.23 | 05-Mar-2002 |
itojun | bring in latest ALTQ from kjc. ALTQify some of the drivers.
|
| 1.22 | 24-Oct-2001 |
itojun | it may fix PR14124.
|
| 1.21 | 02-May-2000 |
sommerfeld | branches: 1.21.6; 1.21.8; 1.21.12;
One more __attribute__((__packed__)) to dissuade egcs from making
unwarranted asumptions about the structure's alignment.
|
| 1.20 | 20-Nov-1999 |
thorpej | Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.19 | 01-Jul-1999 |
itojun | branches: 1.19.2; 1.19.8;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.18 | 10-Feb-1998 |
perry | branches: 1.18.8; 1.18.10; 1.18.12;
add/cleanup multiple inclusion protection.
|
| 1.17 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.16 | 11-Dec-1996 |
mycroft | Minor change to a comment.
|
| 1.15 | 25-Oct-1996 |
thorpej | Make length and offset fields unsigned. From Kevin M. Lahey <kml@nas.nasa.gov>
|
| 1.14 | 21-Sep-1996 |
perry | commit fix in pr 2772 -- the IP input code was assuming that the
reserved (must be zero) flag must necessarily be zero. We now define
an IP_RF (by analogy to IP_DF and IP_MF) and mask it out when necessary.
|
| 1.13 | 14-Sep-1996 |
mrg | move the packet filter hooks in to a saner location. while i'm here, rename
PACKET_FILTER to PFIL_HOOKS.
|
| 1.12 | 12-Sep-1996 |
mrg | forward decl. struct mbuf (for now).
|
| 1.11 | 12-Sep-1996 |
explorer | Move an #ifdef _KERNEL up above all the packet filter stuff. This
could very well break the packet filter stuff, but it will make things
like rcp.c compile, and rcp.c should not need to include sys/mbuf.h
to do so...
|
| 1.10 | 06-Sep-1996 |
mrg | add packet filter interface code. see pfil(9) for more details. you
need the PACKET_FILTER option to enable this code. currently, ipfilter
version 3.1.1-beta has been converted to use this new interface.
|
| 1.9 | 15-May-1995 |
cgd | branches: 1.9.6;
"routine" precedence has a value of 0.
|
| 1.8 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.7 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.6 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.5 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.4 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.9.6.2 | 11-Dec-1996 |
mycroft | From trunk:
Ignore the reserved fragment flag when checking ip_off.
|
| 1.9.6.1 | 10-Nov-1996 |
thorpej | Update from trunk:
- Make ip_len and ip_off unsigned.
- Make sure we don't accept or transmit packets larger than the
maximim IP packet size.
This fixes the so-called `death ping' bug.
Sum of work from Bill Fenner <fenner@parc.xerox.com>,
Kevin Lahey <kml@nas.nasa.gov>, and myself.
Thanks to Curt Sampson, Jukka Marin, and Kevin Lahey for testing
this under NetBSD 1.2
|
| 1.18.12.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.18.12.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.18.10.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.18.8.1 | 05-May-2000 |
cgd | Pull up revisions 1.20-1.21 (requested by sommerfeld):
Add "__attribute__((__packed__))" to structures used to describe
on-the-wire data, to prevent egcs from making unwarranted assumptions
about the alignment of these structures.
|
| 1.19.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.19.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.21.12.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.21.8.2 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.21.8.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.21.6.2 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.21.6.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.24.2.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.24.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.24.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.24.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.26.12.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.26.12.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.26.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.27.8.1 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.27.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.28.4.1 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.28.2.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.29.32.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.29.28.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.29.26.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.29.20.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.32.14.1 | 17-Aug-2017 |
martin | Pull up following revision(s) (requested by mrg in ticket #721):
include/resolv.h: revision 1.40
sys/netinet/ip.h: revision 1.33-1.34
fix typo
make this standalone, like every others (except OpenBSD)
add <netinet/in.h> because it is needed for sockaddr_in.
|
| 1.32.12.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.32.8.1 | 28-Nov-2012 |
riz | Pull up following revision(s) (requested by christos in ticket #721):
include/resolv.h: revision 1.40
sys/netinet/ip.h: revision 1.33
sys/netinet/ip.h: revision 1.34
fix typo
make this standalone, like every others (except OpenBSD)
add <netinet/in.h> because it is needed for sockaddr_in.
|
| 1.32.2.1 | 16-Jan-2013 |
yamt | sync with (a bit old) head
|
| 1.35.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.30 | 07-Mar-2021 |
christos | netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
|
| 1.29 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.28 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.27 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.26 | 27-Jul-2020 |
roy | branches: 1.26.2;
ip6: Remove __packed attribute from ip6 structures
They should naturally align.
Add compile time assertations to ip6_input.c to prove this.
|
| 1.25 | 18-May-2018 |
maxv | branches: 1.25.6;
IP6_EXTHDR_GET performs a basic mbuf operation, which has nothing to do
with IPv6. So declare an IP-independent M_REGION_GET, and make
IP6_EXTHDR_GET an alias to it.
|
| 1.24 | 18-May-2018 |
maxv | Remove IP6_EXTHDR_GET0, remove pointless XXXs, and style.
|
| 1.23 | 25-Dec-2007 |
perry | branches: 1.23.2; 1.23.90; 1.23.96;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.22 | 04-Mar-2007 |
christos | branches: 1.22.16; 1.22.22; 1.22.24; 1.22.28;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.21 | 05-May-2006 |
rpaulo | branches: 1.21.14;
Add support for RFC 3542 Adv. Socket API for IPv6 (which obsoletes 2292).
* RFC 3542 isn't binary compatible with RFC 2292.
* RFC 2292 support is on by default but can be disabled.
* update ping6, telnet and traceroute6 to the new API.
From the KAME project (www.kame.net).
Reviewed by core.
|
| 1.20 | 10-Dec-2005 |
elad | branches: 1.20.4; 1.20.6; 1.20.8; 1.20.10; 1.20.12;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.19 | 29-May-2005 |
christos | branches: 1.19.2;
- add const
- remove bogus casts
- avoid nested variables
|
| 1.18 | 09-Jul-2004 |
itojun | typo. Bruno Rohee
|
| 1.17 | 26-Apr-2004 |
itojun | declare ip6_hdr_pseudo (for kernel only) and use it for TCP MD5 signature
|
| 1.16 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.15 | 06-Jun-2003 |
itojun | branches: 1.15.2;
- sync up MLD declaration with RFC3542 (s/MLD6/MLD/)
- routing header declaration with RFC3542
(note: sizeof(ip6_rthdr0) has changed!)
also, sync up with RFC2460 routing header definition (no "strict" source
routing mode any more)
part of advanced API update (RFC2292 -> 3542).
|
| 1.14 | 14-May-2003 |
itojun | always use PULLDOWN_TEST codepath.
|
| 1.13 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.12 | 05-Jul-2001 |
itojun | IP6_EXTHDR_GET0 had no check against m->m_len (noone was using this macro).
sync with kame
|
| 1.11 | 23-Jan-2001 |
itojun | branches: 1.11.2;
put attribute(packed) for ip6 option headers. they will appear at
strange alignment positions. sync with kame
|
| 1.10 | 10-Oct-2000 |
itojun | sync with kame ($KAME$)
|
| 1.9 | 13-Jul-2000 |
itojun | remove m_pulldown statistics code. it is highly experimental and belong
to kame tree only (not for *bsd).
|
| 1.8 | 02-Jul-2000 |
itojun | typo in previous
|
| 1.7 | 02-Jul-2000 |
itojun | do not touch struct ip6stat on non-INET6 compilation.
From: Paul Goyette <paul@whooppee.com>
|
| 1.6 | 03-Mar-2000 |
itojun | branches: 1.6.4;
comment fix, sync with kame.
|
| 1.5 | 24-Feb-2000 |
itojun | hide declaration of IP6_EXTHDR_{GET,CHECK} from userland.
|
| 1.4 | 06-Feb-2000 |
itojun | to be more rfc2292 complient, move ip6.h and icmp6.h into netinet.
(netinet6/{ip6,icmp6}.h is non-standard path - these files should go away)
it was not possible to use cvsmove in this case.
when you try to look at history, chase it toward netinet6/{ip6,icmp6}.h.
|
| 1.3 | 03-Jul-1999 |
thorpej | branches: 1.3.2;
RCS ID police.
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 28-Jun-1999 |
itojun | branches: 1.1.2;
file ip6.h was initially added on branch kame.
|
| 1.1.2.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 01-Jul-1999 |
thorpej | file ip6.h was added on branch chs-ubc2 on 1999-07-01 23:47:01 +0000
|
| 1.3.2.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.3.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.6.4.3 | 14-Jul-2000 |
itojun | pullup (approved by releng-1-5)
remove m_pulldown statistics code. it is highly experimental and belong
to kame tree only (not for *bsd).
1.4 -> 1.5 syssrc/sys/kern/uipc_mbuf2.c
1.8 -> 1.9 syssrc/sys/netinet/ip6.h
1.13 -> 1.14 syssrc/sys/netinet6/ip6_var.h
|
| 1.6.4.2 | 03-Jul-2000 |
thorpej | Pull up rev. 1.8:
typo in previous
|
| 1.6.4.1 | 03-Jul-2000 |
thorpej | Pull up rev. 1.7:
do not touch struct ip6stat on non-INET6 compilation.
From: Paul Goyette <paul@whooppee.com>
|
| 1.11.2.2 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.11.2.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.15.2.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.15.2.4 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.15.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.15.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.15.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.19.2.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.19.2.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.19.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.20.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.20.10.1 | 11-May-2006 |
elad | sync with head
|
| 1.20.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.20.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.20.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.21.14.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.22.28.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.22.24.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.22.22.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.22.16.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.23.96.1 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.23.90.1 | 07-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1661):
sys/netinet6/ip6_id.c: revision 1.19-1.21
sys/netinet6/ip6_var.h: revision 1.88
sys/netinet/ip_input.c: revision 1.400
sys/netinet/tcp_subr.c: revision 1.285
sys/netinet/ip6.h: revision 1.30
netinet: Enable random IP fragment ids by default (from riastradh)
netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
netinet6: Mark randomid unused.
Will make merging and bisection easier if anything goes wrong with
flow label or fragment id randomization changes.
(from riastradh)
netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
Replace randomid() by cprng_fast32()
|
| 1.23.2.1 | 22-Feb-2008 |
keiichi | imported Mobile IPv6 code developed by the SHISA project
(http://www.mobileip.jp/).
|
| 1.25.6.1 | 07-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1226):
sys/netinet6/ip6_id.c: revision 1.19-1.21
sys/netinet6/ip6_var.h: revision 1.88
sys/netinet/ip_input.c: revision 1.400
sys/netinet/tcp_subr.c: revision 1.285
sys/netinet/ip6.h: revision 1.30
netinet: Enable random IP fragment ids by default (from riastradh)
netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
netinet6: Mark randomid unused.
Will make merging and bisection easier if anything goes wrong with
flow label or fragment id randomization changes.
(from riastradh)
netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
Replace randomid() by cprng_fast32()
|
| 1.26.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.1 | 22-Feb-2008 |
keiichi | branches: 1.1.2;
file ip6mh.h was initially added on branch keiichi-mipv6.
|
| 1.1.2.1 | 22-Feb-2008 |
keiichi | imported Mobile IPv6 code developed by the SHISA project
(http://www.mobileip.jp/).
|
| 1.35 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.34 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.33 | 28-Mar-2004 |
martti | branches: 1.33.2;
Upgraded IPFilter to 4.1.1
|
| 1.32 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.31 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.30 | 19-Sep-2002 |
martti | branches: 1.30.6;
Resync with official IPF
|
| 1.29 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.28 | 09-Jun-2002 |
itojun | whitespace
|
| 1.27 | 02-May-2002 |
martti | branches: 1.27.2; 1.27.4;
Fix compilation problems
|
| 1.26 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.25 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.24 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.23 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.22 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.21 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.20 | 28-Sep-2001 |
chs | don't depend on other headers to include sys/proc.h for us.
|
| 1.19 | 26-Mar-2001 |
mike | branches: 1.19.2; 1.19.4;
Resolve conflicts.
|
| 1.18 | 09-Aug-2000 |
veego | branches: 1.18.2;
Resolve conflicts.
|
| 1.17 | 23-May-2000 |
veego | branches: 1.17.4;
Resolve conflicts.
|
| 1.16 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.15 | 16-Apr-2000 |
chs | remove ifdefs to skip htons() on some big-endian platforms.
|
| 1.14 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.13 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.12 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.11 | 02-Feb-1999 |
cjs | branches: 1.11.2; 1.11.6; 1.11.8; 1.11.14;
Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.10 | 19-Jan-1999 |
mycroft | There's just no plausible reason to byte-swap ip_id internally. It's opaque.
|
| 1.9 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.8 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.7 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.6 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.5 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.4 | 21-Sep-1997 |
veego | branches: 1.4.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.3 | 07-Jul-1997 |
fvdl | branches: 1.3.2;
Get rid of (void) cast to KFREE, as it may be a macro, in which case
the cast will be a syntax error.
|
| 1.2 | 06-Jul-1997 |
thorpej | The sheer number of #ifdef's around it should have been a hint that
#include <machine/mtpr.h> isn't something you're supposed to do in
NetBSD.
|
| 1.1 | 06-Jul-1997 |
thorpej | branches: 1.1.1;
Initial revision
|
| 1.1.1.19 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.18 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.17 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.16 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.15 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.14 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.13 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.12 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.11 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.10 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.9 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.8 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.7 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.6 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.5 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.4 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.3 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.2 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.1 | 06-Jul-1997 |
thorpej | Import yet another missing piece of IPFilter 3.2beta1.
|
| 1.3.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.4.2.3 | 24-Nov-1998 |
cgd | pull up rev(s) 1.9 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.4.2.2 | 22-Jul-1998 |
mellon | Pull up 1.8
|
| 1.4.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.11.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.11.8.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.11.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.11.6.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.11.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.11.2.1 | 20-Dec-1999 |
he | Pull up revision 1.12 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.17.4.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.17.4.2 | 09-Feb-2002 |
he | Pull up revisions 1.19-1.24 (requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.17.4.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.18.2.9 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.18.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.18.2.7 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.18.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.18.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.18.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.18.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.18.2.2 | 08-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.18.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.19.4.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.19.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.19.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.19.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.19.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.27.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.27.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.30.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.30.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.33.2.1 | 13-Aug-2004 |
jmc | branches: 1.33.2.1.2;
Pullup rev 1.34 (requested by christos in ticket #1727)
Sync up w. ipf 4.1.3
|
| 1.33.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.35 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.12 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.11 | 28-Mar-2004 |
martti | branches: 1.11.4;
Upgraded IPFilter to 4.1.1
|
| 1.10 | 24-Jan-2002 |
martti | branches: 1.10.16;
Re-sync with IPFilter
|
| 1.9 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.8 | 26-Mar-2001 |
mike | branches: 1.8.2;
Resolve conflicts.
|
| 1.7 | 23-May-2000 |
veego | branches: 1.7.4; 1.7.6;
Resolve conflicts.
|
| 1.6 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.5 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.4 | 22-Nov-1998 |
mrg | branches: 1.4.4; 1.4.10; 1.4.16;
merge ipf 3.2.10
|
| 1.3 | 13-Sep-1998 |
christos | Fix copyright spacing and 'Van' -> 'van' for consistency.
|
| 1.2 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.1 | 06-Jul-1997 |
thorpej | branches: 1.1.1;
Initial revision
|
| 1.1.1.13 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.12 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.11 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.10 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.9 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.8 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.7 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.6 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.5 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.4 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.3 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.2 | 21-Sep-1997 |
veego | branches: 1.1.1.2.2;
Import ip-filter 3.2beta5
|
| 1.1.1.1 | 06-Jul-1997 |
thorpej | branches: 1.1.1.1.2;
Import ip_auth.h from IPFilter 3.2beta1; this was missed during the
upgrade.
|
| 1.1.1.2.2.3 | 24-Nov-1998 |
cgd | pull up rev(s) 1.4 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.1.1.2.2.2 | 22-Jul-1998 |
mellon | Pull up 1.2 (veego)
|
| 1.1.1.2.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.1.1.1.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.4.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.4.10.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.4.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.4.4.1 | 20-Dec-1999 |
he | Pull up revision 1.5 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.7.6.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.7.6.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.7.4.1 | 09-Feb-2002 |
he | Pull up revisions 1.8-1.10 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.8.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.10.16.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.10.16.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.10.16.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.10.16.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.11.4.1 | 06-Feb-2005 |
jmc | Pull up revision 1.12 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.121 | 20-Dec-2024 |
rin | carp_join_multicast: Stop allocating ip_moptions on stack
It was just a waste of memory. NFC otherwise, and no regression
observed for full ATF run on amd64.
Partially taken from OpenBSD:
https://github.com/openbsd/src/commit/1f237790b75
XXX
Seems like OpenBSD has some fixes to carp, that may improve our
implementation...
|
| 1.120 | 01-Aug-2023 |
mrg | branches: 1.120.6;
fix simple mis-matched function prototype and definitions.
most of these are like, eg
void foo(int[2]);
with either of these
void foo(int*) { ... }
void foo(int[]) { ... }
in some cases (such as stat or utimes* calls found in our header files),
we now match standard definition from opengroup.
found by GCC 12.
|
| 1.119 | 07-Apr-2023 |
mlelstv | Select virtual address as sender if backing interface is anonymous.
Use correct scope for IPv6.
|
| 1.118 | 26-Mar-2023 |
mlelstv | Use backing device to send advertisements. Otherwise the packets originate
from the virtual MAC address, which confuses switches.
|
| 1.117 | 02-Sep-2022 |
thorpej | branches: 1.117.4;
Remove unnecessary inclusion of <net/netisr.h>.
|
| 1.116 | 30-Sep-2021 |
yamaguchi | carp: Register carp_carpdev_state to link-state change hook
|
| 1.115 | 16-Jun-2021 |
riastradh | if_attach and if_initialize cannot fail, don't test return value
These were originally made failable back in 2017 when if_initialize
allocated a softint in every interface for link state changes, so
that it could fail gracefully instead of panicking:
https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html
However, this spawned many seldom- or never-tested error branches,
which are risky to have around. And that softint in every interface
has since been replaced by a single global workqueue, because link
state changes require thread context but not low latency or high
throughput:
https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html
So there is no longer any reason for if_initialize to fail. (The
subroutine if_stats_init can't fail because percpu_alloc can't fail
either.)
There is a snag: the softint_establish in if_percpuq_create could
fail, potentially leading to bad consequences later on trying to use
the softint. This change doesn't introduce any new bugs because of
the snag -- if_percpuq_attach was already broken. However, the snag
can be better addressed without spawning error branches, either by
using a single softint or making softints less scarce.
(Separate commit will change the signatures of if_attach and
if_initialize to return void, scheduled to ride whatever is the next
convenient kernel bump.)
Patch and testing on amd64 and evbmips64-eb by maya@; commit message
soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
|
| 1.114 | 14-Oct-2020 |
roy | branches: 1.114.6;
carp: Don't set a link level address if vhid == -1
Link level address for carp is dervied from vhid.
Until vhid is set, carp is useless, so don't give it a link level address
until a vhid is set.
This fixes recent test case breakage where carp was fixed to actually
print the ethernet address set by default. Note that neither carp nor
the test case itself was actually broken as the error is the common
ATF net code assuming that a cloned interface's link level address is
unique upon creation.
|
| 1.113 | 12-Oct-2020 |
roy | carp: link state is DOWN until it becomes a MASTER
This is consitent with other BSD's handling of CARP and means
we don't have to carry a custom flag for it.
|
| 1.112 | 12-Oct-2020 |
roy | carp: Set ethernet address just before interface registation
Otherwise ifconfig reports SIOCGLIFADDR errors.
|
| 1.111 | 09-Oct-2020 |
roy | carp: Remove media, software should use link status.
carp literally has no media just like ppp, vlan, etc.
|
| 1.110 | 06-Feb-2020 |
thorpej | Perform link state change processing on a work queue, rather than in a
softint.
|
| 1.109 | 04-Feb-2020 |
thorpej | Use ifmedia_fini().
|
| 1.108 | 29-Jan-2020 |
thorpej | Adopt <net/if_stats.h>.
|
| 1.107 | 20-Jan-2020 |
thorpej | Remove FDDI support.
|
| 1.106 | 19-Jan-2020 |
thorpej | Remove Token Ring support.
|
| 1.105 | 16-Jan-2020 |
kardel | Provide SIOCGIFMEDIA ioctl to deliver link status.
Add link0 (IFF_LINK0) flag to map INIT state to LINK_STATE_DOWN
instead of LINK_STATE_UNKNOWN. This allows routing software to
suppress routes to the interface of the carp interface when in
init state (e. g. link down in the parent interface).
|
| 1.104 | 10-Nov-2019 |
chs | branches: 1.104.2;
in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT
and remove code to handle failures that can no longer happen.
|
| 1.103 | 01-Jun-2019 |
joerg | Define carp6_cksum only when it is used, that is under INET6
|
| 1.102 | 14-Mar-2019 |
ozaki-r | carp: don't skip pserialize_read_enter and ifa_release
|
| 1.101 | 22-Dec-2018 |
maxv | Replace M_ALIGN and MH_ALIGN by m_align.
|
| 1.100 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.99 | 26-Jun-2018 |
msaitoh | branches: 1.99.2;
Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward
compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same
as FreeBSD.
This change also fixes a bug that the direction is misunderstand on some
environment by passing the direction to bpf_mtap*() instead of checking
m->m_pkthdr.rcvif.
|
| 1.98 | 14-Jun-2018 |
yamaguchi | Add the lock to refer the list included in ethercom for safety
The lock is already held while adding and deleting
ok ozaki-r@
|
| 1.97 | 14-Jun-2018 |
yamaguchi | Use ether_lookup_multi() instead of the macro
ok ozaki-r@
|
| 1.96 | 18-May-2018 |
maxv | IP6_EXTHDR_GET -> M_REGION_GET, no functional change.
|
| 1.95 | 21-Mar-2018 |
maxv | Fix an untriggerable memory leak. carp_prepare_ad does not fail, so switch
it to void.
|
| 1.94 | 06-Dec-2017 |
ozaki-r | branches: 1.94.2;
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
|
| 1.93 | 22-Nov-2017 |
ozaki-r | Protect IFADDR_READER_FOREACH and obtained ifa with psz/psref
|
| 1.92 | 16-Nov-2017 |
ozaki-r | Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
|
| 1.91 | 23-Oct-2017 |
msaitoh | If if_initialize() failed in the attach function, free resources and return.
|
| 1.90 | 19-May-2017 |
ozaki-r | branches: 1.90.2;
Allow CARP to call the link_state_change handler immediately
If the handler is delayed because of the indirection call via softint,
some operations are executed in reverse and may cause unexpected
behaviors. For example, due to the issue a GARP packet wasn't sent on
a transition from the BACKUP state to the MASTER state; this happened
because IN_IFF_DETACHED flag wasn't cleared on arpannounce, which
had been cleared in the link_state_change handler.
This fixes an issue reported by sborrill@ on tech-net:
http://mail-index.netbsd.org/tech-net/2017/03/14/msg006283.html
|
| 1.89 | 12-May-2017 |
ryo | replace in_fmtaddr() by IN_PRINT(), and delete function in_fmtaddr()
|
| 1.88 | 12-May-2017 |
roy | carp should call if_link_state_change instead of affecting
if_link_state directly.
|
| 1.87 | 19-Apr-2017 |
ozaki-r | branches: 1.87.2;
Fix build without INET6
|
| 1.86 | 14-Mar-2017 |
ozaki-r | Use if_acquire and if_release instead of using psref API directly
- Provide if_release for consistency to if_acquire
- Use if_acquire and if_release for ifp iterations
- Make ifnet_psref_class static
|
| 1.85 | 27-Feb-2017 |
ozaki-r | Make CARP on IPv6 work
It passes ATF tests but no more, no less.
|
| 1.84 | 02-Feb-2017 |
ozaki-r | Defer some pr_input to workqueue
pr_input is currently called in softint. Some pr_input such as ICMP, ICMPv6
and CARP can add/delete/update IP addresses and routing table entries. For
example, icmp6_redirect_input updates an a routing table entry and
nd6_ra_input may delete an IP address.
Basically such operations shouldn't be done in softint. That aside, we have
a reason to avoid the situation; psz/psref waits cannot be used in softint,
however they are required to work in such pr_input in the MP-safe world.
The change implements the workqueue pr_input framework called wqinput which
provides a means to defer pr_input of a protocol to workqueue easily.
Currently icmp_input, icmp6_input, carp_proto_input and carp6_proto_input
are deferred to workqueue by the framework.
Proposed and discussed on tech-kern and tech-net
|
| 1.83 | 16-Jan-2017 |
christos | ip6_sprintf -> IN6_PRINT so that we pass the size.
|
| 1.82 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.81 | 28-Dec-2016 |
ozaki-r | branches: 1.81.2;
Use ether_ifattach in carp_clone_create instead of C&P code
carp_clone_destroy calls ether_ifdetach so not calling ether_ifattach is
inconsistent. If we add something pair of initialization and destruction
to ether_ifattach and ether_ifdetach (e.g., mutex_init/mutex_destroy),
ether_ifdetach of carp_clone_destroy won't work. So use ether_ifattach.
In order to do so, make ether_ifattach accept the 2nd argument (lla) as
NULL to allow carp to initialize its link level address by itself.
|
| 1.80 | 12-Dec-2016 |
ozaki-r | Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.79 | 11-Oct-2016 |
roy | Remove unused variable.
|
| 1.78 | 11-Oct-2016 |
roy | Mark arprequest static and introduce arpannounce so that gratuitous
ARP requests are only send from valid addresses.
|
| 1.77 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.76 | 23-Jul-2016 |
is | Print the IPv6 or IPv4 source addresses of packets with wrong hash, to
help debugging.
|
| 1.75 | 23-Jul-2016 |
is | Workaround for PR 47013 by bouyer@. Only works for mixed IPv4/IPv6
environemnts, not for pure-IPv6 yet. A real fix is still needed.
|
| 1.74 | 07-Jul-2016 |
ozaki-r | branches: 1.74.2;
Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.73 | 06-Jul-2016 |
ozaki-r | Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.72 | 04-Jul-2016 |
ozaki-r | Use pslist(9) for the global in6_ifaddr list
psz and psref will be applied in another commit.
No functional change intended.
|
| 1.71 | 21-Jun-2016 |
ozaki-r | Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.70 | 20-Jun-2016 |
knakahara | fix: i386/ALL build failure
|
| 1.69 | 20-Jun-2016 |
knakahara | apply if_output_lock() to L3 callers which call ifp->if_output() of L2(or L3 tunneling).
|
| 1.68 | 16-Jun-2016 |
ozaki-r | Use curlwp_bind and curlwp_bindx instead of open-coding LP_BOUND
|
| 1.67 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.66 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.65 | 12-May-2016 |
ozaki-r | Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.64 | 28-Apr-2016 |
ozaki-r | Constify rtentry of if_output
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
|
| 1.63 | 04-Apr-2016 |
ozaki-r | Separate nexthop caches from the routing table
By this change, nexthop caches (IP-MAC address pair) are not stored
in the routing table anymore. Instead nexthop caches are stored in
each network interface; we already have lltable/llentry data structure
for this purpose. This change also obsoletes the concept of cloning/cloned
routes. Cloned routes no longer exist while cloning routes still exist
with renamed to connected routes.
Noticeable changes are:
- Nexthop caches aren't listed in route show/netstat -r
- sysctl(NET_RT_DUMP) doesn't return them
- If RTF_LLDATA is specified, it returns nexthop caches
- Several definitions of routing flags and messages are removed
- RTF_CLONING, RTF_XRESOLVE, RTF_LLINFO, RTF_CLONED and RTM_RESOLVE
- RTF_CONNECTED is added
- It has the same value of RTF_CLONING for backward compatibility
- route's -xresolve, -[no]cloned and -llinfo options are removed
- -[no]cloning remains because it seems there are users
- -[no]connected is introduced and recommended
to be used instead of -[no]cloning
- route show/netstat -r drops some flags
- 'L' and 'c' are not seen anymore
- 'C' now indicates a connected route
- Gateway value of a route of an interface address is now not
a L2 address but "link#N" like a connected (cloning) route
- Proxy ARP: "arp -s ... pub" doesn't create a route
You can know details of behavior changes by seeing diffs under tests/.
Proposed on tech-net and tech-kern:
http://mail-index.netbsd.org/tech-net/2016/03/11/msg005701.html
|
| 1.62 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.61 | 20-Aug-2015 |
christos | include "ioconf.h" to get the 'void <driver>attach(int count);' prototype.
|
| 1.60 | 26-Feb-2015 |
roy | Introduce the routing flag RTF_LOCAL to track local address routes.
Add functions rt_ifa_addlocal() and rt_ifa_remlocal() to add and remove
local routes for the address and announce the new address and route
to the routing socket.
Add in_ifaddlocal() and in_ifremlocal() to use these functions.
Rename in6_if{add,rem}loop() to in6_if{add,rem}local() and use these
functions.
rtinit() no longer announces the address, just the network route for the
address. As such, calls to rt_newaddrmsg() have been removed from
in_addprefix() and in_scrubprefix().
This solves the problem of potentially more than one announcement, or no
announcement at all for the address in certain situations.
|
| 1.59 | 31-Jul-2014 |
ozaki-r | branches: 1.59.2; 1.59.4;
Make carp_suppress_preempt global back
It is still accessed by if_pfsync.c.
This unbreaks the build of i386 kernel.
|
| 1.58 | 31-Jul-2014 |
ozaki-r | Make local functions/variables static
No functional change.
|
| 1.57 | 06-Jun-2014 |
rmind | - Eliminate RTFREE() macro in favour of rtfree() function.
- Make rtcache() function static.
|
| 1.56 | 29-May-2014 |
rmind | Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.55 | 17-May-2014 |
rmind | - Move IFNET_*() macros under #ifdef _KERNEL.
- Replace TAILQ_FOREACH on ifnet with IFNET_FOREACH().
|
| 1.54 | 13-May-2014 |
bouyer | Make sure *(if_output)() is called with KERNEL_LOCK held.
Add some KASSERT for this.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details.
|
| 1.53 | 04-Apr-2014 |
bouyer | branches: 1.53.2;
Proper MBUFTRACE handling. Without it, ec_tx_mowner, ec_rx_mowner and
ifp->if_mowner would be used uninitialised.
|
| 1.52 | 25-Feb-2014 |
pooka | Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.51 | 18-Oct-2013 |
christos | remove unused variable
|
| 1.50 | 20-Aug-2012 |
christos | branches: 1.50.2; 1.50.4;
make this compile.
|
| 1.49 | 20-Aug-2012 |
bouyer | Support checksum offloading in carp(4) if the underlying device suports it,
as proposed on tech-net@ on 2 Aug 2012.
|
| 1.48 | 27-Mar-2012 |
bouyer | Do not sleep in callout context, this will hang the clock soft interrupt.
Should fix PR kern/46217.
|
| 1.47 | 19-Nov-2011 |
tls | branches: 1.47.2; 1.47.4;
First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>. This change includes
the following:
An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are
fixed. Some effort is made to accumulate entropy more quickly at
boot time.
A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.
The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.
An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c). This
is based on Greg Rose's implementation from Qualcomm.
A new random stream generator, nist_ctr_drbg, is provided. It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim. This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.
An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness. The arc4random/arc4randbytes API is deprecated for
in-kernel use. It is replaced by "cprng_strong". The current
cprng_fast implementation wraps the existing arc4random
implementation. The current cprng_strong implementation wraps the
new CTR_DRBG implementation. Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.
In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data(). Performance is expected to improve.
The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.
The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot. There is
approximately a 3/10000 chance of a false positive from these
tests. Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.
A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32). A new kernel will require a new rndctl.
The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.
Manual pages for the new kernel interfaces are forthcoming.
|
| 1.46 | 19-Oct-2011 |
dyoung | branches: 1.46.2;
Use if_addr_init() and if_mcast_op() instead of ifp->if_ioctl().
|
| 1.45 | 17-Jul-2011 |
joerg | Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.44 | 11-Aug-2010 |
pooka | ahem, min -> max in previous
|
| 1.43 | 11-Aug-2010 |
pooka | Use kpause() instead of DELAY() and sleep a minimum of 1 tick.
This is possible now since softints have a thread context. It's
also not a very frequent code path. Addresses ABI issue with delay
(kern/40505).
I'm not entire sure what this delay is meant to accomplish, though.
|
| 1.42 | 10-Aug-2010 |
pooka | Include opt_inet since this checks INET/INET6
|
| 1.41 | 05-Apr-2010 |
joerg | Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf
check into the inline functions as well the fourth argument for
bpf_attach.
|
| 1.40 | 19-Jan-2010 |
pooka | branches: 1.40.2; 1.40.4;
Redefine bpf linkage through an always present op vector, i.e.
#if NBPFILTER is no longer required in the client. This change
doesn't yet add support for loading bpf as a module, since drivers
can register before bpf is attached. However, callers of bpf can
now be modularized.
Dynamically loadable bpf could probably be done fairly easily with
coordination from the stub driver and the real driver by registering
attachments in the stub before the real driver is loaded and doing
a handoff. ... and I'm not going to ponder the depths of unload
here.
Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
|
| 1.39 | 16-Sep-2009 |
pooka | Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.38 | 07-Jun-2009 |
taca | Make ip_carp.c compile, fixing usage of CARP_LOG().
|
| 1.37 | 27-May-2009 |
christos | PR/38260: Brian Seklecki: Improve carp logging.
|
| 1.36 | 12-May-2009 |
elad | Fix previous, || -> &&.
Pointed out by cube@, thanks!
|
| 1.35 | 12-May-2009 |
elad | Fix inverted permissions check.
|
| 1.34 | 18-Apr-2009 |
tsutsui | Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.33 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.32 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.31 | 18-Mar-2009 |
cegger | bcmp -> memcmp
|
| 1.30 | 11-Jan-2009 |
christos | branches: 1.30.2;
merge christos-time_t
|
| 1.29 | 19-Dec-2008 |
cegger | use M_ZERO on malloc() and remove subsequent bzero().
|
| 1.28 | 17-Dec-2008 |
cegger | kill MALLOC and FREE macros.
|
| 1.27 | 07-Nov-2008 |
dyoung | *** Summary ***
When a link-layer address changes (e.g., ifconfig ex0 link
02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor
Advertisement to update the network-/link-layer address bindings
on our LAN peers.
Refuse a change of ethernet address to the address 00:00:00:00:00:00
or to any multicast/broadcast address. (Thanks matt@.)
Reorder ifnet ioctl operations so that driver ioctls may inherit
the functions of their "class"---ether_ioctl(), fddi_ioctl(), et
cetera---and the class ioctls may inherit from the generic ioctl,
ifioctl_common(), but both driver- and class-ioctls may override
the generic behavior. Make network drivers share more code.
Distinguish a "factory" link-layer address from others for the
purposes of both protecting that address from deletion and computing
EUI64.
Return consistent, appropriate error codes from network drivers.
Improve readability. KNF.
*** Details ***
In if_attach(), always initialize the interface ioctl routine,
ifnet->if_ioctl, if the driver has not already initialized it.
Delete if_ioctl == NULL tests everywhere else, because it cannot
happen.
In the ioctl routines of network interfaces, inherit common ioctl
behaviors by calling either ifioctl_common() or whichever ioctl
routine is appropriate for the class of interface---e.g., ether_ioctl()
for ethernets.
Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In
the user->kernel interface, SIOCSIFADDR's argument was an ifreq,
but on the protocol->ifnet interface, SIOCSIFADDR's argument was
an ifaddr. That was confusing, and it would work against me as I
make it possible for a network interface to overload most ioctls.
On the protocol->ifnet interface, replace SIOCSIFADDR with
SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to
invoke SIOCINITIFADDR.
In ifioctl(), give the interface the first shot at handling most
interface ioctls, and give the protocol the second shot, instead
of the other way around. Finally, let compatibility code (COMPAT_OSOCK)
take a shot.
Pull device initialization out of switch statements under
SIOCINITIFADDR. For example, pull ..._init() out of any switch
statement that looks like this:
switch (...->sa_family) {
case ...:
..._init();
...
break;
...
default:
..._init();
...
break;
}
Rewrite many if-else clauses that handle all permutations of IFF_UP
and IFF_RUNNING to use a switch statement,
switch (x & (IFF_UP|IFF_RUNNING)) {
case 0:
...
break;
case IFF_RUNNING:
...
break;
case IFF_UP:
...
break;
case IFF_UP|IFF_RUNNING:
...
break;
}
unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and
#ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4).
In ipw(4), remove an if_set_sadl() call that is out of place.
In nfe(4), reuse the jumbo MTU logic in ether_ioctl().
Let ethernets register a callback for setting h/w state such as
promiscuous mode and the multicast filter in accord with a change
in the if_flags: ether_set_ifflags_cb() registers a callback that
returns ENETRESET if the caller should reset the ethernet by calling
if_init(), 0 on success, != 0 on failure. Pull common code from
ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(),
and register if_flags callbacks for those drivers.
Return ENOTTY instead of EINVAL for inappropriate ioctls. In
zyd(4), use ENXIO instead of ENOTTY to indicate that the device is
not any longer attached.
Add to if_set_sadl() a boolean 'factory' argument that indicates
whether a link-layer address was assigned by the factory or some
other source. In a comment, recommend using the factory address
for generating an EUI64, and update in6_get_hw_ifid() to prefer a
factory address to any other link-layer address.
Add a routing message, RTM_LLINFO_UPD, that tells protocols to
update the binding of network-layer addresses to link-layer addresses.
Implement this message in IPv4 and IPv6 by sending a gratuitous
ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD
messages on a change of an interface's link-layer address.
In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address
that is broadcast/multicast or equal to 00:00:00:00:00:00.
Make ether_ioctl() call ifioctl_common() to handle ioctls that it
does not understand.
In gif(4), initialize if_softc and use it, instead of assuming that
the gif_softc and ifp overlap.
Let ifioctl_common() handle SIOCGIFADDR.
Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels
that certain invariants on a struct route are satisfied.
In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit
about the ioctls that we do not allow on an agr(4) member interface.
bzero -> memset. Delete unnecessary casts to void *. Use
sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with
NULL instead of "testing truth". Replace some instances of (type
*)0 with NULL. Change some K&R prototypes to ANSI C, and join
lines.
|
| 1.26 | 04-May-2008 |
thorpej | branches: 1.26.6; 1.26.8; 1.26.10; 1.26.14;
Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.25 | 23-Apr-2008 |
thorpej | branches: 1.25.2;
Use <net/net_stats.h> / netstat_sysctl().
|
| 1.24 | 15-Apr-2008 |
thorpej | branches: 1.24.2;
Make CARP status per-cpu.
|
| 1.23 | 15-Mar-2008 |
ws | branches: 1.23.2;
Set scope on IPv6 multicast address to give carp a chance to work for IPv6, too.
From FreeBSD.
|
| 1.22 | 21-Dec-2007 |
matt | branches: 1.22.2; 1.22.6;
Add fix for ip_id information leakage. Since the leakage information is
primarily used with TCP SYN and RST packets and such packets are less than
the smallest sized packet that an IP stack is allowed to fragment, we simply
set ip_id to 0 for all packets 68 bytes or less.
|
| 1.21 | 20-Dec-2007 |
dyoung | Constify struct ifnet->if_sadl and every use throughout the tree.
Add if_set_sadl() that both sets the link-layer address length and
replaces the current link-layer address with a new one, and use it
throughout the tree.
|
| 1.20 | 11-Dec-2007 |
lukem | use __KERNEL_RCSID()
|
| 1.19 | 04-Dec-2007 |
dyoung | branches: 1.19.2; 1.19.4;
Use IFADDR_FOREACH().
|
| 1.18 | 19-Oct-2007 |
ad | branches: 1.18.2; 1.18.4;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
|
| 1.17 | 19-Sep-2007 |
dyoung | branches: 1.17.4;
Constify sockaddr argument to ether_multiaddr(). Change struct
ifreq * arguments to ether_addmulti() and ether_delmulti() to const
struct sockaddr *, since ether_{add,del}multi() only ever read the
sockaddr ifreq member, ifr_addr. Update uses in carp(4) and in
vlan(4).
|
| 1.16 | 30-Aug-2007 |
dyoung | Use malloc(9) for sockaddrs instead of pool(9), and remove dom_sa_pool
and dom_sa_len members from struct domain. Pools of fixed-size
objects are too rigid for sockaddr_dls, whose size can vary over
a wide range.
Return sockaddr_dl to its "historical" size. Now that I'm using
malloc(9) instead of pool(9) to allocate sockaddr_dl, I can create
a sockaddr_dl of any size in the kernel, so expanding sockaddr_dl
is useless.
Avoid using sizeof(struct sockaddr_dl) in the kernel.
Introduce sockaddr_dl_alloc() for allocating & initializing an
arbitrary sockaddr_dl on the heap.
Add an argument, the sockaddr length, to sockaddr_alloc(),
sockaddr_copy(), and sockaddr_dl_setaddr().
Constify: LLADDR() -> CLLADDR().
Where the kernel overwrites LLADDR(), use sockaddr_dl_setaddr(),
instead. Used properly, sockaddr_dl_setaddr() will not overrun
the end of the sockaddr.
|
| 1.15 | 26-Aug-2007 |
dyoung | branches: 1.15.2;
Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to
identify sockaddr_dl abuse that remains in the kernel, especially
the potential for overwriting memory past the end of a sockaddr_dl
with, e.g., memcpy(LLADDR(), ...).
Use sockaddr_dl_setaddr() in a few places.
|
| 1.14 | 19-Jul-2007 |
dyoung | branches: 1.14.4; 1.14.6;
Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.13 | 09-Jul-2007 |
ad | branches: 1.13.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.12 | 04-Mar-2007 |
christos | branches: 1.12.2; 1.12.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.11 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.10 | 16-Nov-2006 |
christos | branches: 1.10.4;
__unused removal on arguments; approved by core.
|
| 1.9 | 30-Oct-2006 |
christos | Fix typo (hi Elad)
|
| 1.8 | 25-Oct-2006 |
elad | Kill some KAUTH_GENERIC_ISSUSER.
|
| 1.7 | 20-Oct-2006 |
liamjfoy | Remove some dead code - From OpenBSD Rev. 1.129
|
| 1.6 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.5 | 23-Jul-2006 |
ad | branches: 1.5.4; 1.5.6; 1.5.8;
Use the LWP cached credentials where sane.
|
| 1.4 | 13-Jun-2006 |
riz | branches: 1.4.4;
Remove implementation of tvtohz() - since the timecounters branch
was merged, this is now in sys/kern/kern_clock.c .
|
| 1.3 | 25-May-2006 |
liamjfoy | branches: 1.3.2;
remove a little white space
|
| 1.2 | 24-May-2006 |
liamjfoy | branches: 1.2.2;
Add a check for our own advertisements. This is due to non-simplex
interfaces which received the packets they have just sent.
From: OpenBSD (rev. 1.124)
ok: christos@
|
| 1.1 | 18-May-2006 |
liamjfoy | branches: 1.1.2; 1.1.4;
Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.1.4.4 | 11-Aug-2006 |
yamt | sync with head
|
| 1.1.4.3 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.1.4.2 | 24-May-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 18-May-2006 |
yamt | file ip_carp.c was added on branch yamt-pdpolicy on 2006-05-24 10:59:03 +0000
|
| 1.1.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.2.2.2 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.2.2.1 | 24-May-2006 |
tron | file ip_carp.c was added on branch peter-altq on 2006-05-24 15:50:44 +0000
|
| 1.3.2.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.3.2.1 | 25-May-2006 |
kardel | file ip_carp.c was added on branch simonb-timecounters on 2006-06-01 22:38:46 +0000
|
| 1.4.4.9 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.4.4.8 | 21-Jan-2008 |
yamt | sync with head
|
| 1.4.4.7 | 07-Dec-2007 |
yamt | sync with head
|
| 1.4.4.6 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.4.4.5 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.4.4.4 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.4.4.3 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.4.4.2 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.4.4.1 | 13-Jun-2006 |
yamt | file ip_carp.c was added on branch yamt-lazymbuf on 2006-06-21 15:11:01 +0000
|
| 1.5.8.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.5.8.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.5.6.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.5.6.1 | 23-Jul-2006 |
rpaulo | file ip_carp.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:58:47 +0000
|
| 1.5.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.10.4.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.10.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.12.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.12.2.4 | 23-Oct-2007 |
ad | Sync with head.
|
| 1.12.2.3 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.12.2.2 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.12.2.1 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.13.2.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.13.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.14.6.2 | 19-Jul-2007 |
dyoung | Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.14.6.1 | 19-Jul-2007 |
dyoung | file ip_carp.c was added on branch matt-mips64 on 2007-07-19 20:48:55 +0000
|
| 1.14.4.4 | 09-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.14.4.3 | 26-Oct-2007 |
joerg | Sync with HEAD.
Follow the merge of pmap.c on i386 and amd64 and move
pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup
code to restore CR4 before jumping back into kernel space as the large
page option might cover that.
|
| 1.14.4.2 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.14.4.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.15.2.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.15.2.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.15.2.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.17.4.1 | 25-Oct-2007 |
bouyer | Sync with HEAD.
|
| 1.18.4.2 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.18.4.1 | 08-Dec-2007 |
ad | Sync with head.
|
| 1.18.2.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.18.2.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.19.4.2 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.19.4.1 | 13-Dec-2007 |
bouyer | Sync with HEAD
|
| 1.19.2.1 | 11-Dec-2007 |
yamt | sync with head.
|
| 1.22.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.22.6.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.22.6.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.22.2.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.23.2.4 | 27-Dec-2008 |
christos | merge with head.
|
| 1.23.2.3 | 09-Nov-2008 |
christos | merge with head.
|
| 1.23.2.2 | 01-Nov-2008 |
christos | Sync with head.
|
| 1.23.2.1 | 29-Mar-2008 |
christos | Welcome to the time_t=long long dev_t=uint64_t branch.
|
| 1.24.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.25.2.7 | 09-Oct-2010 |
yamt | sync with head
|
| 1.25.2.6 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.25.2.5 | 11-Mar-2010 |
yamt | sync with head
|
| 1.25.2.4 | 20-Jun-2009 |
yamt | sync with head
|
| 1.25.2.3 | 16-May-2009 |
yamt | sync with head
|
| 1.25.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.25.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.26.14.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.26.10.2 | 09-Jun-2009 |
snj | Pull up following revision(s) (requested by taca in ticket #796):
sys/netinet/ip_carp.c: revision 1.38
Make ip_carp.c compile, fixing usage of CARP_LOG().
|
| 1.26.10.1 | 05-Jun-2009 |
snj | Pull up following revision(s) (requested by christos in ticket #785):
sys/netinet/ip_carp.c: revision 1.37
PR/38260: Brian Seklecki: Improve carp logging.
|
| 1.26.8.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.26.8.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.26.6.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.30.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.30.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.40.4.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.40.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.40.2.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.40.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.46.2.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.46.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.46.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.47.4.5 | 28-Aug-2016 |
bouyer | Pull up following revision(s) (requested by is in ticket #1393):
sys/netinet/ip_carp.c: revision 1.75
Workaround for PR 47013 by bouyer@. Only works for mixed IPv4/IPv6
environemnts, not for pure-IPv6 yet. A real fix is still needed.
|
| 1.47.4.4 | 27-Aug-2016 |
bouyer | Pull up following revision(s) (requested by is in ticket #1394):
sys/netinet/ip_carp.c: revision 1.76
Print the IPv6 or IPv4 source addresses of packets with wrong hash, to
help debugging.
|
| 1.47.4.3 | 03-Jun-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1067):
sys/dist/ipf/netinet/ip_fil_netbsd.c 1.9 via patch
sys/net/if_ethersubr.c 1.197 via patch
sys/net/if_loop.c 1.77 via patch
sys/net/if_vlan.c 1.70 via patch
sys/netinet/if_arp.c 1.158
sys/netinet/ip_carp.c 1.54 via patch
sys/netinet6/ip6_flow.c 1.23 via patch
sys/netinet6/nd6.c 1.150 via patch
sys/rump/librump/rumpkern/klock.c 1.4
Make sure *(if_output)() is called with KERNEL_LOCK held to avoid mbuf leak.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details. For netinet6, the problem report, fix and test were done
by njoly@ on current-users@
|
| 1.47.4.2 | 11-Apr-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1043):
sys/netinet/ip_carp.c: revision 1.53
Proper MBUFTRACE handling. Without it, ec_tx_mowner, ec_rx_mowner and
ifp->if_mowner would be used uninitialised.
|
| 1.47.4.1 | 02-Apr-2012 |
riz | branches: 1.47.4.1.4; 1.47.4.1.6;
Pull up following revision(s) (requested by bouyer in ticket #145):
sys/netinet/ip_carp.c: revision 1.48
Do not sleep in callout context, this will hang the clock soft interrupt.
Should fix PR kern/46217.
|
| 1.47.4.1.6.1 | 18-Jun-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1067):
sys/dist/ipf/netinet/ip_fil_netbsd.c 1.9 via patch
sys/net/if_ethersubr.c 1.197 via patch
sys/net/if_loop.c 1.77 via patch
sys/net/if_vlan.c 1.70 via patch
sys/netinet/if_arp.c 1.158
sys/netinet/ip_carp.c 1.54 via patch
sys/netinet6/ip6_flow.c 1.23 via patch
sys/netinet6/nd6.c 1.150 via patch
sys/rump/librump/rumpkern/klock.c 1.4
Make sure *(if_output)() is called with KERNEL_LOCK held to avoid mbuf leak.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details. For netinet6, the problem report, fix and test were done
by njoly@ on current-users@
|
| 1.47.4.1.4.1 | 18-Jun-2014 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1067):
sys/dist/ipf/netinet/ip_fil_netbsd.c 1.9 via patch
sys/net/if_ethersubr.c 1.197 via patch
sys/net/if_loop.c 1.77 via patch
sys/net/if_vlan.c 1.70 via patch
sys/netinet/if_arp.c 1.158
sys/netinet/ip_carp.c 1.54 via patch
sys/netinet6/ip6_flow.c 1.23 via patch
sys/netinet6/nd6.c 1.150 via patch
sys/rump/librump/rumpkern/klock.c 1.4
Make sure *(if_output)() is called with KERNEL_LOCK held to avoid mbuf leak.
See http://mail-index.netbsd.org/tech-net/2014/04/09/msg004511.html
for details. For netinet6, the problem report, fix and test were done
by njoly@ on current-users@
|
| 1.47.2.1 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.50.4.2 | 18-May-2014 |
rmind | sync with head
|
| 1.50.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.50.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.50.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.53.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.59.4.9 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.59.4.8 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.59.4.7 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.59.4.6 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.59.4.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.59.4.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.59.4.3 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.59.4.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.59.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.59.2.5 | 12-May-2017 |
sborrill | Pull up the following revisions(s) (requested by roy in ticket #1420):
sys/netinet/ip_carp.c: revision 1.88
carp should call if_link_state_change instead of affecting
if_link_state directly.
|
| 1.59.2.4 | 27-Aug-2016 |
snj | Pull up following revision(s) (requested by is in ticket #1209):
sys/netinet/ip_carp.c: revision 1.76
Print the IPv6 or IPv4 source addresses of packets with wrong hash, to
help debugging.
|
| 1.59.2.3 | 27-Aug-2016 |
snj | Pull up following revision(s) (requested by is in ticket #1208):
sys/netinet/ip_carp.c: revision 1.75
Workaround for PR 47013 by bouyer@. Only works for mixed IPv4/IPv6
environemnts, not for pure-IPv6 yet. A real fix is still needed.
|
| 1.59.2.2 | 23-Jul-2016 |
is | backout last change (wrong branch).
|
| 1.59.2.1 | 23-Jul-2016 |
is | Log the IPv4/IPv6 source of incorrect hash packets, too. Needed for
meaningful debugging.
|
| 1.74.2.6 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.74.2.5 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.74.2.4 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.74.2.3 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.74.2.2 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.74.2.1 | 26-Jul-2016 |
pgoyette | Sync with HEAD
|
| 1.81.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.87.2.1 | 19-May-2017 |
pgoyette | Resolve conflicts from previous merge (all resulting from $NetBSD
keywork expansion)
|
| 1.90.2.4 | 19-Mar-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1218):
sys/netinet/ip_carp.c: revision 1.102
carp: don't skip pserialize_read_enter and ifa_release
|
| 1.90.2.3 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.90.2.2 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by msaitoh in ticket #427):
sys/arch/amiga/dev/if_bah_zbus.c: 1.17
sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30
sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32
sys/arch/usermode/dev/if_veth.c: 1.9
sys/dev/ic/an.c: 1.66
sys/dev/ic/athn.c: 1.17
sys/dev/ic/atw.c: 1.162
sys/dev/ic/bwi.c: 1.33
sys/dev/ic/dwc_gmac.c: 1.41-1.42
sys/dev/ic/malo.c: 1.10
sys/dev/ic/rt2560.c: 1.31
sys/dev/ic/rt2661.c: 1.36
sys/dev/ic/rt2860.c: 1.29
sys/dev/ic/rtw.c: 1.127
sys/dev/ic/rtwvar.h: 1.46
sys/dev/ic/smc90cx6.c: 1.71
sys/dev/ic/smc90cx6var.h: 1.12
sys/dev/ic/wi.c: 1.244
sys/dev/pci/if_ipw.c: 1.66
sys/dev/pci/if_iwi.c: 1.104
sys/dev/pci/if_iwm.c: 1.76
sys/dev/pci/if_iwn.c: 1.86
sys/dev/pci/if_rtwn.c: 1.13
sys/dev/pci/if_wm.c: 1.541
sys/dev/pci/if_wpi.c: 1.79
sys/dev/pci/ixgbe/ixgbe.c: 1.106
sys/dev/pci/ixgbe/ixv.c: 1.73 via patch
sys/dev/pcmcia/if_malo_pcmcia.c: 1.15
sys/dev/scsipi/if_se.c: 1.95
sys/dev/usb/if_upl.c: 1.60
sys/net/if.c: 1.396
sys/net/if.h: 1.241
sys/net/if_arc.h: 1.23
sys/net/if_arcsubr.c: 1.78
sys/net/if_bridge.c: 1.136-1.137
sys/net/if_etherip.c: 1.39
sys/net/if_faith.c: 1.56
sys/net/if_gif.c: 1.131
sys/net/if_loop.c: 1.96
sys/net/if_mpls.c: 1.30
sys/net/if_pppoe.c: 1.129
sys/net/if_srt.c: 1.27
sys/net/if_stf.c: 1.102
sys/net/if_tap.c: 1.100
sys/net/if_vlan.c: 1.105
sys/netinet/ip_carp.c: 1.91
sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74
sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56
if_initalize() and if_attach() failed when resource allocation failed
(e.g. allocating softint). Without this change, it panics. It's bad because
resource shortage really occured when a lot of pseudo interface is created.
To avoid this problem, don't panic and change return value of if_initialize()
and if_attach() to int. Caller fanction will be recover from error cleanly by
checking the return value.
Return if bah_attach_subr() failed.
If if_attach() failed in the attach function, return.
- If if_initialize() failed in the attach function, free resources and return.
- Add some missing frees in bridge_clone_destroy().
- KNF
If error occured in bcmeth_ccb_attach(), free resources and return.
If error occured in pq3etsec_attach(), free resources and return.
If error occured in the attach function, free resources and return.
- If if_initialize() failed in athn_attach(), free resources and return.
- Add missing pmf_event_deregister() in athn_detach().
- Free resources correctly on some errors in atw_attach().
- Use apint*() insread of printf() in the attach function.
If if_initialize() failed in the attach function, return.
- If if_initialize() failed in the attach function, free resources and return.
- Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach
failed.
- If if_initialize() failed in the attach function, free resources and return.
- ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc.
- If if_initialize() failed in the attach function, free resources and return.
- Fix error path in the attach function correctly.
If if_initialize() failed in the attach function, free resources and return.
If if_attach() failed in the attach function, free resources and return.
- If if_initialize() failed in the attach function, free resources and return.
- KNF
- If if_attach() failed in the attach function, free resources and return.
- KNF
Fix compile error.
Fix compile error.
We don't need '&mii', but just 'mii' for mii_detach().
Don't free sc_rthash twice
|
| 1.90.2.1 | 30-Nov-2017 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #407):
sys/compat/linux32/common/linux32_socket.c: revision 1.28
sys/net/if.c: revision 1.400
sys/netipsec/key.c: revision 1.243
sys/compat/linux/common/linux_socket.c: revision 1.139
sys/netinet/ip_carp.c: revision 1.93
sys/netinet6/in6.c: revision 1.252
sys/netinet6/in6.c: revision 1.253
sys/netinet6/in6.c: revision 1.254
sys/net/if_spppsubr.c: revision 1.173
sys/net/if_spppsubr.c: revision 1.174
sys/compat/common/uipc_syscalls_40.c: revision 1.14
Protect IFADDR_READER_FOREACH and obtained ifa with psz/psref
Fix usage of FOREACH macro
key_sad.lock is held there so SAVLIST_WRITER_FOREACH is enough.
Protect IFADDR_READER_FOREACH and obtained ifa with psz/psref
Protect IFADDR_READER_FOREACH and obtained ifa with psz/psref (more)
Fix and make consistent of usages of psz/psref in ifconf variants
Remove unnecessary goto because there is no cleanup code to share (NFC)
Tweak a condition; we don't need to care ifacount to be negative
Fix a race condition of in6_ifinit
in6_ifinit checks the number of IPv6 addresses on a given interface and
if it's zero (i.e., an IPv6 address being assigned to the interface
is the first one), call if_addr_init. However, the actual assignment of
the address (ifa_insert) is out of in6_ifinit. The check and the
assignment must be done atomically.
Fix it by holding in6_ifaddr_lock during in6_ifinit and ifa_insert.
And also add missing pserialize to IFADDR_READER_FOREACH.
|
| 1.94.2.6 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.94.2.5 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.94.2.4 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.94.2.3 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.94.2.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.94.2.1 | 22-Mar-2018 |
pgoyette | Synch with HEAD, resolve conflicts
|
| 1.99.2.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.99.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.99.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.104.2.3 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.104.2.2 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.104.2.1 | 17-Jan-2020 |
ad | Sync with head.
|
| 1.114.6.1 | 17-Jun-2021 |
thorpej | Sync w/ HEAD.
|
| 1.117.4.1 | 21-Sep-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #902):
sbin/ifconfig/carp.c: revision 1.15
sbin/ifconfig/ifconfig.8: revision 1.125
tests/net/carp/t_basic.sh: revision 1.9
sys/netinet/ip_carp.c: revision 1.118
sys/netinet/ip_carp.c: revision 1.119
Fix parser for carp state.
The state values are uppercase words INIT, BACKUP and MASTER.
Use backing device to send advertisements. Otherwise the packets originate
from the virtual MAC address, which confuses switches.
Select virtual address as sender if backing interface is anonymous.
Use correct scope for IPv6.
Don't expect the net/carp/t_basic/carp_handover_ipv6_halt_nocarpdevip
and carp_handover_ipv6_ifdown_nocarpdevip test cases to fail. At
least on the TNF i386 and amd64 testbeds, they pass more often than
not since the commit of src/sys/netinet/ip_carp.c 1.119 by mlelstv on
2023.04.07.06.44.08.
|
| 1.120.6.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.14 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.13 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.12 | 12-Oct-2020 |
roy | branches: 1.12.2;
carp: link state is DOWN until it becomes a MASTER
This is consitent with other BSD's handling of CARP and means
we don't have to carry a custom flag for it.
|
| 1.11 | 16-Jan-2020 |
kardel | Provide SIOCGIFMEDIA ioctl to deliver link status.
Add link0 (IFF_LINK0) flag to map INIT state to LINK_STATE_DOWN
instead of LINK_STATE_UNKNOWN. This allows routing software to
suppress routes to the interface of the carp interface when in
init state (e. g. link down in the parent interface).
|
| 1.10 | 14-Sep-2018 |
maxv | branches: 1.10.6;
Use non-variadic function pointer in protosw::pr_input.
|
| 1.9 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.8 | 28-Apr-2016 |
ozaki-r | branches: 1.8.16; 1.8.18;
Constify rtentry of if_output
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
|
| 1.7 | 31-Jul-2014 |
ozaki-r | branches: 1.7.4;
KNF
|
| 1.6 | 16-Sep-2009 |
pooka | branches: 1.6.22; 1.6.36;
Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.5 | 16-Apr-2008 |
dyoung | branches: 1.5.4;
C99 does not allow u_int8_t bitfields, so use unsigned int, instead.
|
| 1.4 | 15-Apr-2008 |
thorpej | Make CARP status per-cpu.
|
| 1.3 | 17-Feb-2007 |
dyoung | branches: 1.3.38;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.2 | 13-Jun-2006 |
riz | branches: 1.2.4; 1.2.10; 1.2.16;
Prototype for tvtohz() is no longer needed here.
|
| 1.1 | 18-May-2006 |
liamjfoy | branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8;
Integrate Common Address Redundancy Procotol (CARP) from OpenBSD
'pseudo-device carp'
Thanks to: joerg@ christos@ riz@ and others who tested
Ok: core@
|
| 1.1.8.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.1.8.1 | 18-May-2006 |
kardel | file ip_carp.h was added on branch simonb-timecounters on 2006-06-01 22:38:47 +0000
|
| 1.1.6.2 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.1.6.1 | 18-May-2006 |
tron | file ip_carp.h was added on branch peter-altq on 2006-05-24 15:50:45 +0000
|
| 1.1.4.3 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.1.4.2 | 24-May-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 18-May-2006 |
yamt | file ip_carp.h was added on branch yamt-pdpolicy on 2006-05-24 10:59:03 +0000
|
| 1.1.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.2.16.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.2.10.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.2.10.1 | 13-Jun-2006 |
rpaulo | file ip_carp.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:58:47 +0000
|
| 1.2.4.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.2.4.2 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.2.4.1 | 13-Jun-2006 |
yamt | file ip_carp.h was added on branch yamt-lazymbuf on 2006-06-21 15:11:01 +0000
|
| 1.3.38.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.5.4.1 | 11-Mar-2010 |
yamt | sync with head
|
| 1.6.36.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.6.22.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.6.22.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.7.4.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.8.18.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.8.18.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.8.16.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.8.16.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.10.6.1 | 17-Jan-2020 |
ad | Sync with head.
|
| 1.12.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.40 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.39 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.38 | 09-May-2004 |
christos | PR/25441: Matthew Green: IP-Filter uses M_TEMP when it already has M_IPFILTER
|
| 1.37 | 31-Mar-2004 |
dyoung | Only #define COPYIN copyin, et cetera, in the kernel. That is, only
when when _KERNEL is defined.
|
| 1.36 | 31-Mar-2004 |
darrenr | COPYIN/COPYOUT macros need to call copyin/out on NetBSD rather than just use
bcopy.
|
| 1.35 | 28-Mar-2004 |
martti | branches: 1.35.2;
Upgraded IPFilter to 4.1.1
|
| 1.34 | 26-Jun-2003 |
itojun | branches: 1.34.2;
tabify
|
| 1.33 | 05-Mar-2003 |
ragge | vax -> __vax__. Didn't I fix this a year ago?
|
| 1.32 | 24-Feb-2003 |
thorpej | Comment out the inclusion of <uvm/uvm_extern.h> -- the header is not
necessary.
|
| 1.31 | 19-Sep-2002 |
martti | Resync with official IPF
|
| 1.30 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.29 | 09-Jun-2002 |
itojun | whitespace
|
| 1.28 | 02-May-2002 |
martti | branches: 1.28.2; 1.28.4;
Upgraded IPFilter to 3.4.27
|
| 1.27 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.26 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.25 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.24 | 26-May-2001 |
ragge | branches: 1.24.2;
defined(vax) -> defined(__vax__). This may fix PR#12919.
|
| 1.23 | 12-Apr-2001 |
thorpej | Delete SPL_IMP(). It is not used in IP Filter, and it aids me
on my quest to eliminate the foul beast known as splimp.
|
| 1.22 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.21 | 05-Feb-2001 |
chs | branches: 1.21.2;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.20 | 28-Jun-2000 |
mrg | remove include of <vm/vm.h>
|
| 1.19 | 26-Jun-2000 |
mrg | remove/move more mach vm header files:
<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>
also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.
|
| 1.18 | 03-May-2000 |
veego | branches: 1.18.4;
Resolve conflicts.
|
| 1.17 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.16 | 22-Nov-1998 |
mrg | branches: 1.16.4; 1.16.10; 1.16.16;
merge ipf 3.2.10
|
| 1.15 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.14 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.13 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.12 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.11 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.10 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.9 | 21-Sep-1997 |
veego | branches: 1.9.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.8 | 06-Jul-1997 |
thorpej | branches: 1.8.2;
Restore original RCS IDs.
|
| 1.7 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.6 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.5 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.4 | 15-Apr-1997 |
christos | Fix SPLNET() conditional to work for NetBSD1_0+, not just the named versions.
|
| 1.3 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.21 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.20 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.19 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.18 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.17 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.16 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.15 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.14 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.8.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.16 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.9.2.3 | 22-Jul-1998 |
mellon | Pull up 1.15 (veego)
|
| 1.9.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.9.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.16.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.16.10.4 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.16.10.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.16.10.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.16.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.16.4.1 | 20-Dec-1999 |
he | Pull up revision 1.17 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.18.4.2 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.18.4.1 | 09-Feb-2002 |
he | Pull up revisions 1.19-1.26 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.21.2.7 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.21.2.6 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.21.2.5 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.21.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.21.2.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.21.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.21.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.24.2.3 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.24.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.24.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.28.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.28.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.34.2.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.34.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.35.2.4 | 13-Aug-2004 |
jmc | branches: 1.35.2.4.2;
Pullup rev 1.39 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.35.2.3 | 30-May-2004 |
tron | Pull up revision 1.38 (requested by christos in ticket #416):
PR/25441: Matthew Green: IP-Filter uses M_TEMP when it already has M_IPFILTER
|
| 1.35.2.2 | 01-Apr-2004 |
jmc | Pullup rev 1.37 (requested by mrg in ticket #34)
Only #define COPYIN copyin, et cetera, in the kernel.
|
| 1.35.2.1 | 31-Mar-2004 |
tron | Pull up revision 1.36 (requested by mrg in ticket #32):
COPYIN/COPYOUT macros need to call copyin/out on NetBSD rather than just use
bcopy.
|
| 1.35.2.4.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.40 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.17 | 09-Dec-2017 |
pgoyette | Split ip_ecn code into its own module, so it can be shared between
gif(4), stf(4), and ipsec(4). Without this, loading the if_gif
module can result in redefined global symbols if either ipsec(4) or
stf(4) but not gif(4) is built into the kernel.
Fixes PR kern/52795 (as reported by martin@ via irc).
XXX pullup to netbsd-8
|
| 1.16 | 24-Aug-2015 |
pooka | branches: 1.16.10;
sprinkle _KERNEL_OPT
|
| 1.15 | 05-Sep-2006 |
rpaulo | branches: 1.15.102; 1.15.122;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.14 | 11-Dec-2005 |
christos | branches: 1.14.4; 1.14.8;
merge ktrace-lwp.
|
| 1.13 | 03-Feb-2005 |
perry | branches: 1.13.6;
ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.12 | 10-Apr-2002 |
itojun | branches: 1.12.10; 1.12.18; 1.12.20;
correct variable initialization. reported by fujitsu folks
|
| 1.11 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.10 | 10-May-2001 |
itojun | branches: 1.10.2;
correct ecn consideration on tunnel encap/decap. sync with kame.
|
| 1.9 | 02-Oct-2000 |
itojun | branches: 1.9.2;
fix compilation without INET. fix confusion between ipsecstat and ipsec6stat.
sync with kame.
|
| 1.8 | 06-Jan-2000 |
itojun | remove too much portability code in KAME, to improve readability.
|
| 1.7 | 12-Dec-1999 |
itojun | sync with latest KAME (rcsid only).
|
| 1.6 | 31-Jul-1999 |
itojun | branches: 1.6.2; 1.6.8;
sync with recent KAME.
- loosen ipsec restriction on packet diredtion.
- revise icmp6 redirect handling on IsRouter bit.
- tcp/udp notification processing (link-local address case)
- cosmetic fixes (better code share across *BSD).
|
| 1.5 | 30-Jul-1999 |
itojun | remove reference to in6_systm.h (file itself will be removed afterwords)
|
| 1.4 | 06-Jul-1999 |
itojun | sync with KAME/NetBSD 1.4, SNAP kit 19990705.
key changes are:
- icmp6 redirect fix (dst check)
- revised ip6 multicast check for loopback i/f
- several RCS ID cleanups
|
| 1.3 | 03-Jul-1999 |
thorpej | RCS ID police.
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 28-Jun-1999 |
itojun | branches: 1.1.2;
file ip_ecn.c was initially added on branch kame.
|
| 1.1.2.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.1.2.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 01-Jul-1999 |
thorpej | file ip_ecn.c was added on branch chs-ubc2 on 1999-07-01 23:47:01 +0000
|
| 1.6.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.6.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.9.2.3 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.9.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.9.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.10.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.10.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.12.20.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.12.18.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.12.10.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.13.6.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.14.8.1 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.14.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.15.122.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.15.102.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.16.10.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by pgoyette in ticket #436):
distrib/sets/lists/modules/mi: revision 1.112
sys/modules/Makefile: revision 1.196
sys/modules/ip_ecn/Makefile: revision 1.1
sys/modules/if_gif/Makefile: revision 1.3
sys/net/if_gif.c: revision 1.136
sys/netinet/ip_ecn.c: revision 1.17
Split ip_ecn code into its own module, so it can be shared between
gif(4), stf(4), and ipsec(4). Without this, loading the if_gif
module can result in redefined global symbols if either ipsec(4) or
stf(4) but not gif(4) is built into the kernel.
Fixes PR kern/52795 (as reported by martin@ via irc).
|
| 1.12 | 12-Nov-2008 |
ad | Remove LKMs and switch to the module framework, pass 1.
Proposed on tech-kern@.
|
| 1.11 | 10-Dec-2005 |
elad | branches: 1.11.70; 1.11.74; 1.11.80; 1.11.84;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.10 | 18-Apr-2004 |
matt | branches: 1.10.12;
De __P()
|
| 1.9 | 21-Dec-2001 |
itojun | branches: 1.9.16;
whitespace. protect from multiple inclusion. sync with kame
|
| 1.8 | 10-May-2001 |
itojun | branches: 1.8.2;
correct ecn consideration on tunnel encap/decap. sync with kame.
|
| 1.7 | 12-Dec-1999 |
itojun | branches: 1.7.6;
sync with latest KAME (rcsid only).
|
| 1.6 | 31-Jul-1999 |
itojun | branches: 1.6.2; 1.6.8;
sync with recent KAME.
- loosen ipsec restriction on packet diredtion.
- revise icmp6 redirect handling on IsRouter bit.
- tcp/udp notification processing (link-local address case)
- cosmetic fixes (better code share across *BSD).
|
| 1.5 | 09-Jul-1999 |
thorpej | defopt INET6, and put it in opt_inet.h (most places already include this
file, which is why the file list is so short).
|
| 1.4 | 06-Jul-1999 |
itojun | sync with KAME/NetBSD 1.4, SNAP kit 19990705.
key changes are:
- icmp6 redirect fix (dst check)
- revised ip6 multicast check for loopback i/f
- several RCS ID cleanups
|
| 1.3 | 03-Jul-1999 |
thorpej | RCS ID police.
|
| 1.2 | 01-Jul-1999 |
itojun | branches: 1.2.2;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.1 | 28-Jun-1999 |
itojun | branches: 1.1.2;
file ip_ecn.h was initially added on branch kame.
|
| 1.1.2.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.1.2.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.2.2.1 | 01-Jul-1999 |
thorpej | file ip_ecn.h was added on branch chs-ubc2 on 1999-07-01 23:47:01 +0000
|
| 1.6.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.6.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.7.6.2 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.7.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.8.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.9.16.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.9.16.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.9.16.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.9.16.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.10.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.11.84.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.11.80.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.11.74.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.11.70.1 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.78 | 26-Feb-2025 |
andvar | Fix typos in comments, mainly s/calcurate/calculate/.
|
| 1.77 | 07-Dec-2022 |
knakahara | branches: 1.77.8;
Refactor ip_encap.[ch]
- remove encap_attach() which is no longer used
- remove USE_RADIX code in ip_encap.c, which is used for
encap_attach() only
- remove mask members in encaptab
|
| 1.76 | 07-Dec-2022 |
knakahara | Implement encap_attach_addr() which is used by IP-encaped tunnels.
The tunnels attached by encap_attach() can process receiving packets
fastly as the softc is searched by radix-tree. However, the tunnels
cannot use priority function which decides tunnel's softc by not only
source and destination but also other informations.
On the other hand, the tunnels attached by encap_attach_func() can
use priority function. However, the tunnels can be slow receiving
processing as the softc is searched by linear search (and uses each
priority function).
encap_attach_addr() can be used for tunnels which is fixed tunnel
source address and tunnel destination address. The tunnels attached
by encap_attach_addr() is searched by thmap(9), so the receiving processing
can be fast. Moreover, the tunnels can use priority function.
|
| 1.75 | 07-Dec-2022 |
knakahara | refactor: use typedef for ip_encap priority function
|
| 1.74 | 22-Aug-2020 |
riastradh | Mark KASSERT-only variable __diagused.
|
| 1.73 | 20-Aug-2020 |
riastradh | [ozaki-r] Changes to the kernel core for wireguard
|
| 1.72 | 23-Jan-2020 |
knakahara | Fix PR security/54881. Pointed out by ohishi@IIJ, thanks.
|
| 1.71 | 15-May-2019 |
knakahara | branches: 1.71.2; 1.71.4;
Fix build failure when INET6 is disabled and NET_MPSAFE is enabled. Pointed out by ozaki-r@n.o, thanks.
|
| 1.70 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.69 | 21-Jun-2018 |
knakahara | branches: 1.69.2;
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.68 | 01-May-2018 |
maxv | Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.67 | 14-Jan-2018 |
maxv | branches: 1.67.2;
Fix memory leak, found by Mootja.
|
| 1.66 | 15-Nov-2017 |
knakahara | Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.65 | 01-Jun-2017 |
chs | branches: 1.65.2;
remove checks for failure after memory allocation calls that cannot fail:
kmem_alloc() with KM_SLEEP
kmem_zalloc() with KM_SLEEP
percpu_alloc()
pserialize_create()
psref_class_create()
all of these paths include an assertion that the allocation has not failed,
so callers should not assert that again.
|
| 1.64 | 15-Apr-2017 |
riastradh | No need for membar_datadep_consumer here.
PSLIST_READER_FOREACH takes care of it already.
PSLIST_WRITER_FOREACH is exclusive so doesn't need it.
|
| 1.63 | 07-Apr-2017 |
ozaki-r | Commit a forgotten change for "Prepare netipsec for rump-ification"
- Allow encapinit to be called twice (by ifinit and ipe4_attach)
- ifinit didn't call encapinit if IPSEC is enabled (ipe4_attach called
it instead), however, on a rump kernel ipe4_attach may not be called
even if IPSEC is enabled. So we need to allow ifinit to call it anyway
|
| 1.62 | 22-Dec-2016 |
knakahara | branches: 1.62.2;
pserialize_perform() is required *after* PSLIST_WRITER_REMOVE.
|
| 1.61 | 04-Jul-2016 |
knakahara | branches: 1.61.2;
make gif(4) and ip_encap MP-ify
|
| 1.60 | 04-Jul-2016 |
knakahara | refactor: merge encap_init_once() to encapinit()
|
| 1.59 | 04-Jul-2016 |
knakahara | make encap_lock_{enter,exit} interruptable.
|
| 1.58 | 04-Jul-2016 |
knakahara | remove extra pserialize_perform()
|
| 1.57 | 04-Jul-2016 |
knakahara | use pserialize(9) and psref(9) (2/2) : ip_encap radix tree care
|
| 1.56 | 04-Jul-2016 |
knakahara | use pserialize(9) and psref(9) (1/2) : without ip_encap radix tree care
|
| 1.55 | 04-Jul-2016 |
knakahara | restore ifdef USE_RADIX (revert ip_encap.c:r1.44)
To help future ip_encap optimaization works.
|
| 1.54 | 04-Jul-2016 |
knakahara | let gif(4) promise softint(9) contract (2/2) : ip_encap side
The last commit does not care encaptab. This commit fixes encaptab race which
is used not only gif(4).
|
| 1.53 | 26-Apr-2016 |
ozaki-r | Sweep unnecessary route.h inclusions
|
| 1.52 | 26-Feb-2016 |
knakahara | To eliminate gif_softc_list linear search, add extra argument to encapsw.pr_ctlinput().
|
| 1.51 | 26-Jan-2016 |
knakahara | implement encapsw instead of protosw and uniform prototype.
suggested and advised by riastradh@n.o, thanks.
BTW, It seems in_stf_input() had bugs...
|
| 1.50 | 22-Jan-2016 |
riastradh | Back out previous change to introduce struct encapsw.
This change was intended, but Nakahara-san had already made a better
one locally! So I'll let him commit that one, and I'll try not to
step on anyone's toes again.
|
| 1.49 | 22-Jan-2016 |
riastradh | Don't abuse struct protosw for ip_encap -- introduce struct encapsw.
Mostly mechanical change to replace it, culling some now-needless
boilerplate around all the users.
This does not substantively change the ip_encap API or eliminate
abuse of sketchy pointer casts -- that will come later, and will be
easier now that it is not tangled up with struct protosw.
|
| 1.48 | 20-Jan-2016 |
knakahara | remove unused variable.
|
| 1.47 | 09-Dec-2015 |
knakahara | ip_encap uses kmem_alloc APIs instead of malloc.
|
| 1.46 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.45 | 20-Apr-2015 |
ozaki-r | Remove non-USE_RADIX case and USE_RADIX switch
It seems that we have been using ip_encap only with USE_RADIX
for long years. Let's remove unused non-USE_RADIX case.
No objection on tech-kern and tech-net.
Double-checked by knakahara@
|
| 1.44 | 16-Apr-2015 |
ozaki-r | Remove garbage undef
|
| 1.43 | 15-Apr-2015 |
riastradh | KASSERT x then y, not x && y, to give more specific errors.
|
| 1.42 | 15-Apr-2015 |
ozaki-r | Use LIST_FOREACH_SAFE
We have to use LIST_FOREACH_SAFE because LIST_REMOVE is used
inside the loop through encap_remove.
|
| 1.41 | 15-Apr-2015 |
ozaki-r | Replace DIAGNOSTIC & panic with KASSERT/KASSERTMSG
|
| 1.40 | 15-Apr-2015 |
ozaki-r | Add $NetBSD$ at the top of the file
|
| 1.39 | 17-Jul-2011 |
joerg | branches: 1.39.12; 1.39.30;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.38 | 27-May-2009 |
pooka | Make it possible to register delayed radix tree head inits which
will be processed when the radix "subsystem" is initialized -- all
users must be attached before any inits to know the max keylength.
Use of link sets is no longer required, and only attached domains
need to be considered.
|
| 1.37 | 18-Apr-2009 |
tsutsui | Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.36 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.35 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.34 | 18-Mar-2009 |
cegger | bcmp -> memcmp
|
| 1.33 | 25-Nov-2008 |
pooka | branches: 1.33.4;
Make dom_maxrtkey of inet/inet6domain the size of the ip_encap pack
structures. This is far from optimal, but gets rid of iffy
#ifdef INET in radix.c. The radix bonsai still needs lots of love
before loading domains dynamically is possible...
|
| 1.32 | 24-Apr-2008 |
ad | branches: 1.32.2; 1.32.8; 1.32.10;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.31 | 13-Jun-2007 |
dyoung | branches: 1.31.28; 1.31.30;
Use LIST_FOREACH().
|
| 1.30 | 04-Mar-2007 |
christos | branches: 1.30.2; 1.30.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.29 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.28 | 28-May-2006 |
liamjfoy | branches: 1.28.12;
remove some dead code
ok christos@
|
| 1.27 | 11-Dec-2005 |
christos | branches: 1.27.4; 1.27.6; 1.27.8; 1.27.14;
merge ktrace-lwp.
|
| 1.26 | 06-Jun-2005 |
martin | branches: 1.26.2;
Since we decided "const struct mbuf *" would not do the right thing (tm),
remove ~all const from mbuf pointers.
|
| 1.25 | 03-Jun-2005 |
martin | Sprinkle some const
|
| 1.24 | 02-Jun-2005 |
tron | Change the first argument of the encapsulation check function from
"const struct mbuf *" to "struct mbuf *". Without this change the
actual implementation cannot even use m_copydata() on the mbuf chain
which is broken.
|
| 1.23 | 03-Feb-2005 |
perry | ANSIfy function prototypes. (Still have about 3/5ths of the C files in
netinet to go...)
|
| 1.22 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.21 | 24-Jan-2005 |
enami | branches: 1.21.2;
To fix bad pointer dereference on start up when gif is used,
- Allow rn_init() to be called multiple times, but do nothing except the
first call.
- Include opt_inet.h so that #ifdef INET works.
- Call rn_init() from encap_init() explicitly rather than depending on the
order of initialization.
|
| 1.20 | 24-Jan-2005 |
itojun | get zero-cleared field on malloc. kame-pr-856
|
| 1.19 | 17-Aug-2004 |
itojun | branches: 1.19.4;
initialize max_keylen for ip_encap.c earlier
|
| 1.18 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.17 | 04-Mar-2004 |
wiz | branches: 1.17.4;
No need to include netinet/ip_mroute.h twice.
Closes PR 24652 by Kailash Sethuraman.
|
| 1.16 | 30-Oct-2003 |
simonb | Remove some assigned-to but otherwise unused variables.
|
| 1.15 | 28-Oct-2003 |
mycroft | Do the previous differently.
|
| 1.14 | 25-Oct-2003 |
christos | Fix uninitialized variable warning
|
| 1.13 | 21-Jan-2003 |
itojun | branches: 1.13.2;
correct panic when ip-in-ip encapsulation is used. found by Masanori Kanaoka
|
| 1.12 | 17-Jan-2003 |
itojun | switch from kame-based m_aux mbuf auxiliary data, to openbsd m_tag
implementation. it will simplify porting across *bsd (such as kame/altq),
and make us more synchronized. from Joel Wilsson
|
| 1.11 | 25-Nov-2002 |
thorpej | Avoid strict-alias warnings.
|
| 1.10 | 31-Jul-2002 |
itojun | remove packed attribute as it will cause data be unaligned
|
| 1.9 | 09-Jun-2002 |
itojun | whitespace
|
| 1.8 | 04-Mar-2002 |
sommerfeld | branches: 1.8.6; 1.8.8;
The "gif*" tunnelling interface does everything ipip does.
Move usage example from ipip.4 to gif.4
Excise ipip and stitch up the scars.
|
| 1.7 | 21-Dec-2001 |
itojun | use radix table for inbound tunnel lookup (would increase performance
for machines with a lot of tunnels).
update route cache for IPvX-over-IPv6 tunnel on path MTU discovery.
snyc with kame
|
| 1.6 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.5 | 08-May-2001 |
itojun | branches: 1.5.2;
pull encapsulated packet for vif* via ip_encap framework.
|
| 1.4 | 02-Oct-2000 |
itojun | branches: 1.4.2; 1.4.4;
fix compilation without INET. fix confusion between ipsecstat and ipsec6stat.
sync with kame.
|
| 1.3 | 05-Jul-2000 |
thorpej | Make that note that we really should be checking the viftable
in ip_mroute.c for duplicate tunnel entries, too. Well, what
really needs to happen is that the mrouting code needs to be
changed to work w/ `gif' tunnels... but...
|
| 1.2 | 05-Jul-2000 |
thorpej | Use LIST_HEAD_INITIALIZER(), for correctness sake.
|
| 1.1 | 19-Apr-2000 |
itojun | introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.4.4.8 | 17-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.4.4.7 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.4.4.6 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.4.4.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.4.4.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.4.4.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.4.4.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.4.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.4.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.4.2.1 | 02-Oct-2000 |
bouyer | file ip_encap.c was added on branch thorpej_scsipi on 2000-11-20 18:10:25 +0000
|
| 1.5.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.5.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.5.2.2 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.5.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.8.8.1 | 02-Aug-2002 |
lukem | Pull up revision 1.10 (requested by itojun in ticket #593):
remove packed attribute as it will cause data be unaligned
|
| 1.8.6.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.8.6.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.13.2.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.13.2.6 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.13.2.5 | 24-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.13.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.13.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.13.2.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.13.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.17.4.1 | 06-May-2005 |
riz | Pull up revision 1.20 (requested by bouyer in ticket #1188):
get zero-cleared field on malloc. kame-pr-856
|
| 1.19.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.21.2.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.26.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.26.2.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.26.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.27.14.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.27.8.1 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.27.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.27.4.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.27.4.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.28.12.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.28.12.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.30.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.30.2.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.31.30.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.31.28.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.31.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.32.10.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.32.10.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.32.8.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.32.2.2 | 20-Jun-2009 |
yamt | sync with head
|
| 1.32.2.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.33.4.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.33.4.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.39.30.8 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.39.30.7 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.39.30.6 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.39.30.5 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.39.30.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.39.30.3 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.39.30.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.39.30.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.39.12.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.61.2.2 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.61.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.62.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.65.2.5 | 23-Jan-2020 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1489):
sys/netinet/ip_encap.c: revision 1.72
Fix PR security/54881. Pointed out by ohishi@IIJ, thanks.
|
| 1.65.2.4 | 29-May-2019 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1271):
sys/netinet/ip_encap.c: revision 1.71
Fix build failure when INET6 is disabled. Pointed out by ozaki-r@n.o, thanks.
|
| 1.65.2.3 | 13-Jul-2018 |
martin | Pull up following revision(s) via patch (requested by knakahara in ticket #905):
sys/netinet/ip_mroute.c: revision 1.160
sys/netinet6/in6_l2tp.c: revision 1.16
sys/net/if.h: revision 1.263
sys/netinet/in_l2tp.c: revision 1.15
sys/netinet/ip_icmp.c: revision 1.172
sys/netinet/igmp.c: revision 1.68
sys/netinet/ip_encap.c: revision 1.69
sys/netinet6/ip6_mroute.c: revision 1.129
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.65.2.2 | 30-Mar-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #661):
sys/netinet/ip_encap.c: revision 1.67
Fix memory leak, found by Mootja.
|
| 1.65.2.1 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by knakahara in ticket #419):
sys/net/if_stf.c: revision 1.103
sys/net/if_stf.h: revision 1.8
sys/netinet/in_gif.c: revision 1.89
sys/netinet/in_gif.h: revision 1.17
sys/netinet/in_l2tp.c: revision 1.4
sys/netinet/ip_encap.c: revision 1.66
sys/netinet/ip_encap.h: revision 1.23
sys/netinet/ip_mroute.c: revision 1.148
sys/netinet6/in6_gif.c: revision 1.87
sys/netinet6/in6_gif.h: revision 1.16
sys/netinet6/in6_l2tp.c: revision 1.7
sys/netipsec/xform.h: revision 1.13
sys/netipsec/xform_ipip.c: revision 1.55
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.67.2.3 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.67.2.2 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.67.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.69.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.69.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.71.4.1 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.71.2.1 | 23-Jan-2020 |
martin | Pull up following revision(s) (requested by knakahara in ticket #644):
sys/netinet/ip_encap.c: revision 1.72
Fix PR security/54881. Pointed out by ohishi@IIJ, thanks.
|
| 1.77.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.28 | 07-Dec-2022 |
knakahara | Refactor ip_encap.[ch]
- remove encap_attach() which is no longer used
- remove USE_RADIX code in ip_encap.c, which is used for
encap_attach() only
- remove mask members in encaptab
|
| 1.27 | 07-Dec-2022 |
knakahara | Implement encap_attach_addr() which is used by IP-encaped tunnels.
The tunnels attached by encap_attach() can process receiving packets
fastly as the softc is searched by radix-tree. However, the tunnels
cannot use priority function which decides tunnel's softc by not only
source and destination but also other informations.
On the other hand, the tunnels attached by encap_attach_func() can
use priority function. However, the tunnels can be slow receiving
processing as the softc is searched by linear search (and uses each
priority function).
encap_attach_addr() can be used for tunnels which is fixed tunnel
source address and tunnel destination address. The tunnels attached
by encap_attach_addr() is searched by thmap(9), so the receiving processing
can be fast. Moreover, the tunnels can use priority function.
|
| 1.26 | 07-Dec-2022 |
knakahara | refactor: use typedef for ip_encap priority function
|
| 1.25 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.24 | 28-Feb-2018 |
maxv | branches: 1.24.2; 1.24.4;
Remove unused mbuf tags.
|
| 1.23 | 15-Nov-2017 |
knakahara | Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.22 | 04-Jul-2016 |
knakahara | branches: 1.22.10;
make encap_lock_{enter,exit} interruptable.
|
| 1.21 | 04-Jul-2016 |
knakahara | use pserialize(9) and psref(9) (1/2) : without ip_encap radix tree care
|
| 1.20 | 04-Jul-2016 |
knakahara | let gif(4) promise softint(9) contract (2/2) : ip_encap side
The last commit does not care encaptab. This commit fixes encaptab race which
is used not only gif(4).
|
| 1.19 | 29-Feb-2016 |
knakahara | remove unnecessary declarations and fix KNF
Thanks to riastradh@
|
| 1.18 | 26-Feb-2016 |
knakahara | To eliminate gif_softc_list linear search, add extra argument to encapsw.pr_ctlinput().
|
| 1.17 | 26-Jan-2016 |
knakahara | eliminate variable argument in encapsw
|
| 1.16 | 26-Jan-2016 |
knakahara | implement encapsw instead of protosw and uniform prototype.
suggested and advised by riastradh@n.o, thanks.
BTW, It seems in_stf_input() had bugs...
|
| 1.15 | 22-Jan-2016 |
riastradh | Back out previous change to introduce struct encapsw.
This change was intended, but Nakahara-san had already made a better
one locally! So I'll let him commit that one, and I'll try not to
step on anyone's toes again.
|
| 1.14 | 22-Jan-2016 |
riastradh | Don't abuse struct protosw for ip_encap -- introduce struct encapsw.
Mostly mechanical change to replace it, culling some now-needless
boilerplate around all the users.
This does not substantively change the ip_encap API or eliminate
abuse of sketchy pointer casts -- that will come later, and will be
easier now that it is not tangled up with struct protosw.
|
| 1.13 | 25-Nov-2008 |
pooka | branches: 1.13.26; 1.13.44;
Make dom_maxrtkey of inet/inet6domain the size of the ip_encap pack
structures. This is far from optimal, but gets rid of iffy
#ifdef INET in radix.c. The radix bonsai still needs lots of love
before loading domains dynamically is possible...
|
| 1.12 | 24-Apr-2008 |
ad | branches: 1.12.2; 1.12.8; 1.12.10;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.11 | 17-Feb-2007 |
dyoung | branches: 1.11.38; 1.11.40;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.10 | 10-Dec-2005 |
elad | branches: 1.10.26;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.9 | 06-Jun-2005 |
martin | branches: 1.9.2;
Since we decided "const struct mbuf *" would not do the right thing (tm),
remove ~all const from mbuf pointers.
|
| 1.8 | 03-Jun-2005 |
martin | Sprinkle some const
|
| 1.7 | 02-Jun-2005 |
tron | Change the first argument of the encapsulation check function from
"const struct mbuf *" to "struct mbuf *". Without this change the
actual implementation cannot even use m_copydata() on the mbuf chain
which is broken.
|
| 1.6 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.5 | 17-Aug-2004 |
itojun | branches: 1.5.4; 1.5.6;
initialize max_keylen for ip_encap.c earlier
|
| 1.4 | 18-Apr-2004 |
matt | De __P()
|
| 1.3 | 17-Jan-2003 |
itojun | branches: 1.3.2;
switch from kame-based m_aux mbuf auxiliary data, to openbsd m_tag
implementation. it will simplify porting across *bsd (such as kame/altq),
and make us more synchronized. from Joel Wilsson
|
| 1.2 | 21-Dec-2001 |
itojun | use radix table for inbound tunnel lookup (would increase performance
for machines with a lot of tunnels).
update route cache for IPvX-over-IPv6 tunnel on path MTU discovery.
snyc with kame
|
| 1.1 | 19-Apr-2000 |
itojun | branches: 1.1.6; 1.1.8; 1.1.10;
introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.1.10.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.1.8.2 | 17-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.1.8.1 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.1.6.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.1.6.1 | 19-Apr-2000 |
bouyer | file ip_encap.h was added on branch thorpej_scsipi on 2000-11-20 18:10:25 +0000
|
| 1.3.2.7 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.3.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.3.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.3.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.3.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.3.2.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.3.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.5.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.5.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.9.2.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.9.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.10.26.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.11.40.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.11.38.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.11.38.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.12.10.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.12.8.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.12.2.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.13.44.2 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.13.44.1 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.13.26.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.22.10.1 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by knakahara in ticket #419):
sys/net/if_stf.c: revision 1.103
sys/net/if_stf.h: revision 1.8
sys/netinet/in_gif.c: revision 1.89
sys/netinet/in_gif.h: revision 1.17
sys/netinet/in_l2tp.c: revision 1.4
sys/netinet/ip_encap.c: revision 1.66
sys/netinet/ip_encap.h: revision 1.23
sys/netinet/ip_mroute.c: revision 1.148
sys/netinet6/in6_gif.c: revision 1.87
sys/netinet6/in6_gif.h: revision 1.16
sys/netinet6/in6_l2tp.c: revision 1.7
sys/netipsec/xform.h: revision 1.13
sys/netipsec/xform_ipip.c: revision 1.55
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.24.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.24.2.1 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.22 | 14-Aug-2018 |
maxv | Retire EtherIP, we have L2TP instead.
|
| 1.21 | 26-Jan-2018 |
maxv | branches: 1.21.2; 1.21.4;
A few fixes:
* Style.
* Don't add M_PKTHDR manually, that's absolutely forbidden. Add a
KASSERT to make sure it's already there.
* Add a missing NULL check after m_pullup.
|
| 1.20 | 11-Jan-2017 |
ozaki-r | branches: 1.20.8;
Get rid of unnecessary header inclusions
|
| 1.19 | 15-Dec-2016 |
ozaki-r | Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input
The benefits of the change are:
- We can reduce codes
- We can provide the same behavior between drivers
- Where/When if_ipackets is counted up
- Note that some drivers still update packet statistics in their own
way (periodical update)
- Moved bpf_mtap run in softint
- This makes it easy to MP-ify bpf
Proposed on tech-kern and tech-net
|
| 1.18 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.17 | 10-Jun-2016 |
ozaki-r | branches: 1.17.2;
Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.16 | 09-Feb-2016 |
ozaki-r | Introduce softint-based if_input
This change intends to run the whole network stack in softint context
(or normal LWP), not hardware interrupt context. Note that the work is
still incomplete by this change; to that end, we also have to softint-ify
if_link_state_change (and bpf) which can still run in hardware interrupt.
This change softint-ifies at ifp->if_input that is called from
each device driver (and ieee80211_input) to ensure Layer 2 runs
in softint (e.g., ether_input and bridge_input). To this end,
we provide a framework (called percpuq) that utlizes softint(9)
and percpu ifqueues. With this patch, rxintr of most drivers just
queues received packets and schedules a softint, and the softint
dequeues packets and does rest packet processing.
To minimize changes to each driver, percpuq is allocated in struct
ifnet for now and that is initialized by default (in if_attach).
We probably have to move percpuq to softc of each driver, but it's
future work. At this point, only wm(4) has percpuq in its softc
as a reference implementation.
Additional information including performance numbers can be found
in the thread at tech-kern@ and tech-net@:
http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html
Acknowledgment: riastradh@ greatly helped this work.
Thank you very much!
|
| 1.15 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.14 | 17-Jul-2011 |
joerg | branches: 1.14.12; 1.14.30;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.13 | 05-Apr-2010 |
joerg | Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf
check into the inline functions as well the fourth argument for
bpf_attach.
|
| 1.12 | 19-Jan-2010 |
pooka | branches: 1.12.2; 1.12.4;
Redefine bpf linkage through an always present op vector, i.e.
#if NBPFILTER is no longer required in the client. This change
doesn't yet add support for loading bpf as a module, since drivers
can register before bpf is attached. However, callers of bpf can
now be modularized.
Dynamically loadable bpf could probably be done fairly easily with
coordination from the stub driver and the real driver by registering
attachments in the stub before the real driver is loaded and doing
a handoff. ... and I'm not going to ponder the depths of unload
here.
Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
|
| 1.11 | 19-Oct-2008 |
hans | if_input needs to be called at splnet(). ok by cube.
|
| 1.10 | 16-Oct-2008 |
hans | include bpf headers so that the bpf calls actually do something. ok by cube.
|
| 1.9 | 12-Apr-2008 |
thorpej | branches: 1.9.4; 1.9.10;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.8 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.7 | 20-Dec-2007 |
dyoung | branches: 1.7.6;
Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.6 | 11-Dec-2007 |
lukem | use __KERNEL_RCSID()
|
| 1.5 | 02-May-2007 |
dyoung | branches: 1.5.8; 1.5.16; 1.5.18; 1.5.20;
Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.4 | 17-Feb-2007 |
dyoung | branches: 1.4.4; 1.4.6;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.3 | 15-Dec-2006 |
joerg | branches: 1.3.2; 1.3.4; 1.3.6;
Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.2 | 06-Dec-2006 |
jdc | branches: 1.2.2;
Explicitly include <sys/device.h>, which we need for `struct device'.
This allows us to compile on !i386. (On i386, <machine/cpu.h> pulled
in <sys/device.h> for us, thus hiding the compilation problem.)
OK by rpaulo@.
|
| 1.1 | 23-Nov-2006 |
rpaulo | branches: 1.1.2;
New EtherIP driver based on tap(4) and gif(4) by Hans Rosenfeld.
Notable changes:
* Fixes PR 34268.
* Separates the code from gif(4) (which is more cleaner).
* Allows the usage of STP (Spanning Tree Protocol).
* Removed EtherIP implementation from gif(4)/tap(4).
Some input from Christos.
|
| 1.1.2.1 | 09-Dec-2006 |
bouyer | Pull up following revision(s) (requested by jdc in ticket #259):
sys/netinet6/ip6_etherip.c: revision 1.2
sys/netinet/ip_etherip.c: revision 1.2
Explicitly include <sys/device.h>, which we need for `struct device'.
This allows us to compile on !i386. (On i386, <machine/cpu.h> pulled
in <sys/device.h> for us, thus hiding the compilation problem.)
OK by rpaulo@.
|
| 1.2.2.3 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.2.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.2.2.1 | 06-Dec-2006 |
yamt | file ip_etherip.c was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.3.6.2 | 07-May-2007 |
yamt | sync with head.
|
| 1.3.6.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.3.4.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.3.4.1 | 15-Dec-2006 |
ad | file ip_etherip.c was added on branch newlock2 on 2007-01-12 01:04:14 +0000
|
| 1.3.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.3.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.3.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.3.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.3.2.1 | 15-Dec-2006 |
yamt | file ip_etherip.c was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.4.6.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.4.4.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.5.20.2 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.5.20.1 | 13-Dec-2007 |
bouyer | Sync with HEAD
|
| 1.5.18.1 | 11-Dec-2007 |
yamt | sync with head.
|
| 1.5.16.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.5.8.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.7.6.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.7.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.9.10.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.9.10.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.9.4.3 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.9.4.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.9.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.12.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.12.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.14.30.4 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.14.30.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.14.30.2 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.14.30.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.14.12.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.17.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.17.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.20.8.1 | 05-Apr-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #694):
sys/netinet6/ip6_etherip.c: revision 1.22
sys/net/if_etherip.c: revision 1.41
sys/net/if_etherip.c: revision 1.42
sys/netinet/ip_etherip.c: revision 1.21
Don't call if_attach, do if_initialize+if_register, otherwise when an
EtherIP packet is received the first KASSERT in if_input() fires.
A few fixes:
* Style.
* Don't add M_PKTHDR manually, that's absolutely forbidden. Add a
KASSERT to make sure it's already there.
* Add a missing NULL check after m_pullup.
|
| 1.21.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.21.2.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.2 | 14-Aug-2018 |
maxv | Retire EtherIP, we have L2TP instead.
|
| 1.1 | 23-Nov-2006 |
rpaulo | branches: 1.1.4; 1.1.6; 1.1.8; 1.1.146; 1.1.148;
New EtherIP driver based on tap(4) and gif(4) by Hans Rosenfeld.
Notable changes:
* Fixes PR 34268.
* Separates the code from gif(4) (which is more cleaner).
* Allows the usage of STP (Spanning Tree Protocol).
* Removed EtherIP implementation from gif(4)/tap(4).
Some input from Christos.
|
| 1.1.148.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.1.146.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.1.8.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.1.8.1 | 23-Nov-2006 |
ad | file ip_etherip.h was added on branch newlock2 on 2007-01-12 01:04:14 +0000
|
| 1.1.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.1.6.1 | 23-Nov-2006 |
yamt | file ip_etherip.h was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.1.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.1.4.1 | 23-Nov-2006 |
yamt | file ip_etherip.h was added on branch yamt-splraiseipl on 2006-12-10 07:19:10 +0000
|
| 1.96 | 28-Mar-2004 |
martti | Upgraded IPFilter to 4.1.1
|
| 1.95 | 22-Aug-2003 |
itojun | correct missing inclusion of opt_ipsec.h
|
| 1.94 | 15-Aug-2003 |
martti | Fix return-rst for IPv6 (PR#22157 by Peter Postma).
|
| 1.93 | 30-Jun-2003 |
itojun | branches: 1.93.2;
remove IPv4 hook if IPv6 hook fails (seems to be cut-and-paste error).
|
| 1.92 | 29-Jun-2003 |
fvdl | Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.91 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.90 | 26-Jun-2003 |
itojun | tabify
|
| 1.89 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.88 | 11-Dec-2002 |
atatat | Always recompute the IP checksum, otherwise fast-routed packets that
also get natted leave with an invalid checksum which can prevent
things from working properly.
|
| 1.87 | 23-Oct-2002 |
jdolecek | merge kqueue branch into -current
kqueue provides a stateful and efficient event notification framework
currently supported events include socket, file, directory, fifo,
pipe, tty and device changes, and monitoring of processes and signals
kqueue is supported by all writable filesystems in NetBSD tree
(with exception of Coda) and all device drivers supporting poll(2)
based on work done by Jonathan Lemon for FreeBSD
initial NetBSD port done by Luke Mewburn and Jason Thorpe
|
| 1.86 | 19-Sep-2002 |
martti | Resync with official IPF
|
| 1.85 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.84 | 07-Sep-2002 |
enami | Make usr.sbin/ipf/ipftest compiles again.
|
| 1.83 | 06-Sep-2002 |
gehenna | The device switch ``ipl_cdevsw'' is defined after 1.6H.
|
| 1.82 | 06-Sep-2002 |
gehenna | Merge the gehenna-devsw branch into the trunk.
This merge changes the device switch tables from static array to
dynamically generated by config(8).
- All device switches is defined as a constant structure in device drivers.
- The new grammer ``device-major'' is introduced to ``files''.
device-major <prefix> char <num> [block <num>] [<rules>]
- All device major numbers must be listed up in port dependent majors.<arch>
by using this grammer.
- Added the new naming convention.
The name of the device switch must be <prefix>_[bc]devsw for auto-generation
of device switch tables.
- The backward compatibility of loading block/character device
switch by LKM framework is broken. This is necessary to convert
from block/character device major to device name in runtime and vice versa.
- The restriction to assign device major by LKM is completely removed.
We don't need to reserve LKM entries for dynamic loading of device switch.
- In compile time, device major numbers list is packed into the kernel and
the LKM framework will refer it to assign device major number dynamically.
|
| 1.81 | 09-Jun-2002 |
itojun | whitespace
|
| 1.80 | 29-May-2002 |
itojun | attach nd_ifinfo structure into if_afdata.
split IPv6 link MTU (advertised by RA) from real link MTU.
sync with kame
|
| 1.79 | 02-May-2002 |
martti | branches: 1.79.2; 1.79.4;
Fix compilation problems
|
| 1.78 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.77 | 14-Mar-2002 |
martti | Added (char *) for pointer arithmetic
|
| 1.76 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.75 | 24-Jan-2002 |
martti | Fixed initialization
|
| 1.74 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.73 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.72 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.71 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.70 | 18-Oct-2001 |
thorpej | Deprecate the "m_act" alias of "m_nextpkt" (m_act is a historical
name), and just use m_nextpkt everywhere.
|
| 1.69 | 17-Sep-2001 |
thorpej | Split the pre-computed ifnet checksum flags into Tx and Rx directions.
Add capabilities bits that indicate an interface can only perform
in-bound TCPv4 or UDPv4 checksums. There is at least one Gig-E chip
for which this is true (Level One LXT-1001), and this is also the
case for the Intel i82559 10/100 Ethernet chips.
|
| 1.68 | 02-Jun-2001 |
thorpej | branches: 1.68.2; 1.68.4;
Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.67 | 12-May-2001 |
christos | - Handle realloc failure without leaking memory
(reported by: grendel@heorot.stanford.edu (Ted U)
- Don't cast malloc/realloc/calloc return values because they hide LP64 bugs.
- Don't destroy the whole array when realloc fails
- Use calloc in all cases (malloc was used inconsistently).
- Avoid duplicating code.
Reviewed by: ross
|
| 1.66 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.65 | 05-Feb-2001 |
chs | branches: 1.65.2;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.64 | 28-Dec-2000 |
thorpej | Fix a small typo that would cause IP Filter to not hook in to
pfil_hooks properly on kernels that included IPv6 support.
|
| 1.63 | 28-Dec-2000 |
thorpej | Back out the sledgehammer damage applied by wiz while I was out for
the holiday.
|
| 1.62 | 25-Dec-2000 |
wiz | Back out previous change. It causes NAT to fail, and was CLEARLY
NOT TESTED before it was committed.
|
| 1.61 | 22-Dec-2000 |
thorpej | Slight adjustment to how pfil_head's are registered. Instead of a
"key" and a "dlt", use a "type" (PFIL_TYPE_{AF,IFNET} for now) and
a val/ptr appropriate for that type. This allows for more future
flexibility with the pfil_hook mechanism.
|
| 1.60 | 11-Nov-2000 |
thorpej | Restructure the PFIL_HOOKS mechanism a bit:
- All packets are passed to PFIL_HOOKS as they come off the wire, i.e.
fields in protocol headers in network order, etc.
- Allow for multiple hooks to be registered, using a "key" and a "dlt".
The "dlt" is a BPF data link type, indicating what type of header is
present.
- INET and INET6 register with key == AF_INET or AF_INET6, and
dlt == DLT_RAW.
- PFIL_HOOKS now take an argument for the filter hook, and mbuf **,
an ifnet *, and a direction (PFIL_IN or PFIL_OUT), thus making them
less IP (really, IP Filter) centric.
Maintain compatibility with IP Filter by adding wrapper functions for
IP Filter.
|
| 1.59 | 22-Aug-2000 |
sommerfeld | Fill in next mtu field of NEEDFRAG ICMP error message.
From Marc Horowitz, pr10857
|
| 1.58 | 09-Aug-2000 |
veego | Resolve conflicts.
|
| 1.57 | 01-Aug-2000 |
thorpej | Slight adjustment to last, to allow the userland version to build.
|
| 1.56 | 01-Aug-2000 |
thorpej | - ipl_enable(): -1 is not an error return. If initializing IP Filter
fails, return EIO instead.
- iplioctl(): If performing a NAT operation, and IP Filter is not
yet initialized (e.g. by `ipf -E'), enable it implicitly before
doing the NAT operation.
|
| 1.55 | 12-Jun-2000 |
veego | branches: 1.55.2;
Remove a duplicated check for the NetBSD callout (I think it is a mistake
from a previous conflict resolve which doesn't cause harm).
|
| 1.54 | 12-Jun-2000 |
veego | Resolve conflicts.
|
| 1.53 | 23-May-2000 |
veego | branches: 1.53.2;
Resolve conflicts.
|
| 1.52 | 21-May-2000 |
veego | Add a missing ; at the end of a line.
|
| 1.51 | 21-May-2000 |
veego | Resolve conflicts.
|
| 1.50 | 11-May-2000 |
veego | Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.49 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.48 | 16-Apr-2000 |
chs | remove ifdefs to skip htons() on some big-endian platforms.
|
| 1.47 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.46 | 24-Mar-2000 |
thorpej | Pull in <sys/callout.h> for the benefit of userland.
|
| 1.45 | 23-Mar-2000 |
thorpej | New callout mechanism with two major improvements over the old
timeout()/untimeout() API:
- Clients supply callout handle storage, thus eliminating problems of
resource allocation.
- Insertion and removal of callouts is constant time, important as
this facility is used quite a lot in the kernel.
The old timeout()/untimeout() API has been removed from the kernel.
|
| 1.44 | 07-Mar-2000 |
mycroft | Fix a splx() botch or two.
|
| 1.43 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.42 | 20-Feb-2000 |
darrenr | pass "struct pfil_head *" to pfil_add_hook and pfil_remove hook rather
than "struct protosw *".
|
| 1.41 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.40 | 01-Feb-2000 |
veego | Only print one 'IP Filter:' line when it gets enabled or disabled.
|
| 1.39 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.38 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.37 | 12-Oct-1999 |
sommerfeld | branches: 1.37.2;
in ipfr_fastroute, before calling icmp_error(), put received-interface
back into the packet. (ip_output() clears it since ipsec reuses that
packet field in the output path. by putting it back, we're going to
pretend we're back on the input path now).
|
| 1.36 | 26-Aug-1999 |
itojun | branches: 1.36.2;
clear m->m_pkthdr.rcvif before calling ip_output().
the member is used to pass struct socket to ip{,6}_output for ipsec decisions.
(i agree it is kind of ugly. we need to modify struct mbuf if we are
to do better - which seems to me a bit too much)
|
| 1.35 | 26-Aug-1999 |
marc | when fastrouting a packet which needs fragmentation, the packet passed
to if_output did not have m->m_pkthdr.len set correctly. Add the code
to do this from the similar code in ip_output.c
|
| 1.34 | 02-Feb-1999 |
cjs | branches: 1.34.2; 1.34.6;
Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.33 | 19-Jan-1999 |
mycroft | There's just no plausible reason to byte-swap ip_id internally. It's opaque.
|
| 1.32 | 25-Nov-1998 |
sommerfe | Fragments should start with a header mbuf allocated by MGETHDR()
|
| 1.31 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.30 | 15-Nov-1998 |
drochner | fix the previous: "securelevel" in kernel only
|
| 1.29 | 14-Nov-1998 |
tls | In 'highly secure' mode (securelevel >= 2), the filter lists may not be tampered with. It might be desirable to allow enabling of preset filter lists, but it seems too good a candidate for a denial-of-service attack, so we don't.
|
| 1.28 | 17-Jul-1998 |
sommerfe | Fix PR5559: if fast-forwarding, DF set, and packet too large, send ICMP error..
|
| 1.27 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.26 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.25 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.24 | 21-Sep-1997 |
veego | branches: 1.24.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.23 | 06-Jul-1997 |
thorpej | branches: 1.23.2;
The fingerprint of (*fr_checkp)() is the same if compiling in kernel
or user code.
|
| 1.22 | 06-Jul-1997 |
thorpej | Restore original RCS IDs.
|
| 1.21 | 06-Jul-1997 |
thorpej | - Add a prototype for fixskip() so that this file compiles.
- Fix, ONCE AGAIN, semantics of ipfilterattach(). This time, not only
was it semantically broken, it wasn't even close to compiling!
|
| 1.20 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.19 | 01-Jun-1997 |
thorpej | In ipl_disable(), don't conditionalize the "fr_checkp = fr_savep"
operation, since:
- in ipl_enable(), "fr_savep = fr_checkp" is not conditionalized
in the same way (not at all), and
- without this change, it was not possible to enable, disable,
and reenable ipfilter.
|
| 1.18 | 28-May-1997 |
thorpej | Put the #ifndef _KERNEL prototype of get_unit() in <netinet/ip_fil.h>
since it is needed by other files, in order to compile on 64-bit
architectures.
|
| 1.17 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.16 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures again:
- Fix a really obvious error: ipl_enable() disappeared, but the guts of
the function were scrunched into the "no-op" BSD pseudo-device attach
routine. Would not compile, because of non-void return from a void
function. Fixed by reincarnating ipl_enable(), and reimplementing
the no-op pseudo-device attach.
- #ifdef as appropriate to remove unused variable warnings.
- Call ipl_enable() in iplinit(), rather than the no-op ipfilterattach().
|
| 1.15 | 26-May-1997 |
darrenr | remove extra #endif
|
| 1.14 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.13 | 15-Apr-1997 |
christos | - Fix indentation of the nested conditionals. It was inconsistent in places.
- Make this compile and work without IPFILTER_LOG, and disable logging by
default. This can be re-enabled now as a kernel option.
|
| 1.12 | 04-Apr-1997 |
cgd | include <stdlib.h> if !_KERNEL for malloc declaration/proto
|
| 1.11 | 03-Apr-1997 |
cgd | fix ... potentially fatal typo (s/unix/unit/)
|
| 1.10 | 01-Apr-1997 |
augustss | Make it compile again by removing a cast to void of KFREE(). KFREE expands
to a statement, not an expression.
|
| 1.9 | 29-Mar-1997 |
darrenr | use IPLLOG instead of ipllog to easily mask parameters, fix up prototype
problems for compiling to user programs.
|
| 1.8 | 29-Mar-1997 |
thorpej | Centralize the check for NetBSD PFIL_HOOKS code into ip_fil.h, and use
it consistently.
|
| 1.7 | 29-Mar-1997 |
thorpej | Fix an ... interesting bug that resulted from namespace collision.
Description:
- A BSD pseudo-device initialization routine is declared as
void <pseudo-device name>attach __P((int count));
in ioconf.c by config(8). main() calls these functions
from a table.
- IP Filter has functions iplattach() and ipldetach() (or,
in the NetBSD case, were erroneously renamed ipfilterattach()
and ipfilterdetach()). These functions are used to establish
and disestablish the IP Filter "filter rule check" hook in
the IP input/output stream. They are declared:
int iplattach __P((void));
int ipldetach __P((void));
..and are expected to return a value by iplioctl().
- When main() calls (by sheer coincidence!) iplattach(),
the filter hook is established, and the IP Filter machinery
labeled as "initialized". This causes all packets, whether or
not the user intents to use filter rules, to be passed to
the filter rule checker if "ipfilter" is configured into the
kernel.
- As a result of the above, a kludge existed to default to
passing all packets (I can only assume that when this was
originally committed, the symptom of the bug was noticed by
the integrator, but the bug not actually found/fixed).
- In iplioctl(), if the SIOCFRENB ioctl is issued with an
argument of "enable" (i.e. user executed "ipf -E"), iplattach()
will notice that the machinery is already initialized and
return EBUSY.
Fix:
- Rename iplattach()/ipldetach() to ipl_enable() and ipl_disable().
- Create a pseudo-device entry stub named ipfilterattach()
(NetBSD case) or iplattach() (all other). This is a noop; none
of the machinery should be initialized until the caller expicitly
enables the filter with ipf -E. Add a comment to note that.
|
| 1.6 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.5 | 18-Mar-1997 |
cgd | ioctl commands are u_longs
|
| 1.4 | 08-Jan-1997 |
veego | ipl[attach|detach]->ipfilter[attach|detach] for the pseudo-device change
|
| 1.3 | 07-Jan-1997 |
mrg | remove some old debugging statements.
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.24 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.23 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.22 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.21 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.20 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.19 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.18 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.17 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.16 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.15 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.14 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.13 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.12 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.11 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.10 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.23.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.24.2.5 | 29-Nov-1998 |
cgd | pull up rev 1.32 from trunk (mrg)
|
| 1.24.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.31 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.24.2.3 | 22-Jul-1998 |
mellon | Pull up 1.28 (veego)
|
| 1.24.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.24.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.34.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.34.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.34.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.34.2.5 | 14-Dec-2000 |
he | Apply patch (requested by he):
Fix problem causing only first ipnat rule to be loaded.
Fixes PR#11569.
|
| 1.34.2.4 | 09-Aug-2000 |
thorpej | Pull up rev. 1.57, as was done on netbsd-1-5 branch, to
fix a compilation problem in userland.
|
| 1.34.2.3 | 02-Aug-2000 |
he | Pull up revision 1.56 (via patch, requested by thorpej):
Properly report initialization error. Enable IPF automatically
if it wasn't already enabled before doing a NAT operation.
|
| 1.34.2.2 | 20-Dec-1999 |
he | Pull up patch (requested by he):
Make this compile on the netbsd-1-4 branch.
|
| 1.34.2.1 | 20-Dec-1999 |
he | Pull up revision 1.35 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.36.2.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.37.2.5 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.37.2.4 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.37.2.3 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.37.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.37.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.53.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.55.2.6 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.55.2.5 | 09-Feb-2002 |
he | Pull up revisions 1.59-1.75 (via patch, requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.55.2.4 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.55.2.3 | 31-Aug-2000 |
sommerfeld | Fill in the next mtu field of a NEEDFRAG ICMP error message.
From marc horowitz, pr10857. approved by *someone* on releng-1-5 a while back
|
| 1.55.2.2 | 07-Aug-2000 |
thorpej | Pull up rev. 1.57, to fix userland compilation problem,
as noted by Matthias Drochner.
|
| 1.55.2.1 | 06-Aug-2000 |
thorpej | Pull up rev. 1.56:
- ipl_enable(): -1 is not an error return. If initializing IP Filter
fails, return EIO instead.
- iplioctl(): If performing a NAT operation, and IP Filter is not
yet initialized (e.g. by `ipf -E'), enable it implicitly before
doing the NAT operation.
|
| 1.65.2.14 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.65.2.13 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.65.2.12 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.65.2.11 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.10 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.9 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.65.2.8 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.65.2.7 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.6 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.65.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.65.2.4 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.65.2.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.65.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.65.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.68.4.2 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.68.4.1 | 07-Sep-2001 |
thorpej | Commit my "devvp" changes to the thorpej-devvp branch. This
replaces the use of dev_t in most places with a struct vnode *.
This will form the basic infrastructure for real cloning device
support (besides being architecurally cleaner -- it'll be good
to get away from using numbers to represent objects).
|
| 1.68.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.68.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.68.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.68.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.79.4.2 | 15-Aug-2003 |
tron | Pull up revision 1.94 (requested by martti in ticket #1411):
Fix return-rst for IPv6 (PR#22157 by Peter Postma).
|
| 1.79.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.79.2.5 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.79.2.4 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.79.2.3 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.79.2.2 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.79.2.1 | 16-May-2002 |
gehenna | Add the character device switch.
|
| 1.93.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.93.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.57 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.56 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.55 | 10-May-2004 |
christos | PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.54 | 28-Mar-2004 |
martti | branches: 1.54.2;
Upgraded IPFilter to 4.1.1
|
| 1.53 | 24-Feb-2004 |
wiz | occured -> occurred. From Peter Postma.
|
| 1.52 | 29-Jun-2003 |
fvdl | branches: 1.52.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.51 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.50 | 26-Jun-2003 |
itojun | tabify
|
| 1.49 | 29-Sep-2002 |
martti | Remove unused ipl_usec.
|
| 1.48 | 25-Sep-2002 |
martti | Fix ipmon problems on 64-bit platforms (PR#17403 and PR#17404).
|
| 1.47 | 19-Sep-2002 |
martti | Resync with official IPF
|
| 1.46 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.45 | 01-Jul-2002 |
christos | Fix iplog problem on sparc64 [from Tomi Nylund]
1. size_t is 64 bits, so use a u_32_t for iplused
2. microtime() and friends expect a struct timeval,
passing the first of two unsigned longs will not cut it.
|
| 1.44 | 02-May-2002 |
martti | branches: 1.44.2; 1.44.4;
Upgraded IPFilter to 3.4.27
|
| 1.43 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.42 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.41 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.40 | 16-Sep-2001 |
wiz | Spell 'occurred' with two 'r's.
|
| 1.39 | 26-Mar-2001 |
mike | branches: 1.39.2; 1.39.4;
Resolve conflicts.
|
| 1.38 | 11-Nov-2000 |
thorpej | branches: 1.38.2;
Restructure the PFIL_HOOKS mechanism a bit:
- All packets are passed to PFIL_HOOKS as they come off the wire, i.e.
fields in protocol headers in network order, etc.
- Allow for multiple hooks to be registered, using a "key" and a "dlt".
The "dlt" is a BPF data link type, indicating what type of header is
present.
- INET and INET6 register with key == AF_INET or AF_INET6, and
dlt == DLT_RAW.
- PFIL_HOOKS now take an argument for the filter hook, and mbuf **,
an ifnet *, and a direction (PFIL_IN or PFIL_OUT), thus making them
less IP (really, IP Filter) centric.
Maintain compatibility with IP Filter by adding wrapper functions for
IP Filter.
|
| 1.37 | 12-Jun-2000 |
veego | branches: 1.37.2;
Resolve conflicts.
|
| 1.36 | 23-May-2000 |
veego | branches: 1.36.2;
Resolve conflicts.
|
| 1.35 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.34 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.33 | 28-Dec-1999 |
darrenr | update ipfilter code to 3.3.6
|
| 1.32 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.31 | 11-Dec-1998 |
mrg | branches: 1.31.2; 1.31.8; 1.31.14;
remove this insanity. appeared with ipfilter 3.2.10...
|
| 1.30 | 11-Dec-1998 |
drochner | correction to the previous: protect against _LKM too
pointed out by Todd Whitesel <toddpw@best.com>
|
| 1.29 | 11-Dec-1998 |
drochner | correcton tp previous: don't try to include kernel option headers in
userland
fixes PR kern/6561 (Takahiro Kambe)
|
| 1.28 | 10-Dec-1998 |
christos | defopt
|
| 1.27 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.26 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.25 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.24 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.23 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.22 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.21 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.20 | 21-Sep-1997 |
veego | branches: 1.20.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.19 | 08-Jul-1997 |
mrg | branches: 1.19.2;
put back IPFILTER_DEFAULT_BLOCK, as documented in options(4).
|
| 1.18 | 07-Jul-1997 |
veego | Use FR_PASS for IPF_DEFAULT_PASS. This can be overwritten with an
options IPF_DEFAULT_PASS=FR_BLOCK in your config file.
|
| 1.17 | 06-Jul-1997 |
thorpej | Restore original RCS IDs.
|
| 1.16 | 06-Jul-1997 |
thorpej | - Add a missing #ifdef SOLARIS
- Properly prototype ipfilterattach()/iplattach().
|
| 1.15 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.14 | 28-May-1997 |
thorpej | Put the #ifndef _KERNEL prototype of get_unit() in <netinet/ip_fil.h>
since it is needed by other files, in order to compile on 64-bit
architectures.
|
| 1.13 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.12 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures again:
- Don't prototype functions that don't exist, and do prototype those
that do.
- Get ioctl arguments right (cmd is a u_long in NetBSD).
|
| 1.11 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.10 | 15-Apr-1997 |
christos | - Fix indentation of the nested conditionals. It was inconsistent in places.
- Make this compile and work without IPFILTER_LOG, and disable logging by
default. This can be re-enabled now as a kernel option.
|
| 1.9 | 29-Mar-1997 |
thorpej | Define control device names here; they're needed by kernel and userland.
|
| 1.8 | 29-Mar-1997 |
darrenr | use IPLLOG instead of ipllog to easily mask parameters, fix up prototype
problems for compiling to user programs.
|
| 1.7 | 29-Mar-1997 |
thorpej | Centralize the check for NetBSD PFIL_HOOKS code into ip_fil.h, and use
it consistently.
|
| 1.6 | 29-Mar-1997 |
thorpej | Fix an ... interesting bug that resulted from namespace collision.
Description:
- A BSD pseudo-device initialization routine is declared as
void <pseudo-device name>attach __P((int count));
in ioconf.c by config(8). main() calls these functions
from a table.
- IP Filter has functions iplattach() and ipldetach() (or,
in the NetBSD case, were erroneously renamed ipfilterattach()
and ipfilterdetach()). These functions are used to establish
and disestablish the IP Filter "filter rule check" hook in
the IP input/output stream. They are declared:
int iplattach __P((void));
int ipldetach __P((void));
..and are expected to return a value by iplioctl().
- When main() calls (by sheer coincidence!) iplattach(),
the filter hook is established, and the IP Filter machinery
labeled as "initialized". This causes all packets, whether or
not the user intents to use filter rules, to be passed to
the filter rule checker if "ipfilter" is configured into the
kernel.
- As a result of the above, a kludge existed to default to
passing all packets (I can only assume that when this was
originally committed, the symptom of the bug was noticed by
the integrator, but the bug not actually found/fixed).
- In iplioctl(), if the SIOCFRENB ioctl is issued with an
argument of "enable" (i.e. user executed "ipf -E"), iplattach()
will notice that the machinery is already initialized and
return EBUSY.
Fix:
- Rename iplattach()/ipldetach() to ipl_enable() and ipl_disable().
- Create a pseudo-device entry stub named ipfilterattach()
(NetBSD case) or iplattach() (all other). This is a noop; none
of the machinery should be initialized until the caller expicitly
enables the filter with ipf -E. Add a comment to note that.
|
| 1.5 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.4 | 19-Feb-1997 |
scottr | Don't include ipfilter.h if building an LKM.
|
| 1.3 | 18-Feb-1997 |
mrg | pseudo-device ipfilter brings in PFIL_HOOKS.
|
| 1.2 | 05-Jan-1997 |
veego | branches: 1.2.4;
Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.25 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.24 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.23 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.22 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.21 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.20 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.19 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.18 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.17 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.16 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.15 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.14 | 28-Dec-1999 |
darrenr | update DARRENR branch of netinet to 3.3.6
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.2.4.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.19.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.20.2.5 | 12-Dec-1998 |
cgd | patch to keep opt_pfil_hooks.h from being required. it's not created on
1.3.x. (mrg)
|
| 1.20.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.27 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.20.2.3 | 22-Jul-1998 |
mellon | Pull up 1.26 (veego)
|
| 1.20.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.20.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.31.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.31.8.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.31.8.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.31.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.31.2.2 | 08-Jan-2000 |
he | Pull up revision 1.33 (requested by darrenr):
Update IPF to version 3.3.6.
|
| 1.31.2.1 | 20-Dec-1999 |
he | Pull up revision 1.32 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.36.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.37.2.2 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.37.2.1 | 09-Feb-2002 |
he | Pull up revisions 1.38-1.42 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.38.2.8 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.38.2.7 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.38.2.6 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.38.2.5 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.38.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.38.2.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.38.2.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.38.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.39.4.2 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.39.4.1 | 07-Sep-2001 |
thorpej | Commit my "devvp" changes to the thorpej-devvp branch. This
replaces the use of dev_t in most places with a struct vnode *.
This will form the basic infrastructure for real cloning device
support (besides being architecurally cleaner -- it'll be good
to get away from using numbers to represent objects).
|
| 1.39.2.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.39.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.39.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.39.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.39.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.44.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.44.2.1 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.52.2.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.52.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.52.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.52.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.52.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.54.2.2 | 13-Aug-2004 |
jmc | branches: 1.54.2.2.2;
Pullup rev 1.56 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.54.2.1 | 30-May-2004 |
tron | Pull up revision 1.55 (requested by christos in ticket #416):
PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.54.2.2.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.57 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.17 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.16 | 06-Sep-2004 |
yamt | fr_check_wrapper: as ipf modifies application data as well when
doing application proxy, it's needed to ensure that the whole packet
is writable here.
|
| 1.15 | 06-Sep-2004 |
yamt | fr_check_wrapper, fr_check_wrapper6:
ensure that mbufs are writable beforehand as ipf assumes.
PR/26773 and PR/26850.
|
| 1.14 | 04-Aug-2004 |
christos | remove the avail = 0; assignment which is superfluous. pointed out by enami.
|
| 1.13 | 03-Aug-2004 |
christos | PR/26471: Arto Selonen: ipfilter 4.1.3 crashes the system every few hours
Remove extraneous m = NULL assignment that will cause a NULL dereference
later.
|
| 1.12 | 23-Jul-2004 |
martti | branches: 1.12.2;
Upgraded IPFilter to 4.1.3
|
| 1.11 | 16-Jun-2004 |
tron | Don't leak mbuf if ipfr_fastroute6() fails.
Reviewed by Steve Woodford.
|
| 1.10 | 20-May-2004 |
christos | PR/25622: IPV6 return RST and through cloned interfaces was broken.
- checksum was computed incorrectly.
- ipv6 packet was not initialized properly.
- fixed code to be more similar to the v4 counterpart.
|
| 1.9 | 18-May-2004 |
christos | - remove superfluous assignment
- rt_gateway is already a pointer to struct sockaddr; don't take its address
when assigning it to struct sockaddr_in *
|
| 1.8 | 09-May-2004 |
taca | Make it comiple without warning; void function fr_checkv4sum() and
fr_checkv6sum() should not return value.
|
| 1.7 | 09-May-2004 |
christos | PR/24981: Steven M. Bellovin: ipfilter in 2.0 branch panics the system
patch applied.
|
| 1.6 | 09-May-2004 |
christos | PR/25332: HIROSE yuuji: "fastroute(to)" in ipf.conf doesn't work; patch applied
|
| 1.5 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.4 | 01-Apr-2004 |
martin | Untangle ioctl copyin/copyout confusion. IP-Filter now actually works
on sparc64 (and probably everywhere else).
|
| 1.3 | 28-Mar-2004 |
martti | branches: 1.3.2;
Sync with official IPFilter
|
| 1.2 | 28-Mar-2004 |
martti | Upgraded IPFilter to 4.1.1
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.1 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.3.2.11 | 12-Nov-2004 |
jmc | branches: 1.3.2.11.2;
Pullup patch (requested by darrenr in ticket #934)
build a new fr_info_t structure in fr_send_ip() and pass it through to
the fastroute function so that it uses accurate packet information about
the packet being sent out rather than the packet received (impacts both
return-rst and return-icmp features.) PR#27093
|
| 1.3.2.10 | 11-Sep-2004 |
he | Pull up revision 1.16 (requested by yamt in ticket #833):
Ensure that the whole packet is writable when IPF wants to
modify application data.
|
| 1.3.2.9 | 11-Sep-2004 |
he | Pull up revision 1.15 (requested by yamt in ticket #833):
Ensure that mbufs are writable, as IPF assumes so.
Fixes PR#26773 and PR#26850.
|
| 1.3.2.8 | 13-Aug-2004 |
jmc | Pullup rev 1.12-1.14 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.3.2.7 | 18-Jun-2004 |
grant | Pull up revision 1.11 (requested by tron in ticket #501):
Don't leak mbuf if ipfr_fastroute6() fails.
|
| 1.3.2.6 | 30-May-2004 |
tron | Pull up revision 1.10 (requested by christos in ticket #416):
PR/25622: IPV6 return RST and through cloned interfaces was broken.
- checksum was computed incorrectly.
- ipv6 packet was not initialized properly.
- fixed code to be more similar to the v4 counterpart.
|
| 1.3.2.5 | 30-May-2004 |
tron | Pull up revision 1.9 (requested by christos in ticket #416):
- remove superfluous assignment
- rt_gateway is already a pointer to struct sockaddr; don't take its address
when assigning it to struct sockaddr_in *
|
| 1.3.2.4 | 30-May-2004 |
tron | Pull up revision 1.8 (requested by christos in ticket #416):
Make it comiple without warning; void function fr_checkv4sum() and
fr_checkv6sum() should not return value.
|
| 1.3.2.3 | 30-May-2004 |
tron | Pull up revision 1.7 (requested by christos in ticket #416):
PR/24981: Steven M. Bellovin: ipfilter in 2.0 branch panics the system
patch applied.
|
| 1.3.2.2 | 30-May-2004 |
tron | Pull up revision 1.6 (requested by christos in ticket #416):
PR/25332: HIROSE yuuji: "fastroute(to)" in ipf.conf doesn't work; patch applied
|
| 1.3.2.1 | 01-Apr-2004 |
tron | Pull up revision 1.4 (requested by martin in ticket #39):
Untangle ioctl copyin/copyout confusion. IP-Filter now actually works
on sparc64 (and probably everywhere else).
|
| 1.3.2.11.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.17 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.12.2.7 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.12.2.6 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.12.2.5 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.12.2.4 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.12.2.3 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.12.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.12.2.1 | 23-Jul-2004 |
skrll | file ip_fil_netbsd.c was added on branch ktrace-lwp on 2004-08-03 10:54:38 +0000
|
| 1.86 | 29-Jun-2024 |
riastradh | netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.85 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.84 | 15-Feb-2021 |
knakahara | Fix build failure for options GATEWAY.
|
| 1.83 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.82 | 11-Apr-2018 |
maxv | branches: 1.82.14;
Remove whitespaces/tabs, and one non-ASCII character.
|
| 1.81 | 17-Nov-2017 |
ozaki-r | branches: 1.81.2;
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
|
| 1.80 | 07-Feb-2017 |
ozaki-r | branches: 1.80.6;
Add missing NULL checks for m_get_rcvif
|
| 1.79 | 11-Jan-2017 |
ozaki-r | branches: 1.79.2;
Get rid of unnecessary header inclusions
|
| 1.78 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.77 | 18-Oct-2016 |
ozaki-r | Don't hold global locks if NET_MPSAFE is enabled
If NET_MPSAFE is enabled, don't hold KERNEL_LOCK and softnet_lock in
part of the network stack such as IP forwarding paths. The aim of the
change is to make it easy to test the network stack without the locks
and reduce our local diffs.
By default (i.e., if NET_MPSAFE isn't enabled), the locks are held
as they used to be.
Reviewed by knakahara@
|
| 1.76 | 01-Aug-2016 |
knakahara | improve fast-forward performance when the number of flows exceeds IPFLOW_MAX.
In the fast-forward case, when the number of flows exceeds IPFLOW_MAX, the
performmance degraded to about 50% compared to the case less than IPFLOW_MAX
flows. This modification suppresses the degradation to 65%. Furthermore,
the modified kernel is about the same performance as the original kernel
when the number of flows is less than IPFLOW_MAX.
The original patch is implemented by ryo@n.o. Thanks.
|
| 1.75 | 27-Jul-2016 |
knakahara | remove extra ifdefs. no functional changes.
ip_flow.c becomes build target only if GATEWAY kernel option is on.
So, "#ifdef GATEWAY" in ip_flow.c is not needed.
|
| 1.74 | 26-Jul-2016 |
ozaki-r | Simplify by using atomic_swap instead of mutex
Suggested by kefren@
|
| 1.73 | 11-Jul-2016 |
ozaki-r | branches: 1.73.2;
Run timers in workqueue
Timers (such as nd6_timer) typically free/destroy some data in callout
(softint). If we apply psz/psref for such data, we cannot do free/destroy
process in there because synchronization of psz/psref cannot be used in
softint. So run timer callbacks in workqueue works (normal LWP context).
Doing workqueue_enqueue a work twice (i.e., call workqueue_enqueue before
a previous task is scheduled) isn't allowed. For nd6_timer and
rt_timer_timer, this doesn't happen because callout_reset is called only
from workqueue's work. OTOH, ip{,6}flow_slowtimo's callout can be called
before its work starts and completes because the callout is periodically
called regardless of completion of the work. To avoid such a situation,
add a flag for each protocol; the flag is set true when a work is
enqueued and set false after the work finished. workqueue_enqueue is
called only if the flag is false.
Proposed on tech-net and tech-kern.
|
| 1.72 | 20-Jun-2016 |
knakahara | apply if_output_lock() to L3 callers which call ifp->if_output() of L2(or L3 tunneling).
|
| 1.71 | 13-Jun-2016 |
knakahara | eliminate unnecessary splnet
|
| 1.70 | 13-Jun-2016 |
knakahara | MP-ify fastforward to support GATEWAY kernel option.
I add "ipflow_lock" mutex in ip_flow.c and "ip6flow_lock" mutex in ip6_flow.c
to protect all data in each file. Of course, this is not MP-scalable. However,
it is sufficient as tentative workaround. We should make it scalable somehow
in the future.
ok by ozaki-r@n.o.
|
| 1.69 | 13-Jun-2016 |
knakahara | make ipflow_reap() static function.
|
| 1.68 | 13-Jun-2016 |
knakahara | remove unnecessary splnet before pool_{get,put}
|
| 1.67 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.66 | 23-Mar-2015 |
roy | Add RTF_BROADCAST to mark routes used for the broadcast address when
they are created on the fly. This makes it clear what the route is for
and allows an optimisation in ip_output() by avoiding a call to
in_broadcast() because most of the time we do talk to a host.
It also avoids a needless allocation for the storage of llinfo_arp and
thus vanishes from arp(8) - it showed as incomplete anyway so this
is a nice side effect.
Guard against this and routes marked with RTF_BLACKHOLE in
ip_fastforward().
While here, guard against routes marked with RTF_BLACKHOLE in
ip6_fastforward().
RTF_BROADCAST is IPv4 only, so don't bother checking that here.
|
| 1.65 | 18-Oct-2014 |
snj | branches: 1.65.2;
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.64 | 22-May-2014 |
rmind | branches: 1.64.2;
- Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.63 | 01-Apr-2014 |
pooka | branches: 1.63.2;
Wrap ipflow_create() & ip6flow_create() in kernel lock. Prevents the
interrupt side on another core from seeing the situation while the ipflow
is being modified.
|
| 1.62 | 19-Mar-2014 |
liamjfoy | Move ipflow into ip_var.h and fix confliction
|
| 1.61 | 19-Mar-2014 |
liamjfoy | Remove ipflow_prune and replace with ipflow_reap. ok rmind@
|
| 1.60 | 19-Jan-2012 |
liamjfoy | branches: 1.60.6; 1.60.10;
Remove ipf_start from ipf struct
|
| 1.59 | 01-Apr-2010 |
tls | branches: 1.59.8; 1.59.12;
After discussion with ad@: it appears that KERNEL_LOCK also protects
the driver output path (that is, ifp->if_output()). In the case of
entry through the socket code, we are fine, because pru_usrreq takes
KERNEL_LOCK. However, there are a few other ways to cause output
which require protection:
1) direct calls to tcp_output() in tcp_input()
2) fast-forwarding code (ip_flow) -- protected elsewise
against itself by the softnet lock.
3) *Possibly* the ARP code. I have currently persuaded
myself that it is safe because of how it's called.
4) Possibly the ICMP code.
This change addresses #1 and #2.
|
| 1.58 | 15-Mar-2009 |
cegger | branches: 1.58.2; 1.58.4;
ansify function definitions
|
| 1.57 | 01-Feb-2009 |
pooka | branches: 1.57.2;
Init ipflow pool dynamically instead of using a linkset.
|
| 1.56 | 28-Apr-2008 |
martin | branches: 1.56.8;
Remove clause 3 and 4 from TNF licenses
|
| 1.55 | 24-Apr-2008 |
ad | branches: 1.55.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.54 | 12-Apr-2008 |
thorpej | branches: 1.54.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.53 | 09-Apr-2008 |
thorpej | - ipflow is not used outside ip_flow.c; move its definition there.
- Make ipflow_reap() private to ip_flow.c, and introduce ipflow_prune()
for external callers to use (avoids returning an ipflow * that is never
actually used anyway).
|
| 1.52 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.51 | 04-Jan-2008 |
dyoung | branches: 1.51.6;
Constify a bit.
|
| 1.50 | 04-Jan-2008 |
dyoung | Replace rtcache_down() with rtcache_validate() and update rtcache_down()
uses.
|
| 1.49 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.48 | 20-Aug-2007 |
dyoung | branches: 1.48.2; 1.48.8; 1.48.10; 1.48.14;
Don't call rtcache_check() from the fast-forward code, which runs
at IPL_NET, because rtcache_check() may read the forwarding table.
Elsewhere, the kernel only blocks interrupts at priority IPL_SOFTNET
and below while it modifies the forwarding table, so rtcache_check()
could be reading the table in an inconsistent state. Use
rtcache_done(), instead.
XXX netinet/ip_flow.c and netinet6/ip6_flow.c are virtually identical.
XXX They should share code.
|
| 1.47 | 02-May-2007 |
dyoung | branches: 1.47.2; 1.47.6;
Remove obsolete files netinet/in_route.[ch].
|
| 1.46 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.45 | 05-Apr-2007 |
liamjfoy | use size_t for indexes
just pass a *ip to ipflow_hash instead of members
ok christos@
|
| 1.44 | 26-Mar-2007 |
liamjfoy | Add a small note regarding further commented code in netinet6/ip6_flow.c
|
| 1.43 | 25-Mar-2007 |
liamjfoy | Add net.inet.ip.hashsize to control the IPv4 fast forward hash table size.
|
| 1.42 | 12-Mar-2007 |
ad | branches: 1.42.2; 1.42.4;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.41 | 04-Mar-2007 |
christos | branches: 1.41.2;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.40 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.39 | 26-Jan-2007 |
dyoung | branches: 1.39.2;
bzero -> memset
|
| 1.38 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.37 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.36 | 06-Oct-2006 |
mrg | add a missing semicolon from the previous commit.
|
| 1.35 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.34 | 02-Sep-2006 |
liamjfoy | branches: 1.34.2; 1.34.4;
increment ips_total too.
ok matt thomas
|
| 1.33 | 07-Jun-2006 |
kardel | merge FreeBSD timecounters from branch simonb-timecounters
- struct timeval time is gone
time.tv_sec -> time_second
- struct timeval mono_time is gone
mono_time.tv_sec -> time_uptime
- access to time via
{get,}{micro,nano,bin}time()
get* versions are fast but less precise
- support NTP nanokernel implementation (NTP API 4)
- further reading:
Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf
NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
|
| 1.32 | 24-Dec-2005 |
perry | branches: 1.32.4; 1.32.6; 1.32.8; 1.32.14;
change comment from __const__ to const
|
| 1.31 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.30 | 17-Oct-2005 |
christos | small list macro cleanup:
- remove duplicate LIST_FIRST (Liam Foy)
- change to use LIST_FOREACH or for () instead of while () for consistency
|
| 1.29 | 03-Feb-2005 |
perry | branches: 1.29.6;
KNF + slightly ANSIfy
|
| 1.28 | 25-Apr-2004 |
simonb | branches: 1.28.4; 1.28.6;
Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.27 | 12-Dec-2003 |
scw | Make fast-ipsec and ipflow (Fast Forwarding) interoperate.
The idea is that we only clear M_CANFASTFWD if an SPD exists
for the packet. Otherwise, it's safe to add a fast-forward
cache entry for the route.
To make this work properly, we invalidate the entire ipflow
cache if a fast-ipsec key is added or changed.
|
| 1.26 | 02-Nov-2002 |
perry | branches: 1.26.6;
/*CONTCOND*/ while (0)'ed macros
|
| 1.25 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.24 | 09-Jun-2002 |
itojun | whitespace
|
| 1.23 | 08-Mar-2002 |
thorpej | branches: 1.23.6;
Pool deals fairly well with physical memory shortage, but it doesn't
deal with shortages of the VM maps where the backing pages are mapped
(usually kmem_map). Try to deal with this:
* Group all information about the backend allocator for a pool in a
separate structure. The pool references this structure, rather than
the individual fields.
* Change the pool_init() API accordingly, and adjust all callers.
* Link all pools using the same backend allocator on a list.
* The backend allocator is responsible for waiting for physical memory
to become available, but will still fail if it cannot callocate KVA
space for the pages. If this happens, carefully drain all pools using
the same backend allocator, so that some KVA space can be freed.
* Change pool_reclaim() to indicate if it actually succeeded in freeing
some pages, and use that information to make draining easier and more
efficient.
* Get rid of PR_URGENT. There was only one use of it, and it could be
dealt with by the caller.
From art@openbsd.org.
|
| 1.22 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.21 | 29-Oct-2001 |
simonb | Don't need to include <uvm/uvm_extern.h> just to include <sys/sysctl.h>
anymore.
|
| 1.20 | 17-Sep-2001 |
thorpej | branches: 1.20.2;
Split the pre-computed ifnet checksum flags into Tx and Rx directions.
Add capabilities bits that indicate an interface can only perform
in-bound TCPv4 or UDPv4 checksums. There is at least one Gig-E chip
for which this is true (Level One LXT-1001), and this is also the
case for the Intel i82559 10/100 Ethernet chips.
|
| 1.19 | 12-Jun-2001 |
wiz | branches: 1.19.2; 1.19.4;
receive, not recieve
|
| 1.18 | 02-Jun-2001 |
thorpej | Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.17 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.16 | 30-Jun-2000 |
thorpej | branches: 1.16.2;
Pass the correct destination address for the route-to-gateway case.
From Zdenek Salvet, kern/10483.
|
| 1.15 | 28-Jun-2000 |
mrg | <vm/vm.h> -> <uvm/uvm_extern.h>
|
| 1.14 | 17-Oct-1999 |
sommerfeld | branches: 1.14.2; 1.14.10;
If a packet came in as link-level broadcast or link-level multicast, don't
attempt to fast-forward it out.
|
| 1.13 | 26-Mar-1999 |
proff | branches: 1.13.2; 1.13.8;
security: test for ip_len < ip_hl <<2 and drop packet accordingly
|
| 1.12 | 28-Jan-1999 |
itohy | ~htons(...) is always negative.
|
| 1.11 | 25-Jan-1999 |
mycroft | One more tweak to the checksum hack, and I promise I'm done. B-)
|
| 1.10 | 25-Jan-1999 |
mycroft | Absolutely minor tweak to generate better code.
|
| 1.9 | 24-Jan-1999 |
mycroft | Update the comment about the checksum hack. It was way out of date.
|
| 1.8 | 24-Jan-1999 |
mycroft | Modify the checksum slightly so that the htons()s can all be combined.
|
| 1.7 | 08-Oct-1998 |
thorpej | Use the pool allocator for ipflow entries.
|
| 1.6 | 10-Jun-1998 |
sommerfe | Truncate mbufs to the correct length before forwarding; fixes pr5560
|
| 1.5 | 02-Jun-1998 |
thorpej | In addition to the IP flow hash table, put the flows on a list. The table
is used for fast lookup, the list for traversal of all flows. Also, use
PRT timers.
|
| 1.4 | 18-May-1998 |
matt | Fix two bugs.
|
| 1.3 | 04-May-1998 |
matt | Default IP flow to being enabled. Add a sysctl to control the maximum
number of flows (net.inet.ip.maxflows). If set to 0, will disable fast
path forwarding.
|
| 1.2 | 04-May-1998 |
thorpej | - kern/5380 (Dennis Ferguson): fix incremental IP header checksum.
- kern/5381 (Dennis Ferguson): check IP header checksum in fast forward
code.
- In ipflow_slowtimo(), if no IP flows are in use, don't bother checking
all of the hash buckets.
|
| 1.1 | 29-Apr-1998 |
matt | Add support for "fast" forwarding. Add hooks in if_ethersubr.c and
if_fddisubr.c to fastpath IP forwarding. If ip_forward successfully
forwards a packet, it will create a cache (ipflow) entry. ether_input
and fddi_input will first call ipflow_fastforward with the received
packet and if the packet passes enough tests, it will be forwarded (the
ttl is decremented and the cksum is adjusted incrementally).
|
| 1.13.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.13.2.2 | 01-Jul-2000 |
he | Pull up revision 1.16 (requested by thorpej):
Pass the correct destination address for the route-to-gateway
case. Fixes PR#10483.
|
| 1.13.2.1 | 18-Oct-1999 |
cgd | pull up rev 1.14 from trunk (requested by sommerfeld):
Multicast storm prevention: don't attempt to forward link-level
multicast packets which contain ip unicast packets; these packets
would only be generated from misconfigured/buggy systems.
|
| 1.14.10.1 | 30-Jun-2000 |
thorpej | Pull up rev. 1.16:
Pass the correct destination address for the route-to-gateway case.
>From Zdenek Salvet, kern/10483.
|
| 1.14.2.2 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.14.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.16.2.7 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.16.2.6 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.16.2.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.16.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.16.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.16.2.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.16.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.19.4.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.19.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.19.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.19.2.2 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.19.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.20.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.23.6.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.23.6.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.26.6.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.26.6.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.26.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.26.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.26.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.28.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.28.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.29.6.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.29.6.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.29.6.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.29.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.29.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.32.14.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.32.8.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.32.8.1 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.32.6.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time(), use bintime's for timeout
calculations and use "time_second" instead of "time.tv_sec".
|
| 1.32.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.34.4.3 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.34.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.34.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.34.2.3 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.34.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.34.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.39.2.5 | 07-May-2007 |
yamt | sync with head.
|
| 1.39.2.4 | 15-Apr-2007 |
yamt | sync with head.
|
| 1.39.2.3 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.39.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.39.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.41.2.4 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.41.2.3 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.41.2.2 | 10-Apr-2007 |
ad | Sync with head.
|
| 1.41.2.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.42.4.1 | 29-Mar-2007 |
reinoud | Pullup to -current
|
| 1.42.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.47.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.47.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.48.14.2 | 08-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.48.14.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.48.10.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.48.8.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.48.2.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.51.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.54.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.55.2.3 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.55.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.55.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.56.8.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.56.8.1 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.57.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.58.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.58.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.59.12.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.59.8.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.59.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.60.10.2 | 18-May-2014 |
rmind | sync with head
|
| 1.60.10.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.60.6.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.60.6.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.63.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.64.2.1 | 12-May-2017 |
snj | Pull up following revision(s) (requested by skrll/ozaki-r in ticket #1402):
sys/net/route.c: revision 1.170 via patch
sys/netinet/ip_flow.c: revision 1.73 via patch
sys/netinet6/ip6_flow.c: revision 1.28 via patch
sys/netinet6/nd6.c: revision 1.203 via patch
Run timers in workqueue
Timers (such as nd6_timer) typically free/destroy some data in callout
(softint). If we apply psz/psref for such data, we cannot do free/destroy
process in there because synchronization of psz/psref cannot be used in
softint. So run timer callbacks in workqueue works (normal LWP context).
Doing workqueue_enqueue a work twice (i.e., call workqueue_enqueue before
a previous task is scheduled) isn't allowed. For nd6_timer and
rt_timer_timer, this doesn't happen because callout_reset is called only
from workqueue's work. OTOH, ip{,6}flow_slowtimo's callout can be called
before its work starts and completes because the callout is periodically
called regardless of completion of the work. To avoid such a situation,
add a flag for each protocol; the flag is set true when a work is
enqueued and set false after the work finished. workqueue_enqueue is
called only if the flag is false.
Proposed on tech-net and tech-kern.
|
| 1.65.2.6 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.65.2.5 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.65.2.4 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.65.2.3 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.65.2.2 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.65.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.73.2.4 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.73.2.3 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.73.2.2 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.73.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.79.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.80.6.1 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.81.2.1 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.82.14.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.37 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.36 | 28-Mar-2004 |
martin | branches: 1.36.4;
Cast 64 bit pointers only with (intptr_t) care.
|
| 1.35 | 28-Mar-2004 |
martti | Upgraded IPFilter to 4.1.1
|
| 1.34 | 19-Sep-2002 |
martti | branches: 1.34.6;
Resync with official IPF
|
| 1.33 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.32 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.31 | 09-Jun-2002 |
itojun | whitespace
|
| 1.30 | 02-May-2002 |
martti | branches: 1.30.2; 1.30.4;
Fix compilation problems
|
| 1.29 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.28 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.27 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.26 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.25 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.24 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.23 | 06-Apr-2001 |
darrenr | branches: 1.23.2;
fix fragment cache security hole
|
| 1.22 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.21 | 12-Jun-2000 |
veego | branches: 1.21.2; 1.21.4;
Ups, forgot to resolve one place.
|
| 1.20 | 12-Jun-2000 |
veego | Resolve conflicts.
|
| 1.19 | 11-May-2000 |
veego | branches: 1.19.2;
Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.18 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.17 | 24-Mar-2000 |
thorpej | Pull in <sys/callout.h> for the benefit of userland.
|
| 1.16 | 23-Mar-2000 |
thorpej | New callout mechanism with two major improvements over the old
timeout()/untimeout() API:
- Clients supply callout handle storage, thus eliminating problems of
resource allocation.
- Insertion and removal of callouts is constant time, important as
this facility is used quite a lot in the kernel.
The old timeout()/untimeout() API has been removed from the kernel.
|
| 1.15 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.14 | 02-Feb-1999 |
cjs | branches: 1.14.2; 1.14.6; 1.14.8; 1.14.14;
Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.13 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.12 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.11 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.10 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.9 | 21-Sep-1997 |
veego | branches: 1.9.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.8 | 06-Jul-1997 |
thorpej | branches: 1.8.2;
Restore original RCS IDs.
|
| 1.7 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.6 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.5 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures again:
- garbage-collect unused variables.
|
| 1.4 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.3 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.21 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.20 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.19 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.18 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.17 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.16 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.15 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.14 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.13 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.12 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.11 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.10 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.8.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.13 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.9.2.3 | 22-Jul-1998 |
mellon | Puil up 1.12 (veego)
|
| 1.9.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.9.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.14.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.14.8.3 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.14.8.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.14.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.14.6.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.14.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.14.2.2 | 14-Apr-2001 |
he | Pull up revision 1.23 (via patch, requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.14.2.1 | 20-Dec-1999 |
he | Pull up revision 1.15 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.19.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.21.4.9 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.21.4.8 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.21.4.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.21.4.6 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.21.4.5 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.21.4.4 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.21.4.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.21.4.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.21.4.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.21.2.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.21.2.2 | 09-Feb-2002 |
he | Pull up revisions 1.22,1.24-1.27 (via patch, requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.21.2.1 | 14-Apr-2001 |
he | Pull up revision 1.23 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.23.2.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.23.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.23.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.23.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.23.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.30.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.30.2.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.30.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.34.6.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.34.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.34.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.34.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.36.4.1 | 06-Feb-2005 |
jmc | Pull up revision 1.37 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.21 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.20 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.19 | 28-Mar-2004 |
martti | branches: 1.19.2;
Upgraded IPFilter to 4.1.1
|
| 1.18 | 19-Sep-2002 |
martti | branches: 1.18.6;
Upgraded IPFilter to 3.4.29
|
| 1.17 | 24-Jan-2002 |
martti | branches: 1.17.10;
Upgraded IPFilter to 3.4.23
|
| 1.16 | 06-Apr-2001 |
darrenr | branches: 1.16.2;
fix fragment cache security hole
|
| 1.15 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.14 | 03-May-2000 |
veego | branches: 1.14.4; 1.14.6;
Resolve conflicts.
|
| 1.13 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.12 | 22-Nov-1998 |
mrg | branches: 1.12.4; 1.12.10; 1.12.16;
merge ipf 3.2.10
|
| 1.11 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.10 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.9 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.8 | 21-Sep-1997 |
veego | branches: 1.8.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.7 | 06-Jul-1997 |
thorpej | branches: 1.7.2;
Restore original RCS IDs.
|
| 1.6 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.5 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.4 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.3 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.19 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.18 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.17 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.16 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.15 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.14 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.7.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.8.2.2 | 24-Nov-1998 |
cgd | pull up rev(s) 1.12 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.8.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.12.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.12.10.3 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.12.10.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.12.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.12.4.2 | 14-Apr-2001 |
he | Pull up revision 1.16 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.12.4.1 | 20-Dec-1999 |
he | Pull up revision 1.13 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.14.6.3 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.14.6.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.14.6.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.14.4.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.14.4.2 | 09-Feb-2002 |
he | Pull up revisions 1.15-1.17 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.14.4.1 | 14-Apr-2001 |
he | Pull up revision 1.16 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.16.2.2 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.16.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.17.10.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.18.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.18.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.19.2.1 | 13-Aug-2004 |
jmc | branches: 1.19.2.1.2;
Pullup rev 1.20 (requested by christos in ticket #1727)
Sync up w. ipf 4.1.3
|
| 1.19.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.21 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.29 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.28 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.27 | 28-Mar-2004 |
martti | branches: 1.27.2;
Upgraded IPFilter to 4.1.1
|
| 1.26 | 19-Sep-2002 |
martti | branches: 1.26.6;
Resync with official IPF
|
| 1.25 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.24 | 02-May-2002 |
martti | branches: 1.24.4;
Fix compilation problems
|
| 1.23 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.22 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.21 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.20 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.19 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.18 | 26-Mar-2001 |
mike | branches: 1.18.2;
Resolve conflicts.
|
| 1.17 | 09-Aug-2000 |
veego | branches: 1.17.2;
Resolve conflicts.
|
| 1.16 | 21-May-2000 |
veego | branches: 1.16.4;
Resolve conflicts.
|
| 1.15 | 11-May-2000 |
veego | Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.14 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.13 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.12 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.11 | 22-Nov-1998 |
mrg | branches: 1.11.4; 1.11.10; 1.11.16;
add two more prototypes. noted missing by mjacob.
|
| 1.10 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.9 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.8 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.7 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.6 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.5 | 21-Sep-1997 |
veego | branches: 1.5.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.4 | 05-Jul-1997 |
darrenr | branches: 1.4.2;
fix conflicts from import
|
| 1.3 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.2 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures:
- Add prototypes.
- garbage-collect unused variables.
|
| 1.1 | 26-May-1997 |
darrenr | branches: 1.1.1;
Initial revision
|
| 1.1.1.22 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.21 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.20 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.19 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.18 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.17 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.16 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.15 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.14 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.13 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.12 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.11 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.10 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.9 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.8 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.7 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.6 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.5 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.4 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.3 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.2 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.1 | 26-May-1997 |
darrenr | Import new sources for 3.2alpha7
(blah, someone want to clean away /cvsroot/sys/netinet ?)
|
| 1.4.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.5.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.10-1.11 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.5.2.3 | 22-Jul-1998 |
mellon | Pull up 1.9 (veego)
|
| 1.5.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.5.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.11.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.11.10.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.11.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.11.4.1 | 20-Dec-1999 |
he | Pull up revision 1.12 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.16.4.4 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.16.4.3 | 09-Feb-2002 |
he | Pull up revisions 1.18-1.21 (via patch, requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.16.4.2 | 05-Apr-2001 |
he | Apply patch (requested by mike):
Fix machine hangs with FTP from behind a NetBSD box configured
to do NAT and FTP proxying. Fixes PR#12443.
|
| 1.16.4.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.17.2.6 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.17.2.5 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.17.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.17.2.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.17.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.17.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.18.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.18.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.18.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.18.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.24.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.26.6.3 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.26.6.2 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.26.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.27.2.1 | 13-Aug-2004 |
jmc | branches: 1.27.2.1.2;
Pullup rev 1.28 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.27.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.29 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.48 | 05-Oct-2007 |
dyoung | Work in progress: use a raw socket for GRE in IP encapsulation
instead of adding/subtracting our own IPv4 header.
There are many benefits: gre(4) needn't grok the outer encapsulation
header any longer, so this simplifies the gre(4) code. The IP
stack needn't grok GRE, so it is simplified, too. gre(4) will
benefit from optimizations in the socket code. Eventually, gre(4)
will gain an IPv6 encapsulation with very few new lines of code.
There is a small performance loss. A 133 MHz, 486-class AMD Elan
sinks/sources a TCP stream over GRE with about 93% the throughput
of the old code. TCP throughput on a 266 MHz, 586-class AMD Geode
is about 96% the throughput of the old code. A 175-MHz ADM5120
(MIPS) only sinks a TCP stream over GRE at about 90% of the old
code; I am still investigating that.
I produced stripped-down versions of sosend() and soreceive() for
gre(4) to use. They are guaranteed not to block, so they can be
called from a software interrupt and from a socket upcall,
respectively.
A kernel thread is no longer necessary for socket transmit/receive,
but I didn't get around to removing it, yet.
Thanks to Matt Thomas for suggesting the use of stripped-down socket
code and software interrupts, and to Andrew Doran for advice and
answers concerning software interrupts, threads, and performance.
|
| 1.47 | 02-Sep-2007 |
dyoung | branches: 1.47.2;
Be consistent: use the prefix sc_ for all members of the gre_softc.
|
| 1.46 | 06-May-2007 |
dyoung | branches: 1.46.2; 1.46.6; 1.46.8;
Oops, commit this straggler from the last change to net/if_gre.[ch].
|
| 1.45 | 21-Mar-2007 |
dyoung | If we do not recognize the protocol of a received packet, then
increase ifi_noproto. If the GRE header contains routing options,
increase the input-error count, ifi_ierrors.
While I am here, make some cosmetic changes: remove unnecessary
'proto' argument from gre_input3(). Shorten some staircases.
|
| 1.44 | 16-Nov-2006 |
dyoung | branches: 1.44.4; 1.44.8; 1.44.10; 1.44.12;
Use LIST_FOREACH().
|
| 1.43 | 16-Nov-2006 |
dyoung | Cosmetic: s/g_proto/sc_proto/. Remove superfluous parentheses and
curly braces.
|
| 1.42 | 07-Sep-2006 |
dogcow | branches: 1.42.2; 1.42.4;
remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.
|
| 1.41 | 31-Aug-2006 |
dyoung | Add a mode to gre(4) that sends GRE tunnel packets in UDP datagrams.
Fix MOBILE encapsulation. Add many debugging printfs (mainly
concerning UDP mode). Clean up the gre(4) code a bit. Add the
capability to setup UDP tunnels to ifconfig. Update documentation.
In UDP mode, gre(4) puts a GRE header onto transmitted packets,
and hands them to a UDP socket for transmission. That is, the
encapsulation looks like this: IP+UDP+GRE+encapsulated packet.
There are two ways to set up a UDP tunnel. One way is to tell the
source and destination IP+port to gre(4), and let gre(4) create
the socket. The other way to create a UDP tunnel is for userland
to "delegate" a UDP socket to the kernel.
|
| 1.40 | 28-Jul-2006 |
dyoung | Extract predicate M_UNWRITABLE(m, len), which is true iff len
consecutive bytes at the front of m are writable (i.e., neither
shared nor read-only).
|
| 1.39 | 28-Jul-2006 |
dyoung | Fix mtod() usage. If we will write to the mbuf data, check whether
the data is read-only/shared and call m_pullup(). Otherwise,
extract a const pointer to the mbuf data.
XXX I should extract a new macro, M_WRITABLE(m, len), that is true
if m has len consecutive writable bytes at its front.
KNF slightly.
Use bpf_mtap_af().
|
| 1.38 | 28-Jul-2006 |
dyoung | Use bpf_mtap_af(). KNF slightly.
|
| 1.37 | 31-Jan-2006 |
elad | branches: 1.37.2; 1.37.6;
fix tyop.
pr 32678 from yves emmanuel jutard.
|
| 1.36 | 11-Dec-2005 |
christos | branches: 1.36.2;
merge ktrace-lwp.
|
| 1.35 | 26-Jul-2005 |
christos | PR/30844: Gert Doering: Non-inet traffic is passed to bpf incorrectly (as inet)
|
| 1.34 | 30-Mar-2005 |
is | branches: 1.34.2;
Add IPv6 over GRE (contributed by Gert Doering in PR 29150).
|
| 1.33 | 26-Feb-2005 |
perry | branches: 1.33.2;
nuke trailing whitespace
|
| 1.32 | 03-Feb-2005 |
perry | some ANSIfying, and remove an unsightly tab
|
| 1.31 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.30 | 26-Apr-2004 |
matt | branches: 1.30.4; 1.30.6;
Remove #else clause of __STDC__
|
| 1.29 | 05-Sep-2003 |
itojun | branches: 1.29.4;
u_short -> u_int16_t
|
| 1.28 | 26-Jun-2003 |
itojun | branches: 1.28.2;
tabify
|
| 1.27 | 02-May-2003 |
itojun | clear m_flags just for safety
|
| 1.26 | 21-Apr-2003 |
itojun | remove redundant adjustment of m->m_pkthdr.len
|
| 1.25 | 21-Apr-2003 |
itojun | correct arg to m_pullup (need to count IP header size as well)
|
| 1.24 | 21-Apr-2003 |
itojun | correct (false) assumptions on mbuf chain. not sure if it really helps, but
anyways, it is necessary to perform m_pullup.
|
| 1.23 | 25-Nov-2002 |
simonb | The "osrc" variable in gre_mobile_input() is only ever set but not
referenced; remove it.
|
| 1.22 | 11-Sep-2002 |
itojun | KNF - return is not a function. sync w/kame.
|
| 1.21 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.20 | 10-Aug-2002 |
itojun | inject GRE packet to raw ip socket input, to support userland GRE decapsulator.
discussed on openbsd developers list.
|
| 1.19 | 09-Jun-2002 |
itojun | style
|
| 1.18 | 09-Jun-2002 |
itojun | whitespace
|
| 1.17 | 13-Nov-2001 |
lukem | branches: 1.17.8;
add RCSIDs
|
| 1.16 | 13-Apr-2001 |
thorpej | branches: 1.16.2;
Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.15 | 12-Dec-2000 |
thorpej | branches: 1.15.2;
Adapt to bpfattach() changes, and further centralize the bpfattach()
and bpfdetach() calls into link-type subroutines where possible.
|
| 1.14 | 20-Oct-2000 |
mjl | Mark packets from gre as coming from appropriate gre interface, not
transport interface.
|
| 1.13 | 25-Aug-2000 |
mjl | Add bpf tap to gre interface.
|
| 1.12 | 06-Jul-2000 |
thorpej | Some slight cleanup.
|
| 1.11 | 05-Jul-2000 |
thorpej | Fix an omission in the gre cloning changes.
|
| 1.10 | 30-Mar-2000 |
augustss | branches: 1.10.4;
Remove register declarations.
|
| 1.9 | 25-Oct-1999 |
drochner | defopt the XNS protocol (options NS), clean up the use of related
option headers / defines
|
| 1.8 | 19-Jan-1999 |
mycroft | branches: 1.8.8; 1.8.10; 1.8.12;
Don't screw with ip_len; just subtract from it where we actually use the
value.
|
| 1.7 | 11-Jan-1999 |
thorpej | Pull the IP-in-IP tunneling support out of the GRE code. It's not handled
by a separate IP-IP input path.
XXX Should eventually do the same thing for IPPROTO_MOBILE.
|
| 1.6 | 22-Dec-1998 |
thorpej | Simplify the tunnel lookup routine.
|
| 1.5 | 13-Oct-1998 |
kim | Use ETHERTYPE_ATALK instead of ETHERTYPE_AT. The former seems more common.
Our other constants also use "ATALK".
Added many new ETHERTYPE constants to sys/net/ethertypes.h, including the
ones from libpcap and tcpdump "ethertype.h" files.
|
| 1.4 | 07-Oct-1998 |
thorpej | Fix some typos in comments, and clean up some whitespace.
|
| 1.3 | 02-Oct-1998 |
kleink | Use #error instead of causing a parse error.
|
| 1.2 | 30-Sep-1998 |
hwr | Start supporting IPPROTO_MOBILE (55) encapsulation. This is yet
another tunneling protocol used by the Mobile-IP people. See RFC 2004
for this.
|
| 1.1 | 13-Sep-1998 |
hwr | Add a gre tunnel pseudo network device. Gre = generic route encapsulation.
This device shows up like any other network interface and can be used to
tunnel L3 protocols as e.g. IP over IP.
|
| 1.8.12.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.8.10.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.8.8.3 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.8.8.2 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.8.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.10.4.2 | 20-Oct-2000 |
tv | Pullup 1.14 [mjl]:
Mark packets from gre as coming from appropriate gre interface, not
transport interface.
|
| 1.10.4.1 | 25-Aug-2000 |
mjl | Add bpf tap to gre interfaces. Approved by thorpej.
|
| 1.15.2.7 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.15.2.6 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.15.2.5 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.15.2.4 | 13-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.15.2.3 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.15.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.15.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.16.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.16.2.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.16.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.16.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.17.8.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.17.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.28.2.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.28.2.6 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.28.2.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.28.2.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.28.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.28.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.28.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.29.4.1 | 08-May-2005 |
snj | Pull up revision 1.34 (requested by is in ticket #1382):
Add IPv6 over GRE (contributed by Gert Doering in PR 29150).
|
| 1.30.6.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.30.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.30.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.33.2.1 | 30-Mar-2005 |
tron | Pull up revision 1.34 (requested by is in ticket #80):
Add IPv6 over GRE (contributed by Gert Doering in PR 29150).
|
| 1.34.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.34.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.34.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.34.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.36.2.1 | 01-Feb-2006 |
yamt | sync with head.
|
| 1.37.6.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.37.6.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.37.6.1 | 11-Aug-2006 |
yamt | sync with head
|
| 1.37.2.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.42.4.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.42.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.44.12.1 | 29-Mar-2007 |
reinoud | Pullup to -current
|
| 1.44.10.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.44.8.4 | 19-Oct-2007 |
ad | Sync with head.
|
| 1.44.8.3 | 12-Oct-2007 |
ad | Fix merge errors.
|
| 1.44.8.2 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.44.8.1 | 10-Apr-2007 |
ad | Sync with head.
|
| 1.44.4.2 | 07-May-2007 |
yamt | sync with head.
|
| 1.44.4.1 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.46.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.46.6.2 | 07-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.46.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.46.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.47.2.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.9 | 05-Oct-2007 |
dyoung | Work in progress: use a raw socket for GRE in IP encapsulation
instead of adding/subtracting our own IPv4 header.
There are many benefits: gre(4) needn't grok the outer encapsulation
header any longer, so this simplifies the gre(4) code. The IP
stack needn't grok GRE, so it is simplified, too. gre(4) will
benefit from optimizations in the socket code. Eventually, gre(4)
will gain an IPv6 encapsulation with very few new lines of code.
There is a small performance loss. A 133 MHz, 486-class AMD Elan
sinks/sources a TCP stream over GRE with about 93% the throughput
of the old code. TCP throughput on a 266 MHz, 586-class AMD Geode
is about 96% the throughput of the old code. A 175-MHz ADM5120
(MIPS) only sinks a TCP stream over GRE at about 90% of the old
code; I am still investigating that.
I produced stripped-down versions of sosend() and soreceive() for
gre(4) to use. They are guaranteed not to block, so they can be
called from a software interrupt and from a socket upcall,
respectively.
A kernel thread is no longer necessary for socket transmit/receive,
but I didn't get around to removing it, yet.
Thanks to Matt Thomas for suggesting the use of stripped-down socket
code and software interrupts, and to Andrew Doran for advice and
answers concerning software interrupts, threads, and performance.
|
| 1.8 | 10-Dec-2005 |
elad | branches: 1.8.30; 1.8.44; 1.8.46; 1.8.48;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.7 | 21-Apr-2004 |
itojun | branches: 1.7.12;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.6 | 18-Apr-2004 |
matt | De __P()
|
| 1.5 | 09-Jun-2002 |
itojun | branches: 1.5.6;
whitespace
|
| 1.4 | 06-Jul-2000 |
thorpej | branches: 1.4.2; 1.4.4; 1.4.16;
Some slight cleanup.
|
| 1.3 | 07-Oct-1998 |
thorpej | branches: 1.3.12;
Fix some typos in comments, and clean up some whitespace.
|
| 1.2 | 30-Sep-1998 |
hwr | Start supporting IPPROTO_MOBILE (55) encapsulation. This is yet
another tunneling protocol used by the Mobile-IP people. See RFC 2004
for this.
|
| 1.1 | 13-Sep-1998 |
hwr | Add a gre tunnel pseudo network device. Gre = generic route encapsulation.
This device shows up like any other network interface and can be used to
tunnel L3 protocols as e.g. IP over IP.
|
| 1.3.12.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.4.16.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.4.4.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.4.2.1 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.5.6.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.5.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.5.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.5.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.7.12.2 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.7.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.8.48.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.8.46.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.8.44.1 | 07-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.8.30.1 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.11 | 02-Oct-2004 |
christos | These are ipfilter files, although they don't have the same copyright.
Thanks jaromir.
|
| 1.10 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.9 | 28-Mar-2004 |
martti | branches: 1.9.2;
Upgraded IPFilter to 4.1.1
|
| 1.8 | 23-Jun-2003 |
martin | branches: 1.8.2;
#ifdef _KERNEL_OPT police
|
| 1.7 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.6 | 19-Sep-2002 |
martti | branches: 1.6.2;
Resync with official IPF
|
| 1.5 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.4 | 09-Jun-2002 |
itojun | branches: 1.4.2;
whitespace
|
| 1.3 | 02-May-2002 |
martti | branches: 1.3.2; 1.3.4;
Upgraded IPFilter to 3.4.27
|
| 1.2 | 01-Apr-2002 |
jdolecek | branches: 1.2.2;
add RCS IDs
|
| 1.1 | 01-Apr-2002 |
jdolecek | branches: 1.1.1;
Initial revision
|
| 1.1.1.5 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.4 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.3 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.2 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.1 | 01-Apr-2002 |
jdolecek | Import H.323 proxy of IPFilter 3.4.25. Upon closer examination,
the QNX licence seems to be allow both non-commercial and commercial
use actually.
According to Darren, the H.323 proxy code is buggy ATM, but is imported
here for reference anyway.
|
| 1.2.2.5 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.2.2.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.3 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.2.2.2 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.1 | 01-Apr-2002 |
nathanw | file ip_h323_pxy.c was added on branch nathanw_sa on 2002-04-17 00:06:25 +0000
|
| 1.3.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.3.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.4.2.3 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.4.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.4.2.1 | 09-Jun-2002 |
jdolecek | file ip_h323_pxy.c was added on branch kqueue on 2002-06-23 17:50:50 +0000
|
| 1.6.2.3 | 27-Nov-2002 |
itojun | sys/netinet/ip_h323_pxy.c via patch
sys/netinet/ip_ipsec_pxy.c via patch
sys/netinet/ip_netbios_pxy.c via patch
Fix compilation on a.out systems.
(Thorsten Frueauf)
|
| 1.6.2.2 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.6.2.1 | 19-Sep-2002 |
itojun | file ip_h323_pxy.c was added on branch netbsd-1-5 on 2002-10-18 13:16:46 +0000
|
| 1.8.2.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.8.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.9.2.1 | 13-Aug-2004 |
jmc | branches: 1.9.2.1.2;
Pullup rev 1.10 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.9.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.11 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_htable.c was added on branch ktrace-lwp on 2004-08-03 10:54:39 +0000
|
| 1.2 | 02-Oct-2004 |
jdolecek | move ip_htable.h from sys/netinet/ to sys/dist/ipf/netinet/, it's ipfilter file
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_htable.h was added on branch ktrace-lwp on 2004-08-03 10:54:39 +0000
|
| 1.180 | 05-Jun-2025 |
ozaki-r | Apply if_first_addr() and if_first_addr_psref()
|
| 1.179 | 22-Feb-2025 |
mlelstv | Use canonical M_GETHDR macro. NFCI.
|
| 1.178 | 29-Aug-2022 |
knakahara | branches: 1.178.4; 1.178.10;
Add sysctl entry to control to send routing message for RTM_DYNAMIC.
Some routing daemons require such routing message to keep coherency.
If we want to let kernel send such message, set net.inet.icmp.dynamic_rt_msg=1
for IPv4, net.inet6.icmp6.dynamic_rt_msg=1 for IPv6.
Default(=0) is the same as before, that is, not send such routing message.
|
| 1.177 | 22-Dec-2018 |
maxv | Replace M_ALIGN and MH_ALIGN by m_align.
|
| 1.176 | 15-Nov-2018 |
maxv | Remove the 't' argument from m_tag_find().
|
| 1.175 | 15-Nov-2018 |
maxv | Simplify the mtag API:
- Remove m_tag_init(), m_tag_first(), m_tag_next() and
m_tag_delete_nonpersistent().
- Remove the 't' argument from m_tag_delete_chain().
|
| 1.174 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.173 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.172 | 21-Jun-2018 |
knakahara | branches: 1.172.2;
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.171 | 01-Jun-2018 |
ozaki-r | Fix _rt_free via rtrequest(RTM_DELETE) hangs in rt_timer handlers
A rt_timer handler is passed a rtentry with an extra reference that avoids the
rtentry is accidentally released. So rt_timer handers must release the reference
of a passed rtentry by themselves (but they didn't).
|
| 1.170 | 11-May-2018 |
maxv | Retire ICMPPRINTFS, it's annoying and it doesn't build.
|
| 1.169 | 26-Apr-2018 |
maxv | Use M_UNWRITABLE, no functional change.
|
| 1.168 | 08-Feb-2018 |
maxv | branches: 1.168.2;
Fix a possible buffer overflow in the IPv4 _ctlinput functions.
In _icmp_input we are guaranteeing that the ICMP_ADVLENMIN-byte area
starting from 'icp' is contiguous.
ICMP_ADVLENMIN = 8 + sizeof(struct ip) + 8 = 36
But the _ctlinput functions (eg udp_ctlinput) expect the area to be
larger. These functions read at:
(uint8_t *)icp + 8 + (icp->icmp_ip.ip_hl << 2)
which can be crafted to be:
(uint8_t *)icp + 68
So we end up reading 'icp+68' while the valid area ended at 'icp+36'.
Having said that, it seems pretty complicated to trigger this bug; it
would have to be a fragmented packet with half of the ICMP header in the
first fragment, and we would need to have a driver that did not allocate
a cluster for the first mbuf of the chain.
The check of icmplen against ICMP_ADVLEN(icp) was not sufficient: while it
did guarantee that the ICMP header fit the chain, it did not guarantee
that it fit 'm'.
Fix this bug by pulling up to hlen+ICMP_ADVLEN(icp). No need to log an
error. Rebase the pointers afterwards.
|
| 1.167 | 05-Feb-2018 |
maxv | Declare icmperrppslim in ip_icmp.c, it shouldn't be used elsewhere.
|
| 1.166 | 23-Jan-2018 |
maxv | Don't use global variables, that's obviously incorrect on MP systems.
One remains, because it is imported in tcp_timer.c, and I'm not totally
sure of how it interacts with icmp_mtudisc().
|
| 1.165 | 23-Jan-2018 |
maxv | Style, localify icmp_send, and add a clear KASSERT (that replaces a vague
comment).
|
| 1.164 | 22-Jan-2018 |
maxv | Adapt previous, reintroduce MH_ALIGN. It's used as an optimization - we
can later prepend something to the current mbuf without having to allocate
a new mbuf.
|
| 1.163 | 19-Jan-2018 |
maxv | Fix a buffer overflow in icmp_error. We create in 'm' a packet that must
contain:
IPv4 header | Fixed part of ICMP header | Variable part of ICMP header
But we perform length checks on 'totlen', which does not count the IPv4
header.
So now, add sizeof(struct ip) in totlen, and stop doing this m_data
nonsense, just get the pointers as usual.
|
| 1.162 | 19-Jan-2018 |
maxv | Clarify icmp_error:
* Rename (and constify) oiplen -> oiphlen.
* Rename icmplen -> datalen, it's the size of the variable part of
the ICMP header, not the total size of the ICMP header itself.
* Introduce totlen, this is the total size of the ICMP header (icmp_ip
included).
No real functional change.
|
| 1.161 | 31-Mar-2017 |
ozaki-r | branches: 1.161.6;
Don't use a single global variable to store source route information for multiple incoming packets
It's not MP-safe. So use a m_tag to store the information instead.
Pointed out by knakahara@
The fix is from OpenBSD (originally fixed in FreeBSD)
|
| 1.160 | 06-Mar-2017 |
ozaki-r | Make sure icmp_redirect_timeout_q and ip_mtudisc_timeout_q are initialized on bootup
Fix PR kern/52029
|
| 1.159 | 17-Feb-2017 |
ozaki-r | Protect sysctl_net_inet_ip_pmtudto with icmp_mtx instead of softnet_lock
|
| 1.158 | 13-Feb-2017 |
ozaki-r | Protect mtudisc and redirect stuffs of icmp/icmp6 with mutex
We have to run pr_init of icmp and icmp6 prior to tcp and tcp6 ones
for mutex initialization.
|
| 1.157 | 07-Feb-2017 |
ozaki-r | Add missing NULL checks for m_get_rcvif
|
| 1.156 | 02-Feb-2017 |
ozaki-r | Defer some pr_input to workqueue
pr_input is currently called in softint. Some pr_input such as ICMP, ICMPv6
and CARP can add/delete/update IP addresses and routing table entries. For
example, icmp6_redirect_input updates an a routing table entry and
nd6_ra_input may delete an IP address.
Basically such operations shouldn't be done in softint. That aside, we have
a reason to avoid the situation; psz/psref waits cannot be used in softint,
however they are required to work in such pr_input in the MP-safe world.
The change implements the workqueue pr_input framework called wqinput which
provides a means to defer pr_input of a protocol to workqueue easily.
Currently icmp_input, icmp6_input, carp_proto_input and carp6_proto_input
are deferred to workqueue by the framework.
Proposed and discussed on tech-kern and tech-net
|
| 1.155 | 24-Jan-2017 |
ozaki-r | Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.154 | 12-Dec-2016 |
ozaki-r | branches: 1.154.2;
Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.153 | 25-Oct-2016 |
ozaki-r | Remove unnecessary argument
No functional change.
|
| 1.152 | 19-Oct-2016 |
ozaki-r | Set ia to ensure to call ia4_release
|
| 1.151 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.150 | 08-Jul-2016 |
ozaki-r | branches: 1.150.2;
Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.149 | 07-Jul-2016 |
ozaki-r | Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.148 | 06-Jul-2016 |
ozaki-r | Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.147 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.146 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.145 | 01-Apr-2016 |
ozaki-r | Remove unnecessary casts and do s/0/NULL/ for rtrequest
|
| 1.144 | 21-Jan-2016 |
riastradh | Revert previous: ran cvs commit when I meant cvs diff. Sorry!
Hit up-arrow one too few times.
|
| 1.143 | 21-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.142 | 31-Aug-2015 |
ozaki-r | Make rt_refcnt take into account rt_timer
|
| 1.141 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.140 | 09-May-2015 |
christos | if no address was found, don't check if it is tentative (hi Roy)
|
| 1.139 | 09-May-2015 |
christos | assign sin only when it is needed
|
| 1.138 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.137 | 24-Apr-2015 |
ozaki-r | Use KASSERT instead of if & panic
rt can be NULL only when programming error (and we sure it cannot for now),
so we can use KASSERT here (i.e., check only if DIAGNOSTIC).
|
| 1.136 | 24-Apr-2015 |
ozaki-r | Replace 0 with NULL for pointer variables
|
| 1.135 | 02-Dec-2014 |
christos | use the new printing code.
|
| 1.134 | 30-May-2014 |
christos | branches: 1.134.4;
Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.133 | 19-May-2014 |
rmind | - Split off PRU_ATTACH and PRU_DETACH logic into separate functions.
- Replace malloc with kmem and eliminate M_PCB while here.
- Sprinkle more asserts.
|
| 1.132 | 25-Feb-2014 |
pooka | branches: 1.132.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.131 | 05-Jun-2013 |
christos | branches: 1.131.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.130 | 25-Mar-2013 |
christos | PR/47693: Erik E. Fair: Add missing code to icmp handling.
- While there, add the rest of the missing codes
- Merge groups
- Fix indentation
|
| 1.129 | 22-Mar-2012 |
drochner | branches: 1.129.2;
remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.128 | 09-Jan-2012 |
liamjfoy | branches: 1.128.2;
check against NULL
|
| 1.127 | 31-Dec-2011 |
christos | - fix offsetof usage, and redundant defines
- kill pointer casts to 0
|
| 1.126 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.125 | 17-Jul-2011 |
joerg | branches: 1.125.2; 1.125.6;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.124 | 02-Jul-2010 |
kefren | manually adjust m_data and m_len so it can later be prepended with a
struct ip in case that a cluster is used. icmp len panic is not valid for
cluster case.
Fixes PR/43548
|
| 1.123 | 26-Jun-2010 |
kefren | Add MPLS support, proposed on tech-net@ a couple of days ago
Welcome to 5.99.33
|
| 1.122 | 07-Dec-2009 |
christos | branches: 1.122.2; 1.122.4;
PR/42243: Yasuoka Masahiko: Add "net.inet.icmp.bmcastecho" sysctl support,
to disable icmp replies to the broadcast address.
|
| 1.121 | 16-Sep-2009 |
pooka | Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.120 | 18-Jun-2008 |
yamt | branches: 1.120.6;
merge yamt-pf42 branch.
(import newer pf from OpenBSD 4.2)
ok'ed by peter@. requested by core@
|
| 1.119 | 04-May-2008 |
thorpej | branches: 1.119.2; 1.119.4;
Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.118 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.117 | 23-Apr-2008 |
thorpej | branches: 1.117.2;
Use <net/net_stats.h> / netstat_sysctl().
|
| 1.116 | 12-Apr-2008 |
thorpej | branches: 1.116.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.115 | 06-Apr-2008 |
thorpej | Change ICMP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old icmpstat structure; old netstat
binaries will continue to work properly.
|
| 1.114 | 09-Nov-2007 |
dyoung | branches: 1.114.14;
Use sockaddr_in_init(). KNF. No functional change intended.
|
| 1.113 | 27-Aug-2007 |
dyoung | branches: 1.113.2; 1.113.6; 1.113.8;
Cosmetic: 0 -> NULL. Remove unnecessary cast.
|
| 1.112 | 19-Jul-2007 |
dyoung | branches: 1.112.4; 1.112.6;
Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.111 | 04-Mar-2007 |
christos | branches: 1.111.2; 1.111.10;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.110 | 22-Feb-2007 |
thorpej | TRUE -> true, FALSE -> false
|
| 1.109 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.108 | 29-Jan-2007 |
dyoung | branches: 1.108.2;
bzero -> memset
|
| 1.107 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.106 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.105 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.104 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.103 | 30-Aug-2006 |
christos | branches: 1.103.2; 1.103.4;
fix initializers
|
| 1.102 | 28-Aug-2006 |
yamt | icmp_input: don't assume relations between PRC_ and ICMP_ values.
|
| 1.101 | 10-Jul-2006 |
peter | Wrap long lines, unwrap a short line.
|
| 1.100 | 10-Jul-2006 |
peter | Moves the PF_GENERATED m_tag to the new packet in icmp_error.
This is needed because the pf code can call icmp_error with setting
this tag, but the new packet should not be filtered when it comes back
to pf(4).
ok christos@
|
| 1.99 | 29-Mar-2006 |
dyoung | branches: 1.99.4;
When reflecting an ICMP Echo, do not scribble over read-only/shared
mbuf storage.
|
| 1.98 | 22-Mar-2006 |
matt | An MTU can't be negative so store them in unsigned variables.
|
| 1.97 | 10-Nov-2005 |
christos | branches: 1.97.6; 1.97.8; 1.97.10; 1.97.12; 1.97.14;
Remove redundant assignment (from Liam Foy)
|
| 1.96 | 23-Oct-2005 |
christos | No need to pass an interface when only the mtu is needed. From OpenBSD via
Liam Foy.
|
| 1.95 | 19-Aug-2005 |
christos | branches: 1.95.2;
make ICMPPRINTFS work; from Liam Foy.
|
| 1.94 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.93 | 19-Jul-2005 |
christos | Implement PMTU checks from:
http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.
From OpenBSD.
|
| 1.92 | 29-Apr-2005 |
yamt | branches: 1.92.2;
move decl of inetsw to its own header to avoid array of incomplete type.
found by gcc4. reported by Adam Ciarcinski.
|
| 1.91 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.90 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.89 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.88 | 24-Jan-2005 |
matt | branches: 1.88.2;
Add IFNET_FOREACH and IFADDR_FOREACH macros and start using them.
|
| 1.87 | 03-Aug-2004 |
cube | branches: 1.87.4;
Remove a common (icmpstat).
|
| 1.86 | 25-Jun-2004 |
itojun | icmp_reflect: check if m_pkthdr.rcvif is non-NULL before touching it.
icmp_reflect could be called from the output path, so m_pkthdr.rcvif may not
be set. (found by panic when PF is configured "block return all")
|
| 1.85 | 25-Jun-2004 |
itojun | be careful touching m_pkthdr.rcvif, it could be NULL if the packet was
generated from local node and icmp_error calls icmp_reflect.
|
| 1.84 | 25-May-2004 |
atatat | Sysctl descriptions under net subtree (net.key not done)
|
| 1.83 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.82 | 24-Mar-2004 |
atatat | branches: 1.82.2;
Tango on sysctl_createv() and flags. The flags have all been renamed,
and sysctl_createv() now uses more arguments.
|
| 1.81 | 04-Dec-2003 |
atatat | Dynamic sysctl.
Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(),
vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all
nodes are registered with the tree, and nodes can be added (or
removed) easily, and I/O to and from the tree is handled generically.
Since the nodes are registered with the tree, the mapping from name to
number (and back again) can now be discovered, instead of having to be
hard coded. Adding new nodes to the tree is likewise much simpler --
the new infrastructure handles almost all the work for simple types,
and just about anything else can be done with a small helper function.
All existing nodes are where they were before (numerically speaking),
so all existing consumers of sysctl information should notice no
difference.
PS - I'm sorry, but there's a distinct lack of documentation at the
moment. I'm working on sysctl(3/8/9) right now, and I promise to
watch out for buses.
|
| 1.80 | 13-Nov-2003 |
jonathan | Add m_tag_delete_nonpesrsistent(), for deleting all packet tags on
mbuf chains which are recycled (e.g., ICMP reflection, loopback
interface). A consensus was reached that such recycled packets should
behave (more-or-less) the same way if a new chain had been allocated
and the contents copied to that chain.
Some packet tags may in future be marked as "persistent" (e.g., for
mandatory access controls) and should persist across such deletion.
NetBSD as yet hos no persistent tags, so m_tag_delete_nonpersistent()
just deletes all tags. This should not be relied upon.
|
| 1.79 | 11-Nov-2003 |
jonathan | Change global head-of-local-IP-address list from in_ifaddr to
in_ifaddrhead. Recent changes in struct names caused a namespace
collision in fast-ipsec, which are most cleanly fixed by using
"in_ifaddrhead" as the listhead name.
|
| 1.78 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.77 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.76 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.75 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.74 | 26-Jun-2003 |
itojun | branches: 1.74.2;
fix comment
|
| 1.73 | 17-Apr-2003 |
tron | Clear hardware checksum flags before reusing a mbuf for an ICMP reply as
suggested by Enami Tsugutomo. This fixes PR kern/21203 by myself.
|
| 1.72 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.71 | 23-Sep-2002 |
simonb | Remove breaks after returns, unreachable returns and returns after
returns(!).
|
| 1.70 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.69 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.68 | 13-Jun-2002 |
itojun | set IPv4 parameter to modern value.
- turn on path MTU discovery (previous: turned off)
- ICMPv4 redirect entry timeout = 600 sec (previous: never timeout)
|
| 1.67 | 09-Jun-2002 |
itojun | whitespace
|
| 1.66 | 13-Nov-2001 |
lukem | branches: 1.66.8; 1.66.10;
add RCSIDs
|
| 1.65 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.64 | 04-Nov-2001 |
matt | Keep only one mtu_table (the two were identical except for
one value - 65280).
|
| 1.63 | 30-Oct-2001 |
kml | Add in support for timing out IPv4 routes added due to redirects,
as discussed in tech-net several weeks ago. It turned out that
KAME had already added this functionality to the IPv6 stack, so
I followed their example in adding the sysctl variables
net.inet.icmp.rediraccept and net.inet.icmp.redirtimeout.
|
| 1.62 | 29-Oct-2001 |
simonb | Don't need to include <uvm/uvm_extern.h> just to include <sys/sysctl.h>
anymore.
|
| 1.61 | 20-Oct-2001 |
matt | branches: 1.61.2;
Make the two MTU tables const and change their type to u_int (one was int
and one was u_long!).
|
| 1.60 | 08-Mar-2001 |
itojun | branches: 1.60.2;
Remove a bogus rtfree(); OpenBSD PR 1706.
|
| 1.59 | 01-Mar-2001 |
itojun | branches: 1.59.2;
make sure to enforce inbound ipsec policy checking, for any protocols on top
of ip (check it when final header is visited). sync with kame.
XXX kame team will need to re-check policy engine code
|
| 1.58 | 24-Jan-2001 |
itojun | - record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.57 | 18-Oct-2000 |
itojun | s/mtudisc_callback/icmp_&/ so that we don't feel conflict between IPv4 and
IPv6 counterpart. (or icmp4_&?)
|
| 1.56 | 18-Oct-2000 |
itojun | count successful path MTU changes. good for debugging.
(there could be some discussion on when to increase the counter...)
|
| 1.55 | 18-Oct-2000 |
thorpej | Restructure the Path MTU Discovery code somewhat to avoid
entering rtentry's for hosts we're not actually communicating
with.
Do this by invoking the ctlinput for the protocol, which is
responsible for validating the ICMP message:
* TCP -- Lookup the connection based on the address/port
pairs in the ICMP message.
* AH/ESP -- Lookup the SA based on the SPI in the ICMP message.
If validation succeeds, ctlinput is responsible for calling
icmp_mtudisc(). icmp_mtudisc() then invokes callbacks registered
by protocols (such as TCP) which want to take some sort of special
action when a path's MTU changes. For TCP, this is where we now
refresh cached routes and re-enter slow-start.
As a side-effect, this fixes the problem where TCP would not be
notified when a path's MTU changed if AH/ESP were being used.
XXX Note, this is only a fix for the IPv4 case. For the IPv6
XXX case, we need to wait for the KAME folks.
Reviewed by sommerfeld@netbsd.org and itojun@netbsd.org.
|
| 1.54 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.53 | 27-Jul-2000 |
itojun | do not disable icmp error rate limitation for local address.
local address can be abused too. pps rate limitation should work fine for
moderate amount of icmp errors.
|
| 1.52 | 24-Jul-2000 |
sommerfeld | Improve robustness of icmp_error():
- allow it to work when icmpreturndatabytes is sufficiently large that the
icmp error message doesn't fit in a header mbuf.
- defend against mbuf chains shorter than their contained ip->ip_len.
|
| 1.51 | 10-Jul-2000 |
itojun | implement net.inet.icmp.errppslimit.
make default value for net.inet.icmp.erratelimit to 0, as < 10ms value
does not do the right thing.
|
| 1.50 | 06-Jul-2000 |
itojun | remove unnecessary #include <netkey/key_debug.h>. from kame.
|
| 1.49 | 01-Jul-2000 |
sommerfeld | Don't rate-limit ICMP errors from packets we send to ourselves.
The dns resolver depends on reliably receiving errors to allow it to
quickly detect a dead local nameserver.
|
| 1.48 | 28-Jun-2000 |
mrg | <vm/vm.h> -> <uvm/uvm_extern.h>
|
| 1.47 | 10-Jun-2000 |
darrenr | branches: 1.47.2;
add icmpreturndatabytes kernel variable (default 8) which specifies the
number of extra data bytes to return in ICMP error messages. This is
also available via sysctl as net.icmp.returndatabytes and is limited to
[8,512].
|
| 1.46 | 22-May-2000 |
itojun | branches: 1.46.2;
disallow negative numbers for ratelimit interval (tcp, icmp, icmp6).
|
| 1.45 | 10-May-2000 |
itojun | add missing boundary checks to ip options processing.
correct timestamp option validation (len and ptr upper/lower bound
based on RFC791).
fill "pointer" field for parameter problem in timestamp option processing.
|
| 1.44 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.43 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.42 | 24-Feb-2000 |
itojun | don't transmit ICMPv4 packet back, if the original packet was encyrpted.
|
| 1.41 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.40 | 15-Feb-2000 |
thorpej | Add ICMP error rate limiting, based on the same for ICMP6.
Note, we're reusing the previously unused slot for "MTU discovery" (which
was moved to the "net.inet.ip" branch of the sysctl tree quite some time
ago).
|
| 1.39 | 25-Jan-2000 |
sommerfeld | Pick source address for ICMP errors a bit more intelligently when
there are multiple addresses on the interface.
From Marc Horowitz <marc@netbsd.org>, who left this sitting for too long.
|
| 1.38 | 09-Jul-1999 |
thorpej | branches: 1.38.2;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.37 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.36 | 30-Mar-1999 |
mycroft | branches: 1.36.4; 1.36.6;
Fix a null pointer dereference in the case where forwarding is turned on and
there are interfaces up but with no addresses.
|
| 1.35 | 19-Jan-1999 |
mycroft | There's just no plausible reason to byte-swap ip_id internally. It's opaque.
|
| 1.34 | 19-Jan-1999 |
mycroft | Don't screw with ip_len; just subtract from it where we actually use the
value.
|
| 1.33 | 19-Jan-1999 |
mycroft | Fix byte-swapping of ip_len in returned IP header.
|
| 1.32 | 11-Jan-1999 |
thorpej | Fix byte order and ip_len inconsistencies in ICMP reply code. Also, fix
some formatting and HTONS(foo) vs. foo = htons(foo) inconsistencies.
PR #6602, Darren Reed.
|
| 1.31 | 19-Dec-1998 |
thorpej | Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.30 | 30-Sep-1998 |
tls | branches: 1.30.4;
Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.29 | 29-Apr-1998 |
kml | Add support for deletion of routes added by path MTU discovery;
uses new generic route timeout code. Add sysctl for timeout period.
|
| 1.28 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.27 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.26 | 29-Oct-1997 |
kml | Changes to path MTU discovery to correctly handle "needs
fragmentation" ICMP messages that specify a new MTU size of zero
(from, say, old buggy Linux kernels).
|
| 1.25 | 18-Oct-1997 |
kml | branches: 1.25.2;
change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.24 | 17-Oct-1997 |
kml | Path MTU Discovery support. This is turned off by default.
Use sysctl -w net.inet.icmp.mtudisc=1 to turn on.
Still to come: path removal after some period, black hole detection
|
| 1.23 | 24-Jun-1997 |
thorpej | Increment icmpstat.icps_badlen for bad length of ICMP_MASKREQ, per
Stevens in TCP/IP Illustrated vol. 2, p.319. Submitted by
Koji Imada <koji@math.human.nagoya-u.ac.jp> in PR #3712.
|
| 1.22 | 13-Oct-1996 |
christos | backout previous kprintf changes
|
| 1.21 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.20 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.19 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.18 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.17 | 04-Jun-1995 |
mycroft | Don't cast things unnecessarily.
|
| 1.16 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.15 | 01-Jun-1995 |
mycroft | Don't use INADDR_* constants in case labels.
|
| 1.14 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.13 | 31-May-1995 |
mycroft | Integrate multicast 3.5 distribution, with several bugs fixed and general
cleanup. This is a (working) snapshot of work in progress.
|
| 1.12 | 15-May-1995 |
cgd | KNF
|
| 1.11 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.7 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.6 | 08-Jan-1994 |
mycroft | More prototypes.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.3 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.2 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.25.2.3 | 01-Oct-1998 |
cgd | pull up revisions 1.27-1.28, 1.30 from trunk. (tls)
|
| 1.25.2.2 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.25.2.1 | 30-Oct-1997 |
mellon | Pull rev 1.26 up from trunk
|
| 1.30.4.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.36.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.36.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.36.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.36.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.36.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.38.2.3 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.38.2.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.38.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.46.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.47.2.5 | 06-Apr-2001 |
he | Pull up revision 1.58 (requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.47.2.4 | 11-Mar-2001 |
he | Pull up revision 1.59 (requested by itojun):
Ensure that we enforce inbound IPsec policy on all IP protocols,
not just TCP, UDP and ICMP.
|
| 1.47.2.3 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.47.2.2 | 28-Jul-2000 |
sommerfeld | Pull up UDP, ICMP fixes:
- Drop packet, increment udps_badlen if the udp header length field
reports a size smaller than the udp header; defends against bogus
packets seen by by Assar Westerlund.
- allow icmp_error() to work when icmpreturndatabytes is sufficiently
large that the icmp error message doesn't fit in a header mbuf.
- defend against mbuf chains shorter than their contained ip->ip_len.
Joint work of myself, itojun, and assar
Approved by thorpej
revisions pulled up:
sys/netinet/ip_icmp.c 1.52
sys/netinet/udp_usrreq.c 1.70
|
| 1.47.2.1 | 02-Jul-2000 |
sommerfeld | Pull up 1.49: don't rate-limit ICMP we send to ourselves.
|
| 1.59.2.7 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.59.2.6 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.59.2.5 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.59.2.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.59.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.59.2.2 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.59.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.60.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.60.2.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.60.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.60.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.61.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.66.10.2 | 16-Jun-2003 |
grant | Pull up revision 1.73 (requested by tron in ticket #1260):
Clear hardware checksum flags before reusing a mbuf for an ICMP reply as
suggested by Enami Tsugutomo. This fixes PR kern/21203 by myself.
|
| 1.66.10.1 | 15-Jun-2002 |
lukem | Pull up revision 1.68 (requested by itojun in ticket #266):
set IPv4 parameter to modern value.
- ICMPv4 redirect entry timeout = 600 sec (previous: never timeout)
|
| 1.66.8.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.66.8.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.66.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.74.2.8 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.74.2.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.74.2.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.74.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.74.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.74.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.74.2.2 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.74.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.82.2.2 | 03-Aug-2004 |
jmc | Pullup rev 1.85-1.87 (requested by christos in ticket #732)
icmp_reflect: check if m_pkthdr.rcvif is non-NULL before touching it.
icmp_reflect could be called from the output path, so m_pkthdr.rcvif may not
be set. (found by panic when PF is configured "block return all")
|
| 1.82.2.1 | 28-May-2004 |
tron | Pull up revision 1.84 (requested by atatat in ticket #391):
Sysctl descriptions under net subtree (net.key not done)
|
| 1.87.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.88.2.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.88.2.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.92.2.5 | 15-Nov-2007 |
yamt | sync with head.
|
| 1.92.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.92.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.92.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.92.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.95.2.1 | 26-Oct-2005 |
yamt | sync with head
|
| 1.97.14.2 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.97.14.1 | 28-Mar-2006 |
tron | Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
|
| 1.97.12.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.97.10.3 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.97.10.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.97.10.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.97.8.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.97.6.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.99.4.1 | 13-Jul-2006 |
gdamore | Merge from HEAD.
|
| 1.103.4.3 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.103.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.103.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.103.2.3 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.103.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.103.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.108.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.108.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.111.10.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.111.10.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.111.2.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.111.2.1 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.112.6.2 | 19-Jul-2007 |
dyoung | Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.112.6.1 | 19-Jul-2007 |
dyoung | file ip_icmp.c was added on branch matt-mips64 on 2007-07-19 20:48:55 +0000
|
| 1.112.4.2 | 11-Nov-2007 |
joerg | Sync with HEAD.
|
| 1.112.4.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.113.8.1 | 19-Nov-2007 |
mjf | Sync with HEAD.
|
| 1.113.6.1 | 13-Nov-2007 |
bouyer | Sync with HEAD
|
| 1.113.2.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.114.14.2 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.114.14.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.116.2.2 | 18-May-2008 |
yamt | sync with head.
|
| 1.116.2.1 | 19-Apr-2008 |
yamt | Peter Postma's work-in-progress pf import from OpenBSD 4.2.
updated to -current by me.
|
| 1.117.2.4 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.117.2.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.117.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.117.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.119.4.1 | 18-Jun-2008 |
simonb | Sync with head.
|
| 1.119.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.120.6.1 | 09-Jun-2013 |
msaitoh | Pull up following revision(s) (requested by fair in ticket #1855):
sys/netinet/ip_icmp.c: revision 1.130
PR/47693: Erik E. Fair: Add missing code to icmp handling.
- While there, add the rest of the missing codes
- Merge groups
- Fix indentation
|
| 1.122.4.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.122.2.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.125.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.125.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.125.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.125.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.128.2.1 | 31-Mar-2013 |
riz | Pull up following revision(s) (requested by fair in ticket #860):
sys/netinet/ip_icmp.c: revision 1.130
PR/47693: Erik E. Fair: Add missing code to icmp handling.
- While there, add the rest of the missing codes
- Merge groups
- Fix indentation
|
| 1.129.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.129.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.129.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.131.2.2 | 18-May-2014 |
rmind | sync with head
|
| 1.131.2.1 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.132.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.134.4.9 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.134.4.8 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.134.4.7 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.134.4.6 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.134.4.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.134.4.4 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.134.4.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.134.4.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.134.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.150.2.5 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.150.2.4 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.150.2.3 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.150.2.2 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.150.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.154.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.161.6.3 | 13-Jul-2018 |
martin | Pull up following revision(s) via patch (requested by knakahara in ticket #905):
sys/netinet/ip_mroute.c: revision 1.160
sys/netinet6/in6_l2tp.c: revision 1.16
sys/net/if.h: revision 1.263
sys/netinet/in_l2tp.c: revision 1.15
sys/netinet/ip_icmp.c: revision 1.172
sys/netinet/igmp.c: revision 1.68
sys/netinet/ip_encap.c: revision 1.69
sys/netinet6/ip6_mroute.c: revision 1.129
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.161.6.2 | 08-Jun-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #852):
sys/netinet6/icmp6.c: revision 1.238
sys/netinet/ip_icmp.c: revision 1.171
sys/net/route.c: revision 1.210
Fix _rt_free via rtrequest(RTM_DELETE) hangs in rt_timer handlers
A rt_timer handler is passed a rtentry with an extra reference that avoids the
rtentry is accidentally released. So rt_timer handers must release
the reference of a passed rtentry by themselves (but they didn't).
|
| 1.161.6.1 | 31-Mar-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #675):
sys/netinet/ip_icmp.c: revision 1.168
Fix a possible buffer overflow in the IPv4 _ctlinput functions.
In _icmp_input we are guaranteeing that the ICMP_ADVLENMIN-byte area
starting from 'icp' is contiguous.
ICMP_ADVLENMIN = 8 + sizeof(struct ip) + 8 = 36
But the _ctlinput functions (eg udp_ctlinput) expect the area to be
larger. These functions read at:
(uint8_t *)icp + 8 + (icp->icmp_ip.ip_hl << 2)
which can be crafted to be:
(uint8_t *)icp + 68
So we end up reading 'icp+68' while the valid area ended at 'icp+36'.
Having said that, it seems pretty complicated to trigger this bug; it
would have to be a fragmented packet with half of the ICMP header in the
first fragment, and we would need to have a driver that did not allocate
a cluster for the first mbuf of the chain.
The check of icmplen against ICMP_ADVLEN(icp) was not sufficient: while it
did guarantee that the ICMP header fit the chain, it did not guarantee
that it fit 'm'.
Fix this bug by pulling up to hlen+ICMP_ADVLEN(icp). No need to log an
error. Rebase the pointers afterwards.
|
| 1.168.2.7 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.168.2.6 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.168.2.5 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.168.2.4 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.168.2.3 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.168.2.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.168.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.172.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.178.10.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.178.4.1 | 01-Oct-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1164):
sys/net/link_proto.c: revision 1.41
sys/netinet6/in6.c: revision 1.293
sys/net/if.h: revision 1.307
sys/netinet/ip_icmp.c: revision 1.180
sys/dev/vmt/vmt_subr.c: revision 1.11
sys/netinet6/in6_var.h: revision 1.105
sys/netinet6/in6_var.h: revision 1.106
sys/net/if.c: revision 1.532
sys/net/if.c: revision 1.533
sys/netinet6/mld6.c: revision 1.102
sys/netinet/in_var.h: revision 1.104
sys/net/if_spppsubr.c: revision 1.270
sys/net/if_spppsubr.c: revision 1.271
sys/netinet6/nd6.c: revision 1.284
if: introduce if_first_addr() (and psref variant)
It returns a first address (ifa) of a given family on a given interface.
It will replace a bunch of open codes and make their intent clear.
Apply if_first_addr() and if_first_addr_psref()
in6: introduce in6ifa_first_lladdr() (and psref variant)
It returns a first link-local address (ifa) on a given interface.
Apply in6ifa_first_lladdr() and in6ifa_first_lladdr_psref()
|
| 1.44 | 24-May-2022 |
andvar | fix various typos in comment, documentation and log messages.
|
| 1.43 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.42 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.41 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.40 | 14-Sep-2018 |
maxv | branches: 1.40.12;
Use non-variadic function pointer in protosw::pr_input.
|
| 1.39 | 08-Feb-2018 |
maxv | branches: 1.39.2; 1.39.4;
Use C99 types - in particular, stop using n_time and n_short -, style, and
remove prototype of icmp_sysctl (does not exist). No functional change.
|
| 1.38 | 23-Jan-2018 |
maxv | Style, localify icmp_send, and add a clear KASSERT (that replaces a vague
comment).
|
| 1.37 | 19-Jan-2018 |
maxv | Move the ICMP Extension structures from mpls_ttl.c to ip_icmp.h; that's
part of the ICMP protocol (per RFC4884), and not specific to MPLS. Also
add ih_exthdr in struct icmp, the 'length' field appeared.
While here, style in MPLS.
|
| 1.36 | 19-Jan-2018 |
maxv | Style, explain a bit, and fix icmp_radv, it should be icmp_dun.id_radv.
|
| 1.35 | 17-Feb-2017 |
ozaki-r | Protect sysctl_net_inet_ip_pmtudto with icmp_mtx instead of softnet_lock
|
| 1.34 | 18-Feb-2015 |
christos | branches: 1.34.2; 1.34.4;
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
XXX: pullup-7
|
| 1.33 | 24-Dec-2011 |
christos | branches: 1.33.2; 1.33.6; 1.33.8; 1.33.16; 1.33.22; 1.33.24;
missing comma
|
| 1.32 | 24-Dec-2011 |
christos | add ICMP_STRINGS, a few more missing constants.
|
| 1.31 | 24-Dec-2011 |
christos | fix whitespace only
|
| 1.30 | 24-Dec-2011 |
christos | add SKIP, fix PHOTURIS codes
|
| 1.29 | 23-Dec-2011 |
jmc | Include the ICMP_PHOTURIS error codes if we're going to define ICMP_PHOTURIS
|
| 1.28 | 23-Dec-2011 |
christos | make ICMP_MAXTYPE 18 again to unbreak stats.
|
| 1.27 | 23-Dec-2011 |
christos | add missing icmp types.
|
| 1.26 | 26-Jun-2010 |
kefren | branches: 1.26.8; 1.26.12;
Add MPLS support, proposed on tech-net@ a couple of days ago
Welcome to 5.99.33
|
| 1.25 | 08-Sep-2008 |
gmcgarry | branches: 1.25.4; 1.25.14; 1.25.16;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.
|
| 1.24 | 25-Dec-2007 |
perry | branches: 1.24.6; 1.24.10; 1.24.12; 1.24.16;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.23 | 14-May-2006 |
christos | branches: 1.23.34; 1.23.40; 1.23.44; 1.23.48;
Comment out attribute packed. Gcc4 warns us that the field is too narrow
for packing. Produces the same size struct on i386 (28 bytes)
|
| 1.22 | 10-Dec-2005 |
elad | branches: 1.22.4; 1.22.6; 1.22.8; 1.22.12;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.21 | 23-Oct-2005 |
christos | No need to pass an interface when only the mtu is needed. From OpenBSD via
Liam Foy.
|
| 1.20 | 21-Apr-2004 |
itojun | branches: 1.20.12; 1.20.14;
no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.19 | 18-Apr-2004 |
matt | De __P()
|
| 1.18 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.17 | 09-Jun-2002 |
itojun | branches: 1.17.6;
whitespace
|
| 1.16 | 30-Oct-2001 |
kml | branches: 1.16.8;
Add in support for timing out IPv4 routes added due to redirects,
as discussed in tech-net several weeks ago. It turned out that
KAME had already added this functionality to the IPv6 stack, so
I followed their example in adding the sysctl variables
net.inet.icmp.rediraccept and net.inet.icmp.redirtimeout.
|
| 1.15 | 18-Oct-2000 |
thorpej | branches: 1.15.2; 1.15.4; 1.15.8;
Restructure the Path MTU Discovery code somewhat to avoid
entering rtentry's for hosts we're not actually communicating
with.
Do this by invoking the ctlinput for the protocol, which is
responsible for validating the ICMP message:
* TCP -- Lookup the connection based on the address/port
pairs in the ICMP message.
* AH/ESP -- Lookup the SA based on the SPI in the ICMP message.
If validation succeeds, ctlinput is responsible for calling
icmp_mtudisc(). icmp_mtudisc() then invokes callbacks registered
by protocols (such as TCP) which want to take some sort of special
action when a path's MTU changes. For TCP, this is where we now
refresh cached routes and re-enter slow-start.
As a side-effect, this fixes the problem where TCP would not be
notified when a path's MTU changed if AH/ESP were being used.
XXX Note, this is only a fix for the IPv4 case. For the IPv6
XXX case, we need to wait for the KAME folks.
Reviewed by sommerfeld@netbsd.org and itojun@netbsd.org.
|
| 1.14 | 20-Nov-1999 |
thorpej | Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.13 | 10-Feb-1998 |
perry | branches: 1.13.12; 1.13.14; 1.13.20;
add/cleanup multiple inclusion protection.
|
| 1.12 | 26-Aug-1997 |
thorpej | Add ICMP unreachable code #13 - "Communication Administratively Prohibited",
per RFC 1716. From Havard Eidnes <he@vader.runit.sintef.no>, PR #4038.
|
| 1.11 | 03-Aug-1996 |
neil | branches: 1.11.10;
Prototypes and definitions for ICMP Router Discovery, From FreeBSD.
rdisc coming soon! :-)
|
| 1.10 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.9 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.8 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.7 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.6 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.5 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.4 | 08-Jan-1994 |
mycroft | More prototypes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.11.10.1 | 28-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.13.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.13.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.13.12.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.13.12.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.13.12.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.15.8.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.15.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.15.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.15.2.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.15.2.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.16.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.17.6.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.17.6.4 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.17.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.17.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.17.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.20.14.1 | 26-Oct-2005 |
yamt | sync with head
|
| 1.20.12.2 | 21-Jan-2008 |
yamt | sync with head
|
| 1.20.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.22.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.22.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.22.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.22.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.23.48.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.23.44.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.23.40.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.23.34.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.24.16.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.24.12.1 | 24-Sep-2008 |
wrstuden | Merge in changes between wrstuden-revivesa-base-2 and
wrstuden-revivesa-base-3.
|
| 1.24.10.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.24.10.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.24.6.1 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.25.16.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.25.14.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.25.4.1 | 09-Jun-2013 |
msaitoh | Apply patch (in ticket #1855):
Add some macros. This is a part of ip_icmp.h rev. 1.32.
|
| 1.26.12.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.26.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.33.24.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.33.24.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.33.22.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #537):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
XXX: pullup-7
|
| 1.33.16.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #1258):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
|
| 1.33.8.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #1258):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
|
| 1.33.6.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.33.2.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by christos in ticket #1258):
sys/netinet/icmp_var.h: revision 1.30
sys/netinet/ip_icmp.h: revision 1.34
PR/49676: Ryo Shimizu: ICMP_STATINC() buffer overflows
|
| 1.34.4.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.34.2.1 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.39.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.39.2.1 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.40.12.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.17 | 08-Mar-2021 |
christos | remove now unused pseudo-random ip id code.
|
| 1.16 | 18-Oct-2019 |
msaitoh | branches: 1.16.8;
s/initalize/initialize/ in comment or printf message.
|
| 1.15 | 19-Nov-2011 |
tls | branches: 1.15.50;
First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>. This change includes
the following:
An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are
fixed. Some effort is made to accumulate entropy more quickly at
boot time.
A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.
The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.
An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c). This
is based on Greg Rose's implementation from Qualcomm.
A new random stream generator, nist_ctr_drbg, is provided. It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim. This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.
An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness. The arc4random/arc4randbytes API is deprecated for
in-kernel use. It is replaced by "cprng_strong". The current
cprng_fast implementation wraps the existing arc4random
implementation. The current cprng_strong implementation wraps the
new CTR_DRBG implementation. Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.
In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data(). Performance is expected to improve.
The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.
The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot. There is
approximately a 3/10000 chance of a false positive from these
tests. Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.
A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32). A new kernel will require a new rndctl.
The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.
Manual pages for the new kernel interfaces are forthcoming.
|
| 1.14 | 05-Nov-2010 |
rmind | branches: 1.14.8;
ip_randomid: make mechanism MP-safe and more modular.
OK matt@
|
| 1.13 | 04-Nov-2010 |
matt | Replace the copyright with a new TNF copyright since nothing of the old
ip_id.c remains. Remove old comments which have no relevance anymore.
|
| 1.12 | 06-Feb-2008 |
matt | branches: 1.12.30; 1.12.32;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.11 | 30-Aug-2006 |
christos | branches: 1.11.28; 1.11.34;
static comes first
|
| 1.10 | 07-Jun-2006 |
kardel | merge FreeBSD timecounters from branch simonb-timecounters
- struct timeval time is gone
time.tv_sec -> time_second
- struct timeval mono_time is gone
mono_time.tv_sec -> time_uptime
- access to time via
{get,}{micro,nano,bin}time()
get* versions are fast but less precise
- support NTP nanokernel implementation (NTP API 4)
- further reading:
Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf
NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
|
| 1.9 | 11-Dec-2005 |
christos | branches: 1.9.4; 1.9.6; 1.9.8; 1.9.14;
merge ktrace-lwp.
|
| 1.8 | 23-Mar-2004 |
itojun | branches: 1.8.4; 1.8.18;
typo
|
| 1.7 | 02-Jan-2004 |
itojun | no need for tmp = arc4randomid here
|
| 1.6 | 26-Dec-2003 |
wiz | Niels Provos kindly agreed to drop clauses 3 and 4 from the
license -- thanks.
Based on OpenBSD commit and hints by itojun.
|
| 1.5 | 10-Dec-2003 |
itojun | comment from niels provos;
- seed2 is necessary, but use it as "seed2 + x" not "seed2 ^ x".
- skipping number is not needed, so disable it for 16bit generator (makes
the repetition period to 30000)
|
| 1.4 | 25-Nov-2003 |
itojun | "seed2" was ruining non-repeating property, so remove it. discussed on tech-net
|
| 1.3 | 19-Nov-2003 |
jonathan | Patch back support for (badly) randomized IP ids, by request:
* Include "opt_inet.h" everywhere IP-ids are generated with ip_newid(),
so the RANDOM_IP_ID option is visible. Also in ip_id(), to ensure
the prototype for ip_randomid() is made visible.
* Add new sysctl to enable randomized IP-ids, provided the kernel was
configured with RANDOM_IP_ID. (The sysctl defaults to zero, and is
a read-only zero if RANDOM_IP_ID is not configured).
Note that the implementation of randomized IP ids is still defective,
and should not be enabled at all (even if configured) without
very careful deliberation. Caveat emptor.
|
| 1.2 | 16-Sep-2003 |
itojun | exp is reserved name under posix
|
| 1.1 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.8.18.3 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.8.18.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.8.18.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.8.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.8.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.8.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.8.4.1 | 23-Mar-2004 |
skrll | file ip_id.c was added on branch ktrace-lwp on 2004-08-03 10:54:39 +0000
|
| 1.9.14.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.9.8.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.9.8.1 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.9.6.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time(), use bintime's for timeout
calculations and use "time_second" instead of "time.tv_sec".
|
| 1.9.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.11.34.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.11.28.1 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.12.32.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.12.30.1 | 06-Nov-2010 |
uebayasi | Sync with HEAD.
|
| 1.14.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.15.50.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.16.8.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.406 | 17-Jul-2025 |
ozaki-r | in: avoid racy ia4_acquire(ifatoia(rt->rt_ifa) in ip_rtaddr()
Same as the case of ip_output(), it's racy and should be avoided.
PR kern/59527
|
| 1.405 | 17-Jun-2025 |
ozaki-r | in: avoid packet looping on incoming packets destining to an initializing address
The initialization of an IPv4 address is done by adding a connected route and
a local route (if necessary), and then publishing itself by adding it to the
global list (and the global hashtable). Thus, there can exist a route with an
address that is not published. This inconsistent state allows an incoming
packet destining to one of a host address which is not published but has a
local route to be forwarded and routed to a loopback interface. This results
in forwarding the packet back to ip_input, that is, packet looping.
To avoid the situation, prohibit packets being forwarded via a local route.
This is a workaround for "IPv4 address initialization atomicity" in doc/TODO.smpnet.
|
| 1.404 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.403 | 29-Jun-2024 |
riastradh | branches: 1.403.2;
netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.402 | 02-Sep-2022 |
thorpej | branches: 1.402.4;
pktqueue: Re-factor sysctl handling.
Provide a new pktq_sysctl_setup() function that attaches standard
pktq sysctl nodes below a specified parent node, with either a
fixed node ID or CTL_CREATE to dynamically assign node IDs. Make
all of the sysctl handlers private to pktqueue.c, and remove the
INET- and INET6-specific pktqueue sysctl code from net/if.c.
|
| 1.401 | 08-Mar-2021 |
christos | remove now unused pseudo-random ip id code.
|
| 1.400 | 07-Mar-2021 |
christos | netinet: Enable random IP fragment ids by default (from riastradh)
|
| 1.399 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.398 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.397 | 28-Aug-2020 |
ozaki-r | branches: 1.397.2;
inet: reduce silent packet discards
|
| 1.396 | 28-Aug-2020 |
ozaki-r | inet: pull m_get_rcvif_psref out of ip_input for simplicity
Same as ip6_input.
|
| 1.395 | 28-Aug-2020 |
ozaki-r | ipsec: rename ipsec_ip_input to ipsec_ip_input_checkpolicy
Because it just checks if a packet passes security policies.
|
| 1.394 | 28-Aug-2020 |
ozaki-r | inet, inet6: count packets dropped by IPsec
The counters count packets dropped due to security policy checks.
|
| 1.393 | 13-Nov-2019 |
ozaki-r | Get rid of unnecessary NULL checks for rt_ifa and ifa_ifp
They are always non-NULL nowadays.
|
| 1.392 | 19-Sep-2019 |
ozaki-r | Apply some missing changes lost on the previous commit
|
| 1.391 | 19-Sep-2019 |
ozaki-r | Avoid having a rtcache directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
|
| 1.390 | 15-Sep-2019 |
bouyer | Packet filters can return an mbuf chain with fragmented headers, so
m_pullup() it if needed and remove the KASSERT()s.
|
| 1.389 | 13-May-2019 |
ozaki-r | branches: 1.389.2;
Count packets dropped by pfil
|
| 1.388 | 17-Jan-2019 |
knakahara | Fix ipsecif(4) cannot apply input direction packet filter. Reviewed by ozaki-r@n.o and ryo@n.o.
Add ATF later.
|
| 1.387 | 15-Nov-2018 |
maxv | Remove the 't' argument from m_tag_find().
|
| 1.386 | 02-Sep-2018 |
maxv | remove reference to ipnat, and duplicate comments
|
| 1.385 | 10-Jul-2018 |
maxv | Remove the second argument from ip_reass_packet(). We want the IP header
on the mbuf, not elsewhere. Simplifies the NPF reassembly code a little.
No real functional change.
|
| 1.384 | 17-May-2018 |
maxv | branches: 1.384.2;
Add KASSERTs, related to PR/39794.
|
| 1.383 | 14-May-2018 |
maxv | Merge ipsec4_input and ipsec6_input into ipsec_ip_input. Make the argument
a bool for clarity. Optimize the function: if M_CANFASTFWD is not there
(because already removed by the firewall) leave now.
Makes it easier to see that M_CANFASTFWD is not removed on IPv6.
|
| 1.382 | 10-May-2018 |
maxv | Rename ipsec4_forward -> ipsec_mtu, and switch to void.
|
| 1.381 | 26-Apr-2018 |
maxv | Remove unused mbuf argument from sbsavetimestamp.
|
| 1.380 | 15-Apr-2018 |
maxv | Introduce a m_verify_packet function, that verifies the mbuf chain of a
packet to ensure it is not malformed. Call this function in "points of
interest", that are the IPv4/IPv6/IPsec entry points. There could be more.
We use M_VERIFY_PACKET(m), declared under DIAGNOSTIC only.
This function should not be called everywhere, especially not in places
that temporarily manipulate (and clobber) the mbuf structure; once they're
done they put the mbuf back in a correct format.
|
| 1.379 | 11-Apr-2018 |
maxv | Don't pass IP_ALLOWBROADCAST in ipsec4_input. The flag lands in
ipsec_getpolicybyaddr, and only IP_FORWARDING is taken.
In fact it would be good to change the 'flags' argument of ipsec4_input
to be a boolean, same for ipsec_getpolicybyaddr. It would be less
misleading.
|
| 1.378 | 11-Apr-2018 |
maxv | Add comment about IPsec.
|
| 1.377 | 11-Apr-2018 |
maxv | Small changes in ip_dooptions: replace bcopy by memcpy, the areas can't
overlap.
|
| 1.376 | 24-Feb-2018 |
ozaki-r | branches: 1.376.2;
Avoid a deadlock between softnet_lock and IFNET_LOCK
A deadlock occurs because there is a violation of the rule of lock ordering;
softnet_lock is held with hodling IFNET_LOCK, which violates the rule.
To avoid the deadlock, replace softnet_lock in in_control and in6_control
with KERNEL_LOCK.
We also need to add some KERNEL_LOCKs to protect the network stack surely.
This is required, for example, for PR kern/51356.
Fix PR kern/53043
|
| 1.375 | 09-Feb-2018 |
maxv | Remove dead code.
|
| 1.374 | 07-Feb-2018 |
maxv | Remove null check on ip, it can't be null. (Confuses code scanners.)
|
| 1.373 | 06-Feb-2018 |
maxv | Typos and style a bit, no functional change.
|
| 1.372 | 05-Feb-2018 |
maxv | Exterminate IPSENDREDIRECTS and IPMTUDISCTIMEOUT, neither is documented.
|
| 1.371 | 05-Feb-2018 |
maxv | Nuke DIRECTED_BROADCAST, it is not documented and not enabled anywhere. It
probably wouldn't have built correctly anyway, since there is no associated
defflag.
These ten lines of code in ip_input.c already look a lot better.
|
| 1.370 | 05-Feb-2018 |
maxv | Clean up this mess. This is typically the kind of places where we need to
seriously cut the bullshit. These things are unreadable, undocumented, and
all they bought us was not figuring out we had IPv4 forwarding enabled by
default for 20+ years.
|
| 1.369 | 05-Feb-2018 |
maxv | Be tougher, and don't allow LSRR+SSRR (RFC7126).
|
| 1.368 | 05-Feb-2018 |
maxv | Kick duplicate options, they are not allowed (RFC791).
|
| 1.367 | 05-Feb-2018 |
maxv | Remove unused variable.
|
| 1.366 | 05-Feb-2018 |
maxv | Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.365 | 05-Feb-2018 |
maxv | Style, no functional change.
|
| 1.364 | 01-Jan-2018 |
christos | 1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
|
| 1.363 | 24-Nov-2017 |
roy | Allow local communication over DETACHED addresses.
Allow binding to DETACHED or TENTATIVE addresses as we deny
sending upstream from them anyway.
Prefer non DETACHED or TENTATIVE addresses.
|
| 1.362 | 17-Nov-2017 |
ozaki-r | Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
|
| 1.361 | 27-Sep-2017 |
ozaki-r | Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
|
| 1.360 | 27-Jul-2017 |
ozaki-r | Don't acquire global locks for IPsec if NET_MPSAFE
Note that the change is just to make testing easy and IPsec isn't MP-safe yet.
|
| 1.359 | 19-Jul-2017 |
ozaki-r | Correct a comment
|
| 1.358 | 08-Jul-2017 |
christos | Reorder the controls to the ones that need an interface and the ones that
don't; process the ones that don't first. Add a DIAGNOSTIC if there is no
interface; really this should be a KASSERT/panic because it is a bug if the
interface is not set at this point.
|
| 1.357 | 06-Jul-2017 |
christos | remove unnecessary casts (no functional change)
|
| 1.356 | 06-Jul-2017 |
christos | Merge the two copies SO_TIMESTAMP/SO_OTIMESTAMP processing to a single
function, and add a SOOPT_TIMESTAMP define reducing compat pollution from
5 places to 1.
|
| 1.355 | 01-Jun-2017 |
chs | branches: 1.355.2;
remove checks for failure after memory allocation calls that cannot fail:
kmem_alloc() with KM_SLEEP
kmem_zalloc() with KM_SLEEP
percpu_alloc()
pserialize_create()
psref_class_create()
all of these paths include an assertion that the allocation has not failed,
so callers should not assert that again.
|
| 1.354 | 31-Mar-2017 |
ozaki-r | Don't use a single global variable to store source route information for multiple incoming packets
It's not MP-safe. So use a m_tag to store the information instead.
Pointed out by knakahara@
The fix is from OpenBSD (originally fixed in FreeBSD)
|
| 1.353 | 31-Mar-2017 |
ozaki-r | Don't use a single global variable as a temporal storage for multiple packets
It's not MP-safe. So use local variables instead.
|
| 1.352 | 06-Mar-2017 |
ozaki-r | Make sure icmp_redirect_timeout_q and ip_mtudisc_timeout_q are initialized on bootup
Fix PR kern/52029
|
| 1.351 | 17-Feb-2017 |
ozaki-r | Fix return value
|
| 1.350 | 17-Feb-2017 |
ozaki-r | Protect sysctl_net_inet_ip_pmtudto with icmp_mtx instead of softnet_lock
|
| 1.349 | 07-Feb-2017 |
ozaki-r | Add missing NULL checks for m_get_rcvif
|
| 1.348 | 24-Jan-2017 |
ozaki-r | Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.347 | 12-Dec-2016 |
ozaki-r | branches: 1.347.2;
Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.346 | 08-Dec-2016 |
ozaki-r | Use psref for ip_rtaddr
ip_rtaddr will be sleepable soon. So use psref instead of pserialize.
|
| 1.345 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.344 | 18-Oct-2016 |
ozaki-r | Don't hold global locks if NET_MPSAFE is enabled
If NET_MPSAFE is enabled, don't hold KERNEL_LOCK and softnet_lock in
part of the network stack such as IP forwarding paths. The aim of the
change is to make it easy to test the network stack without the locks
and reduce our local diffs.
By default (i.e., if NET_MPSAFE isn't enabled), the locks are held
as they used to be.
Reviewed by knakahara@
|
| 1.343 | 18-Oct-2016 |
ozaki-r | Avoid double frees of mbuf
May fix one of panicks reported by Tom Ivar Helbekkmo in PR kern/51522
|
| 1.342 | 11-Oct-2016 |
ozaki-r | Fix kernel builds with IFA_STATS
|
| 1.341 | 07-Sep-2016 |
roy | Disallow input to detached addresses because they are not yet valid.
|
| 1.340 | 31-Aug-2016 |
ozaki-r | Make ipforward_rt and ip6_forward_rt percpu
Sharing one rtcache between CPUs is just a bad idea.
Reviewed by knakahara@
|
| 1.339 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.338 | 26-Jul-2016 |
ozaki-r | Fix downmatch increment
|
| 1.337 | 08-Jul-2016 |
ozaki-r | branches: 1.337.2;
CID 1363344: remove dead code
We may need to reconsider a case when m_get_rcvif_psref returns NULL.
|
| 1.336 | 07-Jul-2016 |
ozaki-r | Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.335 | 06-Jul-2016 |
ozaki-r | Switch the IPv4 address list to pslist(9)
Note that we leave the old list just in case; it seems there are some
kvm(3) users accessing the list. We can remove it later if we confirmed
nobody does actually.
|
| 1.334 | 06-Jul-2016 |
ozaki-r | Add and use pslist(9)-based hashtable for IPv4 addresses
Note that we leave the old hashtable to keep vmstat -H working.
|
| 1.333 | 04-Jul-2016 |
ozaki-r | Separate IP address matching functions
No functional change intended.
|
| 1.332 | 30-Jun-2016 |
ozaki-r | Tidy up goto lables
No functional change.
|
| 1.331 | 30-Jun-2016 |
ozaki-r | Fix error paths
Some error paths did m_put_rcvif_psref twice.
|
| 1.330 | 28-Jun-2016 |
ozaki-r | Add missing NULL checks for m_get_rcvif_psref
|
| 1.329 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.328 | 21-Jan-2016 |
riastradh | Revert previous: ran cvs commit when I meant cvs diff. Sorry!
Hit up-arrow one too few times.
|
| 1.327 | 21-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.326 | 08-Jan-2016 |
knakahara | eliminate ip_input.c and ip6_input.c dependency on gif(4)
|
| 1.325 | 13-Oct-2015 |
roy | Include arp.h to restore the sysctl net.inet.ip.dad_count.
Fixes PR kern/49883 thanks to HITOSHI Osada.
|
| 1.324 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.323 | 07-Aug-2015 |
ozaki-r | Use time_uptime instead of time_second to avoid time leaps
Some codes in sys/net* use time_second to manage time periods such as
cache expirations. However, time_second doesn't increase monotonically
and can leap by say settimeofday(2) according to time_second(9). We
should use time_uptime instead of it to avoid such time leaps.
This change replaces time_second with time_uptime. Additionally it
converts a time based on time_uptime to a time based on time_second
when the kernel passes the time to userland programs that expect
the latter, and vice versa.
Note that we shouldn't leak time_uptime to other hosts over the
netowrk. My investigation shows there is no such leak:
http://mail-index.netbsd.org/tech-net/2015/08/06/msg005332.html
Discussed on tech-kern and tech-net.
|
| 1.322 | 02-May-2015 |
joerg | Fix !ARP build.
|
| 1.321 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.320 | 26-Mar-2015 |
ozaki-r | Tidy up the regular path of ip_forward
No functional change is intended.
|
| 1.319 | 16-Jun-2014 |
ozaki-r | branches: 1.319.2; 1.319.4; 1.319.6; 1.319.10;
Add 3rd argument to pktq_create to pass sc
It will be used to pass bridge sc for bridge_forward softint.
ok rmind@
|
| 1.318 | 05-Jun-2014 |
rmind | - Implement pktqueue interface for lockless IP input queue.
- Replace ipintrq and ip6intrq with the pktqueue mechanism.
- Eliminate kernel-lock from ipintr() and ip6intr().
- Some preparation work to push softnet_lock out of ipintr().
Discussed on tech-net.
|
| 1.317 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.316 | 29-May-2014 |
rmind | Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.315 | 28-May-2014 |
christos | CID 12164{49,51}: Remove bogus ifp == NULL checks; if ifp was really NULL,
we would have been dead a few lines before the tests.
|
| 1.314 | 23-May-2014 |
rmind | ip_input(), ip_savecontrol(): cache m->m_pkthdr.rcvif in a variable.
|
| 1.313 | 23-May-2014 |
rmind | Make ip_forward() static, there is no need to expose it.
|
| 1.312 | 23-May-2014 |
rmind | Make ip_input() static, there is no need to expose it.
|
| 1.311 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.310 | 19-Mar-2014 |
liamjfoy | branches: 1.310.2;
Remove ipflow_prune and replace with ipflow_reap. ok rmind@
|
| 1.309 | 25-Feb-2014 |
pooka | Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.308 | 29-Jun-2013 |
rmind | - Rewrite parts of pfil(9): use array to store hooks and thus be more cache
friendly (there are only few hooks in the system). Make the structures
opaque and the interface more strict.
- Remove PFIL_HOOKS option by making pfil(9) mandatory.
|
| 1.307 | 27-Jun-2013 |
christos | branches: 1.307.2;
flip src/dst
|
| 1.306 | 27-Jun-2013 |
christos | implement IP_PKTINFO and IP_RECVPKTINFO.
|
| 1.305 | 08-Jun-2013 |
rmind | Split IPsec code in ip_input() and ip_forward() into the separate routines
ipsec4_input() and ipsec4_forward(). Tested by christos@.
|
| 1.304 | 05-Jun-2013 |
christos | IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.303 | 29-Nov-2012 |
christos | Add a new sysctl to mark ports as reserved, so that they are not used in
the anonymous or reserved port allocation.
|
| 1.302 | 25-Jun-2012 |
christos | branches: 1.302.2;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.301 | 22-Jun-2012 |
christos | PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.300 | 02-Jun-2012 |
dsl | Add some pre-processor magic to verify that the type of the data item
passed to sysctl_createv() actually matches the declared type for
the item itself.
In the places where the caller specifies a function and a structure
address (typically the 'softc') an explicit (void *) cast is now needed.
Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c
sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting
AcpiGbl_EnableAmlDebugObject.
(mostly passing the address of a uint64_t when typed as CTLTYPE_INT).
I've test built quite a few kernels, but there may be some unfixed MD
fallout. Most likely passing &char[] to char *.
Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
|
| 1.299 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.298 | 09-Jan-2012 |
liamjfoy | branches: 1.298.2; 1.298.6; 1.298.8;
check against NULL
|
| 1.297 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.296 | 31-Aug-2011 |
plunky | branches: 1.296.2; 1.296.6;
NULL does not need a cast
|
| 1.295 | 03-May-2011 |
dyoung | *_drain() routines may be called with locks held, so instead of doing
any work in *_drain(), set a drain-needed flag. Do the work in the
fasttimo handler.
Contributed by Coyote Point Systems, Inc.
|
| 1.294 | 14-Apr-2011 |
dyoung | In ipintr(), don't overwrite ipintrq.ifq_maxlen with IFQ_MAXLEN.
Initialize ipintrq.ifq_maxlen using IFQ_MAXLEN directly instead of using
the global ipqmaxlen. Get rid of the global ipqmaxlen.
Now it works again to override the maximum IP queue length with, for
example, sysctl -w net.inet.ip.ifq.maxlen=5.
|
| 1.293 | 13-Dec-2010 |
matt | branches: 1.293.2;
Back out rev that shouldn't have been committed.
|
| 1.292 | 11-Dec-2010 |
matt | Add routines to calculate a checkesum if the driver concludes that the
h/w can't do it.
|
| 1.291 | 05-Nov-2010 |
rmind | ip_randomid: make mechanism MP-safe and more modular.
OK matt@
|
| 1.290 | 05-Nov-2010 |
rmind | ip_reass_packet: finish abstraction; some clean-up.
Discussed some time ago with matt@.
|
| 1.289 | 19-Jul-2010 |
rmind | Abstract IP reassembly into single generic routine - ip_reass_packet().
Make struct ipq private and struct ipqent not visible to userland.
Push ip_len adjustment into reassembly layer.
OK matt@
|
| 1.288 | 13-Jul-2010 |
rmind | Split-off IPv4 re-assembly mechanism into a separate module. Abstract
into ip_reass_init(), ip_reass_lookup(), etc (note: abstraction is not
yet complete). No functional changes to the actual mechanism.
OK matt@
|
| 1.287 | 09-Jul-2010 |
rmind | ip_input: move lookup for fragment queue a little bit further. OK matt@.
|
| 1.286 | 01-Apr-2010 |
tls | As suggested by at least 3 different people (the guilty parties know who
they are) avoid repeated kernel_lock/unlock by using an intrq on the stack.
About 5%-10% better from run to run, on my *very* simpleminded test. Can't
possibly be worse.
|
| 1.285 | 31-Mar-2010 |
tls | Don't hold kernel lock across call to ip_input() -- it blocked *all*
hardware interrupts for the length of time it took for all dequeued
packets to flow up the stack (on multiprocessors only). Initial testing
shows performance impact is minimal -- since this temporary fix actually
means taking/releasing the kernel lock per-packet, that seems
acceptable.
Holding the kernel lock across the ip_input() call duplicated the
exclusion intended to be provided by the socket locks/softnet lock
(same lock, for INET/INET6 sockets) and could mask serious bugs. Several
hours' testing didn't turn any up but I'd be surprised if some don't now
appear.
Damon Permezel noticed the problem. Temporary fix suggested by matt@.
|
| 1.284 | 16-Sep-2009 |
pooka | branches: 1.284.2; 1.284.4;
Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.283 | 17-Jul-2009 |
minskim | Delete trailing whitespace.
|
| 1.282 | 16-Jul-2009 |
minskim | Add the IP_RECVTTL option support.
If the IP_RECVTTL option is enabled on a SOCK_DGRAM socket, the
recvmsg(2) call will return the TTL of the received datagram. The
msg_control field in the msghdr structure points to a buffer that
contains a cmsghdr structure followed by the TTL value.
Modeled after FreeBSD implementation.
|
| 1.281 | 18-Apr-2009 |
tsutsui | Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.280 | 15-Apr-2009 |
elad | Remove a few KAUTH_GENERIC_ISSUSER in favor of more descriptive
alternatives.
Discussed on tech-kern:
http://mail-index.netbsd.org/tech-kern/2009/04/11/msg004798.html
Input from ad@, christos@, dyoung@, tsutsui@.
Okay ad@.
|
| 1.279 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.278 | 19-Jan-2009 |
christos | branches: 1.278.2;
Provide compatibility to the old timeval SCM_TIMESTAMP messages.
|
| 1.277 | 17-Dec-2008 |
cegger | kill MALLOC and FREE macros.
|
| 1.276 | 23-Nov-2008 |
rmind | ip_input: fix an IPQ "lock" leak. (hi <matt>!)
|
| 1.275 | 04-Oct-2008 |
pooka | branches: 1.275.2; 1.275.4;
POOL_INIT -> pool_init
|
| 1.274 | 05-Sep-2008 |
seanb | Wrong route being consulted in one place
in ip_forward() after change to rtcache_*().
Restore previous behaviour.
|
| 1.273 | 20-Aug-2008 |
matt | Make the sysctl routines take out softnet_lock before dealing with
any data structures.
Change inet6ctlerrmap and zeroin6_addr to const.
|
| 1.272 | 05-May-2008 |
ad | branches: 1.272.2; 1.272.6;
- Convert hashinit() to use kmem_alloc(). The hash tables can be large
and it's better to not have them in kmem_map.
- Convert a couple of minor items along the way to kmem_alloc().
- Fix some memory leaks.
|
| 1.271 | 04-May-2008 |
thorpej | Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.270 | 02-May-2008 |
ad | PR kern/38497 Out of memory allocating ksiginfo
Work around: don't acquire softnet_lock in protocol drain routines.
|
| 1.269 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.268 | 24-Apr-2008 |
ad | branches: 1.268.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.267 | 23-Apr-2008 |
thorpej | Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.266 | 12-Apr-2008 |
thorpej | branches: 1.266.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.265 | 09-Apr-2008 |
thorpej | - ipflow is not used outside ip_flow.c; move its definition there.
- Make ipflow_reap() private to ip_flow.c, and introduce ipflow_prune()
for external callers to use (avoids returning an ipflow * that is never
actually used anyway).
|
| 1.264 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.263 | 27-Mar-2008 |
cube | - Make sure we send a reasonable fragment size when IPSEC is configured.
Otherwise we end up sending a dubious "0" whenever we cannot find a
proper association for the packet.
- Reset sack_newdata along with snd_nxt to avoid improper integer
arithmetics that lead to sending data from an incorrect place in the
stream, making it appear as corrupted.
Patch by Michael Van Elst, based on an analysis by Michael for the IPSEC
stuff and I for the SACK issue.
|
| 1.262 | 06-Feb-2008 |
matt | branches: 1.262.6;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.261 | 14-Jan-2008 |
dyoung | Use rtcache_validate() instead of rtcache_getrt(). Shorten staircase
in in_losing().
|
| 1.260 | 22-Dec-2007 |
matt | Fix offset calculation.
Make sure that all frags use the same TOS.
|
| 1.259 | 21-Dec-2007 |
matt | Also make sure the first is at 68 bytes long.
|
| 1.258 | 21-Dec-2007 |
matt | Prevent TCP blind data attacks by not allowing non-initial fragments to
start at less than 68 bytes (minimal fragment size).
|
| 1.257 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.256 | 26-Nov-2007 |
yamt | branches: 1.256.2; 1.256.6;
inetctlerrmap: use designated initializer.
|
| 1.255 | 09-Nov-2007 |
kefren | Don't MCLAIM in ipintr() because we do it anyway in ip_input()
|
| 1.254 | 02-Oct-2007 |
dyoung | branches: 1.254.2; 1.254.4;
Delete the unused second argument to ip_stripoptions(), move it
closer to its single caller in if_eon.c, try to move fewer bytes
by moving the IP header forward instead of moving the tail of the
mbuf backward, and use m_adj(9) instead of fiddling directly with
mbuf data members.
|
| 1.253 | 11-Sep-2007 |
degroote | branches: 1.253.2;
In some FAST_IPSEC, spl level is not restored correctly. Fix that.
Spotted by Wolfgang Stukenbrock in pr/36800
|
| 1.252 | 30-Aug-2007 |
dyoung | Use malloc(9) for sockaddrs instead of pool(9), and remove dom_sa_pool
and dom_sa_len members from struct domain. Pools of fixed-size
objects are too rigid for sockaddr_dls, whose size can vary over
a wide range.
Return sockaddr_dl to its "historical" size. Now that I'm using
malloc(9) instead of pool(9) to allocate sockaddr_dl, I can create
a sockaddr_dl of any size in the kernel, so expanding sockaddr_dl
is useless.
Avoid using sizeof(struct sockaddr_dl) in the kernel.
Introduce sockaddr_dl_alloc() for allocating & initializing an
arbitrary sockaddr_dl on the heap.
Add an argument, the sockaddr length, to sockaddr_alloc(),
sockaddr_copy(), and sockaddr_dl_setaddr().
Constify: LLADDR() -> CLLADDR().
Where the kernel overwrites LLADDR(), use sockaddr_dl_setaddr(),
instead. Used properly, sockaddr_dl_setaddr() will not overrun
the end of the sockaddr.
|
| 1.251 | 10-Aug-2007 |
dyoung | branches: 1.251.2;
Use sockaddr_dl_init().
|
| 1.250 | 19-Jul-2007 |
dyoung | branches: 1.250.4; 1.250.6;
Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.249 | 02-May-2007 |
dyoung | branches: 1.249.2;
Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.248 | 25-Mar-2007 |
liamjfoy | Add net.inet.ip.hashsize to control the IPv4 fast forward hash table size.
|
| 1.247 | 24-Mar-2007 |
liamjfoy | Don't call ip*flow_reap if we're just looking up maxflows
|
| 1.246 | 12-Mar-2007 |
ad | branches: 1.246.2; 1.246.4;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.245 | 05-Mar-2007 |
liamjfoy | branches: 1.245.2;
Move ipflow_slowtimo from ip_slowtimo and into in_proto.c
ok matt@
|
| 1.244 | 04-Mar-2007 |
christos | Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.243 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.242 | 29-Jan-2007 |
dyoung | branches: 1.242.2;
Cosmetic: remove extraneous, non-KNF parentheses. Change a
sizeof(type) to a sizeof(*ptr) so the correctness of the statement
is correct "at a glance" (or so I hope).
|
| 1.241 | 22-Dec-2006 |
ad | ipintr(): check if the queue is empty before looping. Hardly a giant
win, but removed 30% of splnet() calls in one local test.
|
| 1.240 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.239 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.238 | 06-Dec-2006 |
dyoung | KNF.
|
| 1.237 | 06-Dec-2006 |
dyoung | KNF.
|
| 1.236 | 16-Nov-2006 |
christos | branches: 1.236.2; 1.236.4;
__unused removal on arguments; approved by core.
|
| 1.235 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.234 | 10-Oct-2006 |
dogcow | change the MOWNER_INIT define to take two args; fix extant struct mowner
decls to use it. Makes options MBUFTRACE compile again and not whinge about
missing structure declarations. (Also makes initialization consistent.)
|
| 1.233 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.232 | 19-Sep-2006 |
elad | Remove ugly (void *) casts from network scope authorization wrapper and
calls to it.
While here, adapt code for system scope listeners to avoid some more
casts (forgotten in previous run).
Update documentation.
|
| 1.231 | 13-Sep-2006 |
elad | branches: 1.231.2;
Don't use KAUTH_RESULT_* where it's not applicable.
Prompted by yamt@.
|
| 1.230 | 08-Sep-2006 |
elad | First take at security model abstraction.
- Add a few scopes to the kernel: system, network, and machdep.
- Add a few more actions/sub-actions (requests), and start using them as
opposed to the KAUTH_GENERIC_ISSUSER place-holders.
- Introduce a basic set of listeners that implement our "traditional"
security model, called "bsd44". This is the default (and only) model we
have at the moment.
- Update all relevant documentation.
- Add some code and docs to help folks who want to actually use this stuff:
* There's a sample overlay model, sitting on-top of "bsd44", for
fast experimenting with tweaking just a subset of an existing model.
This is pretty cool because it's *really* straightforward to do stuff
you had to use ugly hacks for until now...
* And of course, documentation describing how to do the above for quick
reference, including code samples.
All of these changes were tested for regressions using a Python-based
testsuite that will be (I hope) available soon via pkgsrc. Information
about the tests, and how to write new ones, can be found on:
http://kauth.linbsd.org/kauthwiki
NOTE FOR DEVELOPERS: *PLEASE* don't add any code that does any of the
following:
- Uses a KAUTH_GENERIC_ISSUSER kauth(9) request,
- Checks 'securelevel' directly,
- Checks a uid/gid directly.
(or if you feel you have to, contact me first)
This is still work in progress; It's far from being done, but now it'll
be a lot easier.
Relevant mailing list threads:
http://mail-index.netbsd.org/tech-security/2006/01/25/0011.html
http://mail-index.netbsd.org/tech-security/2006/03/24/0001.html
http://mail-index.netbsd.org/tech-security/2006/04/18/0000.html
http://mail-index.netbsd.org/tech-security/2006/05/15/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/01/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/25/0000.html
Many thanks to YAMAMOTO Takashi, Matt Thomas, and Christos Zoulas for help
stablizing kauth(9).
Full credit for the regression tests, making sure these changes didn't break
anything, goes to Matt Fleming and Jaime Fournier.
Happy birthday Randi! :)
|
| 1.229 | 30-Aug-2006 |
christos | branches: 1.229.2;
fix initializer
|
| 1.228 | 30-Jul-2006 |
elad | ugh.. more stuff that's overdue and should not be in 4.0: remove the
sysctl(9) flags CTLFLAG_READONLY[12]. luckily they're not documented
so it's only half regression.
only two knobs used them; proc.curproc.corename (check added in the
existing handler; its CTLFLAG_ANYWRITE, yay) and net.inet.ip.forwsrcrt,
that got its own handler now too.
|
| 1.227 | 07-Jun-2006 |
kardel | merge FreeBSD timecounters from branch simonb-timecounters
- struct timeval time is gone
time.tv_sec -> time_second
- struct timeval mono_time is gone
mono_time.tv_sec -> time_uptime
- access to time via
{get,}{micro,nano,bin}time()
get* versions are fast but less precise
- support NTP nanokernel implementation (NTP API 4)
- further reading:
Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf
NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
|
| 1.226 | 08-May-2006 |
liamjfoy | branches: 1.226.2;
#if -> #ifdef
ok christos
|
| 1.225 | 15-Apr-2006 |
christos | Coverity CID 1134: Protect against NULL deref.
|
| 1.224 | 18-Feb-2006 |
joerg | branches: 1.224.2; 1.224.4; 1.224.6;
Print the source and destination IP in ip_forward's DIAGNOSTIC code
with inet_ntoa, making it more human friendly.
From Liam J. Foy in private mail.
|
| 1.223 | 24-Dec-2005 |
perry | branches: 1.223.2; 1.223.4; 1.223.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.222 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.221 | 01-Nov-2005 |
christos | Don't decrement the ttl, until we are sure that we can forward this packet.
Before if there was no route, we would call icmp_error with a datagram
packet that has an incorrect checksum. (From Liam Foy)
|
| 1.220 | 23-Oct-2005 |
christos | No need to pass an interface when only the mtu is needed. From OpenBSD via
Liam Foy.
|
| 1.219 | 05-Aug-2005 |
elad | branches: 1.219.2;
Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.218 | 28-Jun-2005 |
seanb | branches: 1.218.2;
- Return ICMP_UNREACH_NET when no route found as per
section 4.3.3.1 of rfc1812.
|
| 1.217 | 09-Jun-2005 |
atatat | Properly fix the constipated lossage wrt -Wcast-qual and the sysctl
code. I know it's not the prettiest code, but it seems to work rather
well in spite of itself.
|
| 1.216 | 01-Jun-2005 |
blymn | Unconstify rnode to prevent compile error when GATEWAY option set.
|
| 1.215 | 29-Apr-2005 |
yamt | move decl of inetsw to its own header to avoid array of incomplete type.
found by gcc4. reported by Adam Ciarcinski.
|
| 1.214 | 18-Apr-2005 |
yamt | fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.213 | 29-Mar-2005 |
yamt | ip_reass: clear stale csum_flags.
|
| 1.212 | 26-Feb-2005 |
perry | branches: 1.212.2;
nuke trailing whitespace
|
| 1.211 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.210 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.209 | 24-Jan-2005 |
matt | branches: 1.209.2;
Add IFNET_FOREACH and IFADDR_FOREACH macros and start using them.
|
| 1.208 | 19-Dec-2004 |
christos | branches: 1.208.2;
yamt's changes seem to fix all the checksumming issues. Turn the loopback
checksums back off so we can make sure that everything works.
|
| 1.207 | 17-Dec-2004 |
christos | Turn checksumming on loopback back on until we fix the bugs in it.
Connect over tcp on the loopback is broken:
4729 amq 0.000007 CALL connect(4,0x804f2a0,0x1c)
4729 amq 75.007420 RET connect -1 errno 60 Connection timed out
|
| 1.206 | 15-Dec-2004 |
thorpej | Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.205 | 06-Oct-2004 |
darrenr | Add a comment to document what setting "srcrt" is really on about in ipintr()
|
| 1.204 | 29-Sep-2004 |
christos | PR/27081: Sean Boudreau: ip_input() bad csum count not incremented on sw csum
|
| 1.203 | 25-May-2004 |
atatat | Sysctl descriptions under net subtree (net.key not done)
|
| 1.202 | 02-May-2004 |
darrenr | at line 543, we do a pullup here of hlen bytes into the mbuf,
so these later ones are superfluous.
|
| 1.201 | 01-May-2004 |
matt | Use EVCNT_ATTACH_STATIC{,2}
|
| 1.200 | 25-Apr-2004 |
simonb | Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.199 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.198 | 01-Apr-2004 |
matt | In ip_reass_ttl_descr, make i signed since it's compared to >= 0
|
| 1.197 | 24-Mar-2004 |
atatat | branches: 1.197.2;
Tango on sysctl_createv() and flags. The flags have all been renamed,
and sysctl_createv() now uses more arguments.
|
| 1.196 | 15-Jan-2004 |
itojun | correct typo in 1.94 -> 1.95. pointed out by Shiva Shenoy
|
| 1.195 | 14-Dec-2003 |
thorpej | Fix syntax errors in CHECK_NMBCLUSTER_PARAMS().
|
| 1.194 | 14-Dec-2003 |
jonathan | Second part of hashed IP_reassembly changes:
When under pressure for mbufs or we have too many fragments in the IP
reassembly queue, drop half of all fragments. This multiplicative-drop
strategy ensures we return to a healthy state, even under borderline
denial-of-service from extremely lossy NFS-over-UDP peers.
The multiplicative-drop phase currently drops 50% of fragments, but
has pre-placed support for implementing drop-fractions other than 50%
The threshhold for the `drop-half' phase is the new variable,
ip_maxfrags which is calculated as nmbclusters/4.
ip_input.c now keeps ip_nmbclusters, a cached copy of nmbclusters.
Before using limits derived from nmbclusters, we check if nmbclusters
and ip_nmclusters are equal. If not, we recompute Ip parameters
derived from nmbclusters. Based on a suggestion by Jason Thorpe.
ip_maxfrags is currently auto-recalcuated.
The counters ip_nfrags and ip_nfragpacketsr are now declared static
and uninitialized (bss), to discourage tampering with them.
|
| 1.193 | 12-Dec-2003 |
scw | Make fast-ipsec and ipflow (Fast Forwarding) interoperate.
The idea is that we only clear M_CANFASTFWD if an SPD exists
for the packet. Otherwise, it's safe to add a fast-forward
cache entry for the route.
To make this work properly, we invalidate the entire ipflow
cache if a fast-ipsec key is added or changed.
|
| 1.192 | 08-Dec-2003 |
jonathan | Add new field ipq_nfrags to struct ipq. Maintain count of fragments
(fragments, not fragmented packets) in each queue entry.
Use ipq_nfrags to maintain a count of total fragments in reassembly queue.
|
| 1.191 | 07-Dec-2003 |
jonathan | KNF: s/unsigned/u_int/, in a couple of places I missed.
|
| 1.190 | 06-Dec-2003 |
jonathan | Replace the single global IP reassembly list/listhead, with a
hashtable of list-heads. Independently re-invented, then reworked to
match similar code in FreeBSD.
|
| 1.189 | 04-Dec-2003 |
atatat | Dynamic sysctl.
Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(),
vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all
nodes are registered with the tree, and nodes can be added (or
removed) easily, and I/O to and from the tree is handled generically.
Since the nodes are registered with the tree, the mapping from name to
number (and back again) can now be discovered, instead of having to be
hard coded. Adding new nodes to the tree is likewise much simpler --
the new infrastructure handles almost all the work for simple types,
and just about anything else can be done with a small helper function.
All existing nodes are where they were before (numerically speaking),
so all existing consumers of sysctl information should notice no
difference.
PS - I'm sorry, but there's a distinct lack of documentation at the
moment. I'm working on sysctl(3/8/9) right now, and I promise to
watch out for buses.
|
| 1.188 | 04-Dec-2003 |
scw | ipflow (IP fast forwarding) is not compatible with FAST_IPSEC either.
XXX: The decision whether or not to fast forward should be made
XXX: dynamically. Using the current approach seriously reduces
XXX: routing performance on gateways with IPsec enabled.
|
| 1.187 | 26-Nov-2003 |
itojun | define RANDOM_IP_ID by default (unifdef -DRANDOM_IP_ID).
one use remains in sys/netipsec, which is kept for freebsd source code compat.
|
| 1.186 | 24-Nov-2003 |
scw | For FAST_IPSEC, ipfilter gets to see wire-format IPsec-encapsulated packets
only. Decapsulated packets bypass ipfilter. This mimics current behaviour
for Kame IPsec.
|
| 1.185 | 19-Nov-2003 |
fvdl | Correct number of arguments to sysctl_rdint.
|
| 1.184 | 19-Nov-2003 |
jonathan | Patch back support for (badly) randomized IP ids, by request:
* Include "opt_inet.h" everywhere IP-ids are generated with ip_newid(),
so the RANDOM_IP_ID option is visible. Also in ip_id(), to ensure
the prototype for ip_randomid() is made visible.
* Add new sysctl to enable randomized IP-ids, provided the kernel was
configured with RANDOM_IP_ID. (The sysctl defaults to zero, and is
a read-only zero if RANDOM_IP_ID is not configured).
Note that the implementation of randomized IP ids is still defective,
and should not be enabled at all (even if configured) without
very careful deliberation. Caveat emptor.
|
| 1.183 | 17-Nov-2003 |
jonathan | Diff to netinet/ip_input.c (restore ip_id, initialize) for ip_id fix:
Revert the (default) ip_id algorithm to the pre-randomid algorithm,
due to demonstrated low-period repeated IDs from the randomized IP_id
code. Consensus is that the low-period repetition (much less than
2^15) is not suitable for general-purpose use.
Allocators of new IPv4 IDs should now call the function ip_newid().
Randomized IP_ids is now a config-time option, "options RANDOM_IP_ID".
ip_newid() can use ip_random-id()_IP_ID if and only if configured
with RANDOM_IP_ID. A sysctl knob should be provided.
This API may be reworked in the near future to support linear ip_id
counters per (src,dst) IP-address pair.
|
| 1.182 | 12-Nov-2003 |
itojun | KNF
|
| 1.181 | 11-Nov-2003 |
jonathan | Change global head-of-local-IP-address list from in_ifaddr to
in_ifaddrhead. Recent changes in struct names caused a namespace
collision in fast-ipsec, which are most cleanly fixed by using
"in_ifaddrhead" as the listhead name.
|
| 1.180 | 10-Nov-2003 |
jonathan | Make per-protocol network input queue stats visible to userland via
sysctl. Add a protocol-independent sysctl handler to show the per-protocol
"struct ifq' statistics. Add IP(v4) specific call to the handler.
Other protocols can show their per-protocol input statistics by
allocating a sysclt node and calling sysctl_ifq() with their own struct ifq *.
As posted to tech-kern plus improvements/cleanup suggested by Andrew Brown.
|
| 1.179 | 28-Sep-2003 |
mycroft | Remove some code that breaks AH tunnels completely. The comment describing
the purpose of this code appears to be on crack -- it's talking about
end-to-end authentication, but the purpose of an AH tunnel is NOT end-to-end
authentication; it's authentication of the tunnel endpoints.
NB: This does not fix the fact that IPsec leaks "packet tags."
|
| 1.178 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.177 | 06-Sep-2003 |
itojun | backout previous, we don't know if arc4random() corrides on reboot.
|
| 1.176 | 05-Sep-2003 |
itojun | initialize fragment ID with arc4random, not by time.tv_sec
|
| 1.175 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.174 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.173 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.172 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.171 | 14-Jul-2003 |
itojun | correct igmp. from love
|
| 1.170 | 03-Jul-2003 |
itojun | minor KNF
|
| 1.169 | 30-Jun-2003 |
itojun | branches: 1.169.2;
do not generate ICMP redirect when packet filter alters ip_dst to an
address that reside on the same link. Cedric Berger convinced me that
it is necessary.
|
| 1.168 | 30-Jun-2003 |
itojun | fix indent
|
| 1.167 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.166 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.165 | 11-Apr-2003 |
christos | PR/991: Darren Reed: Add a sysctl (checkinteface) to implement this. This
implementation is taken from FreeBSD, but we default to off.
XXX: We should really do this on a per ifaddr basis as jason suggested.
|
| 1.164 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.163 | 12-Nov-2002 |
itojun | remove all entries in rt timer queue on ip_mtudisc change, instead of
destroying the queue.
|
| 1.162 | 12-Nov-2002 |
itojun | ckout previous - doesn't compile
|
| 1.161 | 12-Nov-2002 |
itojun | update ip_mtudisc sysctl change handling.
|
| 1.160 | 10-Nov-2002 |
itojun | always create pmtud timeout queue, as ip_mtudisc can be tweaked via
sysctl at runtime. From lha@stacken.kth.se
|
| 1.159 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.158 | 23-Sep-2002 |
itojun | revert mtudisc_timeout value to the old one if update falis
|
| 1.157 | 11-Sep-2002 |
itojun | KNF - return is not a function. sync w/kame.
|
| 1.156 | 11-Sep-2002 |
itojun | correct signedness mixup in pointer passing. sync w/kame
|
| 1.155 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.154 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.153 | 13-Jun-2002 |
itojun | set IPv4 parameter to modern value.
- turn on path MTU discovery (previous: turned off)
- ICMPv4 redirect entry timeout = 600 sec (previous: never timeout)
|
| 1.152 | 09-Jun-2002 |
itojun | whitespace
|
| 1.151 | 07-Jun-2002 |
itojun | look at rmx_mtu on IPsec tunnel MTU computation.
From: David Waitzman <djw@bbn.com>
|
| 1.150 | 12-May-2002 |
matt | branches: 1.150.2; 1.150.4;
Eliminate commons.
|
| 1.149 | 12-May-2002 |
wiz | Spelling fixes, from Sergey Svishchev in kern/16650.
|
| 1.148 | 07-May-2002 |
matt | Change struct ipqe to use TAILQ's instead of LIST's (primarily for TCP's
benefit currently). Rework tcp_reass code to optimize the 4 most likely causes
of out-of-order packets: first OoO pkt, next OoO pkt in seq, OoO pkt is part
of new chuck of OoO packets, and the OoO pkt fills the first hole. Add evcnts
to instrument tcp_reass (enabled by the options TCP_REASS_COUNTERS). This is
part 1/2 of tcp_reass changes.
|
| 1.147 | 18-Apr-2002 |
matt | Change test for M_EXT to M_READONLY for MROUTING. We only need to to do
a pullup if we aren't allowed to modify the packet.
|
| 1.146 | 08-Mar-2002 |
thorpej | Pool deals fairly well with physical memory shortage, but it doesn't
deal with shortages of the VM maps where the backing pages are mapped
(usually kmem_map). Try to deal with this:
* Group all information about the backend allocator for a pool in a
separate structure. The pool references this structure, rather than
the individual fields.
* Change the pool_init() API accordingly, and adjust all callers.
* Link all pools using the same backend allocator on a list.
* The backend allocator is responsible for waiting for physical memory
to become available, but will still fail if it cannot callocate KVA
space for the pages. If this happens, carefully drain all pools using
the same backend allocator, so that some KVA space can be freed.
* Change pool_reclaim() to indicate if it actually succeeded in freeing
some pages, and use that information to make draining easier and more
efficient.
* Get rid of PR_URGENT. There was only one use of it, and it could be
dealt with by the caller.
From art@openbsd.org.
|
| 1.145 | 25-Feb-2002 |
itojun | correctly enforce ipsec policy check on forwarding case.
From: Greg Troxel <gdt@ir.bbn.com>, Bill Chiarchiaro <wjc@work.cleartech.com>
|
| 1.144 | 24-Feb-2002 |
martin | Clear M_BCAST and M_MCAST on outgoing mbufs.
Don't copy ttl from the inner packet to the encapsulating packet. Make
the outer ttl sysctl'able. This should close PR 14269 from Jasper Wallace
(change partly from there) and it makes traceroute work over gre tunnels.
|
| 1.143 | 21-Feb-2002 |
itojun | suppress source quence message, based on router-req RFC (also could be abused
as DoS traffic generator). from kjc/kame
|
| 1.142 | 28-Nov-2001 |
darrenr | recompute hlen after calling pfil_run_hooks() in case ip_hl was changed.
|
| 1.141 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.140 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.139 | 04-Nov-2001 |
matt | Change a few variable/tables to const since they are read-only.
|
| 1.138 | 29-Oct-2001 |
simonb | Don't need to include <uvm/uvm_extern.h> just to include <sys/sysctl.h>
anymore.
|
| 1.137 | 17-Sep-2001 |
thorpej | branches: 1.137.2;
Split the pre-computed ifnet checksum flags into Tx and Rx directions.
Add capabilities bits that indicate an interface can only perform
in-bound TCPv4 or UDPv4 checksums. There is at least one Gig-E chip
for which this is true (Level One LXT-1001), and this is also the
case for the Intel i82559 10/100 Ethernet chips.
|
| 1.136 | 06-Aug-2001 |
itojun | branches: 1.136.2;
cache IPsec policy on in6?pcb. most of the lookup operations can be bypassed,
especially when it is a connected SOCK_STREAM in6?pcb. sync with kame.
|
| 1.135 | 02-Jun-2001 |
thorpej | branches: 1.135.2;
Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.134 | 21-May-2001 |
lukem | fix spelo in comment
|
| 1.133 | 16-Apr-2001 |
itojun | give a default value to net.inet.ip.maxfragpackets, to protect us from
"lots of fragmented packets" DoS attack.
the current default value is derived from ipv6 counterpart, which is
a magical value "200". it should be enough for normal systems, not sure
if it is enough when you take hundreds of thousands of tcp connections on
your system. if you have proposal for a better value with concrete reasons,
let me know.
|
| 1.132 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.131 | 27-Mar-2001 |
itojun | net.inet.ip.maxfragpackets defines the maximum size of ip reass queue
(prevents fragment flood from chewing up mbuf memory space).
derived from KAME net.inet6.ip6.maxfragpackets.
|
| 1.130 | 02-Mar-2001 |
itojun | branches: 1.130.2;
increase ipstat.ips_badaddr if the packet fails to pass address checks.
|
| 1.129 | 02-Mar-2001 |
itojun | reject packets with 127/8 on IPv4 src/dst, they must not appear on wire
(RFC1122). torture-tests will be welcomed.
XXX do we want to check source routing headers as well?
|
| 1.128 | 01-Mar-2001 |
itojun | make sure to enforce inbound ipsec policy checking, for any protocols on top
of ip (check it when final header is visited). sync with kame.
XXX kame team will need to re-check policy engine code
|
| 1.127 | 24-Jan-2001 |
itojun | - record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.126 | 28-Dec-2000 |
thorpej | Back out the sledgehammer damage applied by wiz while I was out for
the holiday.
|
| 1.125 | 25-Dec-2000 |
wiz | Back out previous change. It causes NAT to fail, and was CLEARLY
NOT TESTED before it was committed.
|
| 1.124 | 22-Dec-2000 |
thorpej | Slight adjustment to how pfil_head's are registered. Instead of a
"key" and a "dlt", use a "type" (PFIL_TYPE_{AF,IFNET} for now) and
a val/ptr appropriate for that type. This allows for more future
flexibility with the pfil_hook mechanism.
|
| 1.123 | 14-Dec-2000 |
thorpej | Add ALTQ glue. XXX Temporary until ALTQ is changed to use a pfil hook.
|
| 1.122 | 24-Nov-2000 |
itojun | IFA_STATS stability (not complete); don't touch ip if it is NULL.
|
| 1.121 | 11-Nov-2000 |
thorpej | Restructure the PFIL_HOOKS mechanism a bit:
- All packets are passed to PFIL_HOOKS as they come off the wire, i.e.
fields in protocol headers in network order, etc.
- Allow for multiple hooks to be registered, using a "key" and a "dlt".
The "dlt" is a BPF data link type, indicating what type of header is
present.
- INET and INET6 register with key == AF_INET or AF_INET6, and
dlt == DLT_RAW.
- PFIL_HOOKS now take an argument for the filter hook, and mbuf **,
an ifnet *, and a direction (PFIL_IN or PFIL_OUT), thus making them
less IP (really, IP Filter) centric.
Maintain compatibility with IP Filter by adding wrapper functions for
IP Filter.
|
| 1.120 | 08-Nov-2000 |
ad | Update for hashinit() change.
|
| 1.119 | 13-Oct-2000 |
itojun | make sure we don't share external mbuf between m and mcopy, in ip_forward().
should solve PR 11201.
|
| 1.118 | 26-Aug-2000 |
itojun | make sure anonport{min,max} is not negative number
|
| 1.117 | 25-Aug-2000 |
tron | Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
|
| 1.116 | 06-Jul-2000 |
itojun | remove unnecessary #include <netkey/key_debug.h>. from kame.
|
| 1.115 | 28-Jun-2000 |
mrg | <vm/vm.h> -> <uvm/uvm_extern.h>
|
| 1.114 | 10-May-2000 |
itojun | branches: 1.114.4;
add missing boundary checks to ip options processing.
correct timestamp option validation (len and ptr upper/lower bound
based on RFC791).
fill "pointer" field for parameter problem in timestamp option processing.
|
| 1.113 | 10-May-2000 |
itojun | correct more out-of-bounds memory access, if cnt == 1 and optlen > 1.
|
| 1.112 | 06-May-2000 |
sommerfeld | Handle large offsets with very small options correctly.
|
| 1.111 | 31-Mar-2000 |
jdolecek | Slighly improve previous - only include <netinet/ip_mroute.h> if MROUTING
is defined.
|
| 1.110 | 31-Mar-2000 |
jdolecek | include <netinet/ip_mroute.h> for ip_mforward() - needed after
last duplicate prototype sweep (prototype for ip_mforward() used to be in <netinet/ip_var.h>)
|
| 1.109 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.108 | 30-Mar-2000 |
simonb | Delete uninitialised declaration of ip_defttl - there's an initialised
decl earlier in this file.
|
| 1.107 | 10-Mar-2000 |
thorpej | Back out previous, and adjust a comment.
|
| 1.106 | 07-Mar-2000 |
thorpej | Back out part of 1.104 which isn't actually needed.
|
| 1.105 | 03-Mar-2000 |
itojun | remove unnecessary ttl initialization which I mistakingly bringed in
during KAME merge (this is part of WIDE's expeirmental reass code...)
NetBSD PR: 9412
From: Wolfgang Rupprecht <wolfgang@wsrcc.com>
Fix from: ho@crt.se
itojun was notified from: theo
|
| 1.104 | 02-Mar-2000 |
thorpej | Avoid a bug in GCC which manifests itself when processing unaligned
IP options. Problem pointed out by Matt Hargett and Erik Fair, analyzed
by me.
|
| 1.103 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.102 | 20-Feb-2000 |
darrenr | pass "struct pfil_head *" to pfil_add_hook and pfil_remove hook rather
than "struct protosw *".
|
| 1.101 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.100 | 16-Feb-2000 |
itojun | - if ip_dst matches address on !IFF_UP interface, and
- there's no match against addresses on IFF_UP interface,
send icmp unreach if I'm router. drop it if I'm host.
Revised version of PR: 9387 from nrt@iij.ad.jp. Discussed with thorpej+nrt.
|
| 1.99 | 12-Feb-2000 |
thorpej | Typo (Thanks, Havard :-)
|
| 1.98 | 12-Feb-2000 |
thorpej | Small cosmetic change, and note a place where a statistic should be
gathered.
|
| 1.97 | 11-Feb-2000 |
itojun | fix in-kernel packet forwarding loop (till TTL becomes 0) when:
- a packet is delivered to an address X,
- and the address X is configured on my !IFF_UP interface
- and ipforwarding=1
NetBSD PR: 9387
From: nrt@iij.ad.jp
|
| 1.96 | 01-Feb-2000 |
thorpej | Use ifatoia() and sintosa() consistently, rather than using home-grown
casting macros intermixed.
|
| 1.95 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.94 | 26-Oct-1999 |
itojun | disable ipflow (IPv4 fast fowarding) when IPsec is configured into the kernel.
|
| 1.93 | 17-Oct-1999 |
sommerfeld | branches: 1.93.2; 1.93.4;
In ip_forward():
Avoid forwarding ip unicast packets which were contained inside
link-level multicast packets; having M_MCAST still set in the packet
header flags will mean that the packet will get multicast to a bogus
group instead of unicast to the next hop.
Malformed packets like this have occasionally been spotted "in the
wild" on a mediaone cable modem segment which also had multiple netbsd
machines running as router/NAT boxes.
Without this, any subnet with multiple netbsd routers receiving all
multicasts will generate a packet storm on receipt of such a
multicast. Note that we already do the same check here for link-level
broadcasts; ip6_forward already does this as well.
Note that multicast forwarding does not go through ip_forward().
Adding some code to if_ethersubr to sanity check link-level
vs. ip-level multicast addresses might also be worthwhile.
|
| 1.92 | 23-Jul-1999 |
itojun | branches: 1.92.2;
do not include unnecessary include files.
|
| 1.91 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.90 | 06-Jul-1999 |
itojun | sync with KAME/NetBSD 1.4, SNAP kit 19990705.
key changes are:
- icmp6 redirect fix (dst check)
- revised ip6 multicast check for loopback i/f
- several RCS ID cleanups
|
| 1.89 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.88 | 26-Jun-1999 |
sommerfeld | If the new global variable hostzerobroadcast is zero, no longer assume
address zero of each net/subnet is a broadcast address.
(The default value is nonzero, which preserves the current behavior).
This can be set using sysctl; the boot-time default can also be
configured using the HOSTZEROBROADCAST kernel config option.
While we're here, defopt HOSTZEROBROADCAST and SUBNETSARELOCAL
|
| 1.87 | 04-May-1999 |
hwr | It does not make much sense to increase a "output" counter on input.
|
| 1.86 | 03-May-1999 |
thorpej | In INADDR_TO_IA(), skip interfaces which are not up. Revert previous change
to ip_input.c to check the interface status after INADDR_TO_IA().
Fix cooked up by Heiko Rupp and myself.
Fixes PR 7480.
|
| 1.85 | 03-May-1999 |
hwr | Drop packets, that have a Class-D address as source address.
Implements the first half of PR 7003.
|
| 1.84 | 07-Apr-1999 |
proff | tiny KNF change
|
| 1.83 | 07-Apr-1999 |
proff | Prevent reception of packets on downed interfaces (via an up interface).
fixes kern/7327
|
| 1.82 | 27-Mar-1999 |
aidan | branches: 1.82.2;
Added per-addr input/output statistics. Currently just support netatalk
and netinet, currently only tested under netinet.
Disabled by default, enabled by compiling the kernel with option
IFA_STATS. Enabling this feature seems to make the ip_output function
take 13% longer than before, which should be OK for people that need
this feature.
|
| 1.81 | 26-Mar-1999 |
proff | security: test for ip_len < ip_hl <<2 and drop packet accordingly
|
| 1.80 | 19-Jan-1999 |
mycroft | There's just no plausible reason to byte-swap ip_id internally. It's opaque.
|
| 1.79 | 19-Jan-1999 |
mycroft | Don't screw with ip_len; just subtract from it where we actually use the
value.
|
| 1.78 | 19-Jan-1999 |
mycroft | Don't overwrite the checksum fields when checking them. There's no reason to
do this, and it screws up ICMP replies.
XXX The returned IP checksum and length are still wrong.
|
| 1.77 | 11-Jan-1999 |
thorpej | Fix byte order and ip_len inconsistencies in ICMP reply code. Also, fix
some formatting and HTONS(foo) vs. foo = htons(foo) inconsistencies.
PR #6602, Darren Reed.
|
| 1.76 | 19-Dec-1998 |
thorpej | Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.75 | 18-Dec-1998 |
thorpej | Add a lock around the IP fragment reassembly queue, to prevent ip_drain()
from corrupting the queue if called from a device's interrupt context.
Should fix PR #5684.
|
| 1.74 | 13-Nov-1998 |
thorpej | branches: 1.74.2;
Once a fragmented IP packet has been reassembled, recompute the packet
length before passing it up the stack. From FreeBSD.
|
| 1.73 | 08-Oct-1998 |
thorpej | Use the pool allocator for ipflow entries.
|
| 1.72 | 08-Oct-1998 |
thorpej | Use the pool allocator for ipqent structures.
|
| 1.71 | 30-Sep-1998 |
tls | Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.70 | 09-Sep-1998 |
thorpej | Make a diagnostic printf more sensible, PR #5951, Heiko W. Rupp.
|
| 1.69 | 09-Aug-1998 |
mrg | defopt PFIL_HOOKS.
|
| 1.68 | 17-Jul-1998 |
sommerfe | Fix PR5508: ipfil cut-through forwarding causes panic
|
| 1.67 | 01-Jun-1998 |
thorpej | Protect the ipflow_reap() call with splsoftnet.
|
| 1.66 | 24-May-1998 |
thorpej | Fix OBOB in IP timestamp option processing, as noted in FreeBSD PR 6738,
from Jennifer Dawn Meyers <jdm@enteract.com>.
|
| 1.65 | 04-May-1998 |
matt | Default IP flow to being enabled. Add a sysctl to control the maximum
number of flows (net.inet.ip.maxflows). If set to 0, will disable fast
path forwarding.
|
| 1.64 | 01-May-1998 |
thorpej | Allow packet filters to prevent a packet from creating a fast-forwarding
flow, by setting the "can fast forward" flag in the packet header, and
giving a chance for filters to clear the flag. If the flag is still
set after the filters have given it a chance, the packet will be used
to create a fast-forward flow entry.
|
| 1.63 | 29-Apr-1998 |
matt | Add support for "fast" forwarding. Add hooks in if_ethersubr.c and
if_fddisubr.c to fastpath IP forwarding. If ip_forward successfully
forwards a packet, it will create a cache (ipflow) entry. ether_input
and fddi_input will first call ipflow_fastforward with the received
packet and if the packet passes enough tests, it will be forwarded (the
ttl is decremented and the cksum is adjusted incrementally).
|
| 1.62 | 29-Apr-1998 |
matt | defopt GATEWAY
|
| 1.61 | 29-Apr-1998 |
kml | change path MTU timeout value to match RFC 1191
|
| 1.60 | 29-Apr-1998 |
kml | Add support for deletion of routes added by path MTU discovery;
uses new generic route timeout code. Add sysctl for timeout period.
|
| 1.59 | 19-Mar-1998 |
mrg | convert pfil(9) in and out lists from <sys/queue.h> LISTs to TAILQs, and
change pfil_add_hook to put output filters at the tail of the queue,
while continuing to place input filters at the head of the queue. update
the two users of these functions, and document these changes.
fixes PR#4593.
|
| 1.58 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.57 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.56 | 28-Jan-1998 |
thorpej | Use offsetof() from libkern.h
|
| 1.55 | 12-Jan-1998 |
scottr | Use option header file for MROUTING
|
| 1.54 | 05-Jan-1998 |
lukem | enhance ephemeral port allocation code:
* support sysctl net.inet.ip.anonportmin (lowest ephemeral port)
and net.inet.ip.anonportmax (highest ephemeral port).
these can't be set to >65535, < IPPORT_RESERVED (unless IPNOPRIVPORTS
is defined), and anonportmin has to be < anonportmax.
* use a cleaner way of only cycling through the available set once;
this will be useful for when a random allocation scheme is used
* define IPPORT_ANON{MIN,MAX} instead of IPPORT_USER{LOW,HIGH}
|
| 1.53 | 18-Oct-1997 |
kml | branches: 1.53.2;
change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.52 | 17-Oct-1997 |
thorpej | Allow `subnetsarelocal' to be changed via sysctl.
|
| 1.51 | 29-Aug-1997 |
gwr | Tweaks to allow operation with an interface address of 0.0.0.0
(needed for NFS mountroot using BOOTP to get boot parameters)
|
| 1.50 | 24-Jun-1997 |
thorpej | branches: 1.50.4;
Eliminate use of dtom() from the network code, allowing more flexible
use of mbuf external storage and increasing performance (by eliminating
an m_pullup() for clusters in the IP reassembly code).
Changes from Koji Imada <koji@math.human.nagoya-u.ac.jp>, in PR #3628
and #3480, with ever-so-slight integration changes by me.
|
| 1.49 | 15-Apr-1997 |
christos | Move the mtod calls *after* we've made sure that the packet has passed the
filter successfully. Otherwise it can be NULL if the filter blocked it,
and we die. How did this ever work?
|
| 1.48 | 26-Feb-1997 |
mrg | allow src-routed packetd by default, per host requirements
|
| 1.47 | 25-Feb-1997 |
cjs | Add net.inet.ip.allowsrcrt option which allows/drops all source
routed packets. This currently defaults to `drop,' but once we
verify that all applications that rely on determining remote IP
addresses for authentication are dropping the connection when they
see a source route option (not just disabling the source route
option), we can turn this back on and conform with the host
requirements.
|
| 1.46 | 19-Feb-1997 |
cjs | Fix bug in sysctl net.inet.ip.forwsrcrt handing: now you can read it
if securelevel > 0. (Thanks, cgd.)
|
| 1.45 | 18-Feb-1997 |
mrg | pseudo-device ipfilter brings in PFIL_HOOKS.
|
| 1.44 | 11-Jan-1997 |
thorpej | branches: 1.44.4;
Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.43 | 20-Dec-1996 |
mrg | in pfil_hooks: always reassign ip after calling hook.
|
| 1.42 | 20-Dec-1996 |
mrg | remove pfil_bad.
|
| 1.41 | 25-Oct-1996 |
thorpej | Before concatenating frags, sanity check the length of the packet. If it's
larger than IP_MAXPACKET, discard it.
Based on a patch from Bill Fenner <fenner@parc.xerox.com>
|
| 1.40 | 22-Oct-1996 |
veego | Fix a panic from the pfil_hooks.
|
| 1.39 | 13-Oct-1996 |
christos | backout previous kprintf changes
|
| 1.38 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.37 | 21-Sep-1996 |
perry | commit fix in pr 2772 -- the IP input code was assuming that the
reserved (must be zero) flag must necessarily be zero. We now define
an IP_RF (by analogy to IP_DF and IP_MF) and mask it out when necessary.
|
| 1.36 | 14-Sep-1996 |
mrg | move the packet filter hooks in to a saner location. while i'm here, rename
PACKET_FILTER to PFIL_HOOKS.
|
| 1.35 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.34 | 08-Sep-1996 |
mycroft | Save 68 bytes of the packet for ICMP, not 64. From Laine Stump, PR 2296.
|
| 1.33 | 06-Sep-1996 |
mrg | add packet filter interface code. see pfil(9) for more details. you
need the PACKET_FILTER option to enable this code. currently, ipfilter
version 3.1.1-beta has been converted to use this new interface.
|
| 1.32 | 14-Aug-1996 |
thorpej | Fix some DIAGNOSTIC printf() formats; ntohl() provides a 32-bit quantity,
and should be printed with %x, not %lx.
|
| 1.31 | 10-Jul-1996 |
cgd | print result of ntohl/htonl as a long. (makes -Wformat work on the
Alpha.)
|
| 1.30 | 16-Mar-1996 |
christos | branches: 1.30.4;
Fix printf format args.
|
| 1.29 | 26-Feb-1996 |
mrg | two more local addr changes, all done differently now (idea from charles)
|
| 1.28 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.27 | 16-Jan-1996 |
thorpej | Add a net.inet.ip.directed-broadcast sysctl as suggested by
Darren Reed <darrenr@vitruvius.arbld.unimelb.edu.au> in PR #1227.
This change is slightly different than the one submitted by Darren in
that the DIRECTED_BROADCAST compile-time option will behave like it used
to so that existing configurations utilizing it won't have to change.
|
| 1.26 | 15-Jan-1996 |
thorpej | Add net.inet.ip.forwsrcrt: if zero, the system will not forward
source-routed packets. Note this value is protected by kernel security
level; it can only be changed if securelevel < 1.
|
| 1.25 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.24 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.23 | 12-Jun-1995 |
mycroft | Change in_pcbnotify*() to take an errno value. Make inetctlerrmap[] an
array on ints, not u_chars.
|
| 1.22 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.21 | 07-Jun-1995 |
mycroft | Remove ip_ifmatrix completely.
|
| 1.20 | 04-Jun-1995 |
mycroft | Don't cast things unnecessarily.
|
| 1.19 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.18 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.17 | 15-May-1995 |
cgd | oops; forgot a '{'
|
| 1.16 | 14-May-1995 |
cgd | drop (and record) malformed IP fragments. Fixes pr 1030 (differently).
|
| 1.15 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.14 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.13 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.12 | 14-Feb-1994 |
mycroft | PARANOID --> DIAGNOSTIC for inexpensive tests.
|
| 1.11 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.10 | 29-Jan-1994 |
brezak | Fix some cases of NOT dealing with m_pkthdr's. This code is still suspect though, at least this fixes some panics.
|
| 1.9 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.8 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.7 | 08-Jan-1994 |
mycroft | More prototypes.
|
| 1.6 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 20-May-1993 |
cgd | branches: 1.3.4;
more rcsid additions and file header cleanups
|
| 1.2 | 04-May-1993 |
cgd | make ip_input recursion checking be for -DPARANOID, and make it panic
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.3.4.2 | 14-Nov-1993 |
mycroft | PARANOID --> DIAGNOSTIC. These are not expensive tests.
|
| 1.3.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.30.4.3 | 11-Dec-1996 |
mycroft | From trunk:
Save 68 bytes of the packet for ICMP, not 64.
|
| 1.30.4.2 | 11-Dec-1996 |
mycroft | From trunk:
Ignore the reserved fragment flag when checking ip_off.
|
| 1.30.4.1 | 10-Nov-1996 |
thorpej | Update from trunk:
- Make ip_len and ip_off unsigned.
- Make sure we don't accept or transmit packets larger than the
maximim IP packet size.
This fixes the so-called `death ping' bug.
Sum of work from Bill Fenner <fenner@parc.xerox.com>,
Kevin Lahey <kml@nas.nasa.gov>, and myself.
Thanks to Curt Sampson, Jukka Marin, and Kevin Lahey for testing
this under NetBSD 1.2
|
| 1.44.4.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.50.4.1 | 01-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.53.2.4 | 15-Nov-1998 |
cgd | pull up rev 1.74 from trunk (thorpej)
|
| 1.53.2.3 | 01-Oct-1998 |
cgd | pull up revisions 1.57-1.58 (via patch), 1.71 (via patch) from trunk. (tls)
|
| 1.53.2.2 | 22-Jul-1998 |
mellon | Pull up 1.59 and 1.68 (veego)
|
| 1.53.2.1 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.74.2.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.82.2.7 | 30-May-2001 |
he | Pull up revisions 1.131,1.133 (via patch, requested by he):
Introduce net.inet.ip.maxfragpackets, which controls the maximum
number of IPv4 fragment reassembly queue entries. Defends against
certain DoS attacks. Fixes SA#2001-006.
|
| 1.82.2.6 | 06-May-2000 |
he | Pull up revision 1.112 (requested by sommerfeld):
Handle large offsets inside very small options correctly.
|
| 1.82.2.5 | 02-Mar-2000 |
he | Pull up revision 1.104 (requested by thorpej):
Work around a compiler bug that causes a security vulnerability
in our IP stack on some platforms.
|
| 1.82.2.4 | 12-Feb-2000 |
he | Apply patch (requested by thorpej):
Adhere to RFC 1112 and RFC 1122 by dropping incoming packets with
a multicast source address. Fixes part of PR#7003.
|
| 1.82.2.3 | 17-Oct-1999 |
cgd | pull up rev 1.93 from trunk (requested by sommerfeld):
Multicast storm prevention: don't attempt to forward link-level
multicast packets which contain ip unicast packets; these packets
would only be generated from misconfigured/buggy systems.
|
| 1.82.2.2 | 03-May-1999 |
perry | branches: 1.82.2.2.2; 1.82.2.2.4;
pullup 1.85->1.86 (thorpej)
|
| 1.82.2.1 | 07-Apr-1999 |
proff | pullup 1.82 - 1.83; don't receive packets on downed interface addresses
|
| 1.82.2.2.4.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.82.2.2.4.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.82.2.2.4.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.82.2.2.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.82.2.2.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.82.2.2.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.92.2.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.93.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.93.2.8 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.93.2.7 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.93.2.6 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.93.2.5 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.93.2.4 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.93.2.3 | 08-Dec-2000 |
bouyer | Sync with HEAD.
|
| 1.93.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.93.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.114.4.10 | 13-Nov-2002 |
itojun | sys/net/route.c 1.55 via patch
sys/net/route.h 1.32
sys/netinet/ip_input.c 1.163
Remove all entries on rt timer queue on ip_mtudisc change, instead
of destroying the queue.
(itojun, redo)
|
| 1.114.4.9 | 10-Nov-2002 |
itojun | sys/netinet/ip_input.c 1.160 via patch
Always create PMTUD timeout queue, as PMTUD can be turned on via
sysctl at runtime. From lha@stacken.kth.se.
(itojun)
|
| 1.114.4.8 | 26-Feb-2002 |
he | Pull up revision 1.145 (requested by itojun):
Correctly enforce ipsec policy check in IPv4 forwarding case.
|
| 1.114.4.7 | 26-Feb-2002 |
he | Pull up revision 1.144 (requested by martin):
Clear M_BCAST and M_MCAST on encapsulated packets on outgoing
mbufs. Also do not copy TTL from the inner packet, and make the
outer TTL sysctl'able. Fixes PR#14269, and makes traceroute work
over GRE tunnels.
|
| 1.114.4.6 | 24-Apr-2001 |
he | Pull up revisions 1.131,1.133 (requested by itojun):
Introduce net.inet.ip.maxfragpackets, which controls the maximum
number of IPv4 fragment reassembly queue entries. Defends against
certain DoS attacks.
|
| 1.114.4.5 | 06-Apr-2001 |
he | Pull up revision 1.127 (via patch, requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.114.4.4 | 11-Mar-2001 |
he | Pull up revision 1.128 (requested by itojun):
Ensure that we enforce inbound IPsec policy on all IP protocols,
not just TCP, UDP and ICMP.
|
| 1.114.4.3 | 17-Oct-2000 |
tv | Pullup 1.119 [itojun]:
make sure we don't share external mbuf between m and mcopy, in ip_forward().
should solve PR 11201.
|
| 1.114.4.2 | 27-Aug-2000 |
itojun | pullup 1.117 -> 1.118 (approved by releng-1-5)
> make sure anonport{min,max} is not negative number
|
| 1.114.4.1 | 26-Aug-2000 |
tron | Pull up from current (approved by thorpej):
Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
syssrc/sys/netinet/in.h 1.49 -> 1.50
syssrc/sys/netinet/in_pcb.c 1.66 -> 1.67
syssrc/sys/netinet/ip_input.c 1.116 -> 1.117
syssrc/sys/netinet/ip_var.h 1.41 -> 1.42
|
| 1.130.2.16 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.130.2.15 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.130.2.14 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.13 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.12 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.11 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.10 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.9 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.130.2.8 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.130.2.7 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.6 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.130.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.130.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.130.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.130.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.130.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.135.2.6 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.135.2.5 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.135.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.135.2.3 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.135.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.135.2.1 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.136.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.137.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.150.4.3 | 17-Jun-2003 |
msaitoh | Pullup rev. 1.163 via patch (requested by itojun in ticket #984):
remove all entries in rt timer queue on ip_mtudisc change, instead of
destroying the queue.
|
| 1.150.4.2 | 12-Nov-2002 |
tron | Pull up revision 1.160 (requested by itojun in ticket #977):
always create pmtud timeout queue, as ip_mtudisc can be tweaked via
sysctl at runtime. From lha@stacken.kth.se
|
| 1.150.4.1 | 07-Jun-2002 |
thorpej | pullup-1-6 ticket #202:
syssrc/sys/netinet/ip_input.c 1.151
Original log message:
look at rmx_mtu on IPsec tunnel MTU computation.
From: David Waitzman <djw@bbn.com>
|
| 1.150.2.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.150.2.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.150.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.169.2.8 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.169.2.7 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.169.2.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.169.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.169.2.4 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.169.2.3 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.169.2.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.169.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.197.2.1 | 28-May-2004 |
tron | Pull up revision 1.203 (requested by atatat in ticket #391):
Sysctl descriptions under net subtree (net.key not done)
|
| 1.208.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.209.2.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.209.2.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.212.2.3 | 17-Sep-2007 |
bouyer | Pull up following revision(s) (requested by degroote in ticket #1840):
sys/netinet/ip_input.c: revision 1.253
In some FAST_IPSEC, spl level is not restored correctly. Fix that.
Spotted by Wolfgang Stukenbrock in pr/36800
|
| 1.212.2.2 | 06-May-2005 |
tron | branches: 1.212.2.2.2; 1.212.2.2.4;
Pull up revision 1.214 (requested by yamt in ticket #251):
fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.212.2.1 | 04-Apr-2005 |
tron | Pull up revision 1.213 (requested by yamt in ticket #88):
ip_reass: clear stale csum_flags.
|
| 1.212.2.2.4.1 | 17-Sep-2007 |
bouyer | Pull up following revision(s) (requested by degroote in ticket #1840):
sys/netinet/ip_input.c: revision 1.253
In some FAST_IPSEC, spl level is not restored correctly. Fix that.
Spotted by Wolfgang Stukenbrock in pr/36800
|
| 1.212.2.2.2.1 | 17-Sep-2007 |
bouyer | Pull up following revision(s) (requested by degroote in ticket #1840):
sys/netinet/ip_input.c: revision 1.253
In some FAST_IPSEC, spl level is not restored correctly. Fix that.
Spotted by Wolfgang Stukenbrock in pr/36800
|
| 1.218.2.9 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.218.2.8 | 21-Jan-2008 |
yamt | sync with head
|
| 1.218.2.7 | 07-Dec-2007 |
yamt | sync with head
|
| 1.218.2.6 | 15-Nov-2007 |
yamt | sync with head.
|
| 1.218.2.5 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.218.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.218.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.218.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.218.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.219.2.2 | 02-Nov-2005 |
yamt | sync with head.
|
| 1.219.2.1 | 26-Oct-2005 |
yamt | sync with head
|
| 1.223.6.3 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.223.6.2 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.223.6.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time(), use bintime's for timeout
calculations and use "time_second" instead of "time.tv_sec".
|
| 1.223.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.223.2.1 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.224.6.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.224.4.2 | 11-May-2006 |
elad | sync with head
|
| 1.224.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.224.2.5 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.224.2.4 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.224.2.3 | 11-Aug-2006 |
yamt | sync with head
|
| 1.224.2.2 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.224.2.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.226.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.229.2.3 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.229.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.229.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.231.2.3 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.231.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.231.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.236.4.2 | 03-Jun-2008 |
skrll | Sync with netbsd-4.
|
| 1.236.4.1 | 23-Sep-2007 |
wrstuden | Sync with somewhat-recent netbsd-4.
|
| 1.236.2.2 | 30-Mar-2008 |
jdc | Pull up revisions:
src/sys/netinet/ip_input.c 1.263
src/sys/netinet/tcp_subr.c 1.225
(requested by cube in ticket #1109).
- Make sure we send a reasonable fragment size when IPSEC is configured.
Otherwise we end up sending a dubious "0" whenever we cannot find a
proper association for the packet.
- Reset sack_newdata along with snd_nxt to avoid improper integer
arithmetics that lead to sending data from an incorrect place in the
stream, making it appear as corrupted.
Patch by Michael Van Elst, based on an analysis by Michael for the IPSEC
stuff and I for the SACK issue.
|
| 1.236.2.1 | 16-Sep-2007 |
xtraeme | branches: 1.236.2.1.4;
Pull up following revision(s) (requested by degroote in ticket #881):
sys/netinet/ip_input.c: revision 1.253
sys/netinet6/ip6_input.c: revision 1.110
In some FAST_IPSEC, spl level is not restored correctly. Fix that.
Spotted by Wolfgang Stukenbrock in pr/36800
|
| 1.236.2.1.4.1 | 30-Mar-2008 |
jdc | Pull up revisions:
src/sys/netinet/ip_input.c 1.263
src/sys/netinet/tcp_subr.c 1.225
(requested by cube in ticket #1109).
- Make sure we send a reasonable fragment size when IPSEC is configured.
Otherwise we end up sending a dubious "0" whenever we cannot find a
proper association for the packet.
- Reset sack_newdata along with snd_nxt to avoid improper integer
arithmetics that lead to sending data from an incorrect place in the
stream, making it appear as corrupted.
Patch by Michael Van Elst, based on an analysis by Michael for the IPSEC
stuff and I for the SACK issue.
|
| 1.242.2.5 | 07-May-2007 |
yamt | sync with head.
|
| 1.242.2.4 | 15-Apr-2007 |
yamt | sync with head.
|
| 1.242.2.3 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.242.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.242.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.245.2.5 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.245.2.4 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.245.2.3 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.245.2.2 | 10-Apr-2007 |
ad | Sync with head.
|
| 1.245.2.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.246.4.1 | 29-Mar-2007 |
reinoud | Pullup to -current
|
| 1.246.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.249.2.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.249.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.250.6.2 | 19-Jul-2007 |
dyoung | Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.250.6.1 | 19-Jul-2007 |
dyoung | file ip_input.c was added on branch matt-mips64 on 2007-07-19 20:48:56 +0000
|
| 1.250.4.6 | 27-Nov-2007 |
joerg | Sync with HEAD. amd64 Xen support needs testing.
|
| 1.250.4.5 | 11-Nov-2007 |
joerg | Sync with HEAD.
|
| 1.250.4.4 | 04-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.250.4.3 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.250.4.2 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.250.4.1 | 16-Aug-2007 |
jmcneill | Sync with HEAD.
|
| 1.251.2.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.251.2.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.251.2.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.253.2.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.254.4.4 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.254.4.3 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.254.4.2 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.254.4.1 | 19-Nov-2007 |
mjf | Sync with HEAD.
|
| 1.254.2.1 | 13-Nov-2007 |
bouyer | Sync with HEAD
|
| 1.256.6.2 | 19-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.256.6.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.256.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.262.6.5 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.262.6.4 | 05-Oct-2008 |
mjf | Sync with HEAD.
|
| 1.262.6.3 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.262.6.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.262.6.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.266.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.268.2.6 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.268.2.5 | 11-Mar-2010 |
yamt | sync with head
|
| 1.268.2.4 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.268.2.3 | 18-Jul-2009 |
yamt | sync with head.
|
| 1.268.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.268.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.272.6.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.272.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.272.2.2 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.272.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.275.4.1 | 25-Nov-2008 |
snj | branches: 1.275.4.1.8;
Pull up following revision(s) (requested by rmind in ticket #119):
sys/netinet/ip_input.c: revision 1.276
ip_input: fix an IPQ "lock" leak. (hi <matt>!)
|
| 1.275.4.1.8.2 | 07-Jan-2011 |
matt | Backout an inadverdant change.
|
| 1.275.4.1.8.1 | 07-Jan-2011 |
matt | If using hardware checksum offload and the packet can't be h/w checksumed
(for whatever reason, some hardware is stupid) allow the driver to calculate
the checksum instead.
|
| 1.275.2.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.275.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.278.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.278.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.284.4.4 | 31-May-2011 |
rmind | sync with head
|
| 1.284.4.3 | 21-Apr-2011 |
rmind | sync with head
|
| 1.284.4.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.284.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.284.2.3 | 06-Nov-2010 |
uebayasi | Sync with HEAD.
|
| 1.284.2.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.284.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.293.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.296.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.296.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.296.2.4 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.296.2.3 | 16-Jan-2013 |
yamt | sync with (a bit old) head
|
| 1.296.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.296.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.298.8.1 | 09-Feb-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1526):
sys/netinet/ip_input.c: revision 1.366
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.298.6.1 | 09-Feb-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1526):
sys/netinet/ip_input.c: revision 1.366
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.298.2.1 | 09-Feb-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1526):
sys/netinet/ip_input.c: revision 1.366
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.302.2.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.302.2.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.302.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.302.2.1 | 25-Feb-2013 |
tls | resync with head
|
| 1.307.2.3 | 18-May-2014 |
rmind | sync with head
|
| 1.307.2.2 | 28-Aug-2013 |
rmind | sync with head
|
| 1.307.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.310.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.319.10.2 | 17-Sep-2019 |
martin | Pull up following revision(s) (requested by bouyer in ticket #1708):
sys/netinet6/ip6_input.c: revision 1.209 via patch
sys/netinet/ip_input.c: revision 1.390 via patch
Packet filters can return an mbuf chain with fragmented headers, so
m_pullup() it if needed and remove the KASSERT()s.
|
| 1.319.10.1 | 09-Feb-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1563):
sys/netinet/ip_input.c: revision 1.366 (via patch)
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.319.6.2 | 17-Sep-2019 |
martin | Pull up following revision(s) (requested by bouyer in ticket #1708):
sys/netinet6/ip6_input.c: revision 1.209 via patch
sys/netinet/ip_input.c: revision 1.390 via patch
Packet filters can return an mbuf chain with fragmented headers, so
m_pullup() it if needed and remove the KASSERT()s.
|
| 1.319.6.1 | 09-Feb-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1563):
sys/netinet/ip_input.c: revision 1.366 (via patch)
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.319.4.10 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.319.4.9 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.319.4.8 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.319.4.7 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.319.4.6 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.319.4.5 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.319.4.4 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.319.4.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.319.4.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.319.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.319.2.2 | 17-Sep-2019 |
martin | Pull up following revision(s) (requested by bouyer in ticket #1708):
sys/netinet6/ip6_input.c: revision 1.209 via patch
sys/netinet/ip_input.c: revision 1.390 via patch
Packet filters can return an mbuf chain with fragmented headers, so
m_pullup() it if needed and remove the KASSERT()s.
|
| 1.319.2.1 | 09-Feb-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1563):
sys/netinet/ip_input.c: revision 1.366 (via patch)
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.337.2.5 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.337.2.4 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.337.2.3 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.337.2.2 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.337.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.347.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.355.2.9 | 07-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1661):
sys/netinet6/ip6_id.c: revision 1.19-1.21
sys/netinet6/ip6_var.h: revision 1.88
sys/netinet/ip_input.c: revision 1.400
sys/netinet/tcp_subr.c: revision 1.285
sys/netinet/ip6.h: revision 1.30
netinet: Enable random IP fragment ids by default (from riastradh)
netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
netinet6: Mark randomid unused.
Will make merging and bisection easier if anything goes wrong with
flow label or fragment id randomization changes.
(from riastradh)
netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
Replace randomid() by cprng_fast32()
|
| 1.355.2.8 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1385):
sys/net/if.c 1.461
sys/net/if.h 1.277
sys/net/if_gif.c 1.149
sys/net/if_gif.h 1.33
sys/net/if_ipsec.c 1.19,1.20,1.24
sys/net/if_ipsec.h 1.5
sys/net/if_l2tp.c 1.33,1.36-1.39
sys/net/if_l2tp.h 1.7,1.8
sys/net/route.c 1.220,1.221
sys/net/route.h 1.125
sys/netinet/in_gif.c 1.95
sys/netinet/in_l2tp.c 1.17
sys/netinet/ip_input.c 1.391,1.392
sys/netinet/wqinput.c 1.6
sys/netinet6/in6_gif.c 1.94
sys/netinet6/in6_l2tp.c 1.18
sys/netinet6/ip6_forward.c 1.97
sys/netinet6/ip6_input.c 1.210,1.211
sys/netipsec/ipsec_output.c 1.82,1.83 (patched)
sys/netipsec/ipsecif.c 1.12,1.13,1.15,1.17 (patched)
sys/netipsec/key.c 1.259,1.260
ipsecif(4) support input drop packet counter.
ipsecif(4) should not increment drop counter by errors not related to if_snd. Pointed out by ozaki-r@n.o, thanks.
Remove unnecessary addresses in PF_KEY message.
MOBIKE Extensions for PF_KEY draft-schilcher-mobike-pfkey-extension-01.txt says
|
| 1.355.2.7 | 17-Sep-2019 |
martin | Pull up following revision(s) (requested by bouyer in ticket #1378):
sys/netinet6/ip6_input.c: revision 1.209 (patch)
sys/netinet/ip_input.c: revision 1.390 (patch)
Packet filters can return an mbuf chain with fragmented headers, so
m_pullup() it if needed and remove the KASSERT()s.
|
| 1.355.2.6 | 18-Mar-2018 |
martin | Pull up following revision(s) (requested by tih in ticket #639):
sys/kern/uipc_socket.c: revision 1.258
sys/kern/uipc_socket.c: revision 1.259
sys/netinet/ip_input.c: revision 1.364 (via patch)
sys/netinet/ip_output.c: revision 1.289
sys/netinet/in.h: revision 1.102
sys/netinet/in_pcb.c: revision 1.181
share/man/man9/sockopt.9: revision 1.11
sys/netinet/in_pcb.h: revision 1.65
sys/sys/socketvar.h: revision 1.146
sys/kern/uipc_syscalls.c: revision 1.189
sys/netinet/ip_output.c: revision 1.290
share/man/man4/ip.4: revision 1.41
share/man/man4/ip.4: revision 1.42
sys/kern/uipc_syscalls.c: revision 1.190
pass valsize for getsockopt like we do for setsockopt
make sure that we have enough space, don't require the exact size
(Tom Ivar Helbekkmo)
1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
new sentence-new line
Remove comment now that the getsockopt code passes the size.
Add a new sockopt member to keep track of the actual size of the option
that should be returned to the caller in getsockopt(2).
(Tom Ivar Helbekkmo)
|
| 1.355.2.5 | 26-Feb-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #588):
sys/netinet6/in6.c: revision 1.260
sys/netinet/in.c: revision 1.219
sys/netinet/wqinput.c: revision 1.4
sys/rump/net/lib/libnetinet/netinet_component.c: revision 1.11
sys/netinet/ip_input.c: revision 1.376
sys/netinet6/ip6_input.c: revision 1.193
Avoid a deadlock between softnet_lock and IFNET_LOCK
A deadlock occurs because there is a violation of the rule of lock ordering;
softnet_lock is held with hodling IFNET_LOCK, which violates the rule.
To avoid the deadlock, replace softnet_lock in in_control and in6_control
with KERNEL_LOCK.
We also need to add some KERNEL_LOCKs to protect the network stack surely.
This is required, for example, for PR kern/51356.
Fix PR kern/53043
|
| 1.355.2.4 | 12-Feb-2018 |
snj | Pull up following revision(s) (requested by maxv in ticket #547):
sys/netinet/ip_input.c: 1.366
Disable ip_allowsrcrt and ip_forwsrcrt. Enabling them by default was a
completely dumb idea, because they have security implications.
By sending an IPv4 packet containing an LSRR option, an attacker will
cause the system to forward the packet to another IPv4 address - and
this way he white-washes the source of the packet.
It is also possible for an attacker to reach hidden networks: if a server
has a public address, and a private one on an internal network (network
which has several internal machines connected), the attacker can send a
packet with:
source = 0.0.0.0
destination = public address of the server
LSRR first address = address of a machine on the internal network
And the packet will be forwarded, by the server, to the internal machine,
in some cases even with the internal IP address of the server as a source.
|
| 1.355.2.3 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.355.2.2 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by roy in ticket #390):
sys/netinet/ip_input.c: 1.363
sys/netinet6/ip6_input.c: 1.184-1.185
sys/netinet6/ip6_output.c: 1.194-1.195
sys/netinet6/in6_src.c: 1.83-1.84
Allow local communication over DETACHED addresses.
Allow binding to DETACHED or TENTATIVE addresses as we deny
sending upstream from them anyway.
Prefer non DETACHED or TENTATIVE addresses.
--
Attempt to restore v6 networking. Not 100% certain that these
changes are all that is needed, but they're certainly a big part of it
(especially the ip6_input.c change.)
--
Treat unvalidated addresses as deprecated in rule 3.
|
| 1.355.2.1 | 21-Oct-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #300):
crypto/dist/ipsec-tools/src/setkey/parse.y: 1.19
crypto/dist/ipsec-tools/src/setkey/token.l: 1.20
distrib/sets/lists/tests/mi: 1.754, 1.757, 1.759
doc/TODO.smpnet: 1.12-1.13
sys/net/pfkeyv2.h: 1.32
sys/net/raw_cb.c: 1.23-1.24, 1.28
sys/net/raw_cb.h: 1.28
sys/net/raw_usrreq.c: 1.57-1.58
sys/net/rtsock.c: 1.228-1.229
sys/netinet/in_proto.c: 1.125
sys/netinet/ip_input.c: 1.359-1.361
sys/netinet/tcp_input.c: 1.359-1.360
sys/netinet/tcp_output.c: 1.197
sys/netinet/tcp_var.h: 1.178
sys/netinet6/icmp6.c: 1.213
sys/netinet6/in6_proto.c: 1.119
sys/netinet6/ip6_forward.c: 1.88
sys/netinet6/ip6_input.c: 1.181-1.182
sys/netinet6/ip6_output.c: 1.193
sys/netinet6/ip6protosw.h: 1.26
sys/netipsec/ipsec.c: 1.100-1.122
sys/netipsec/ipsec.h: 1.51-1.61
sys/netipsec/ipsec6.h: 1.18-1.20
sys/netipsec/ipsec_input.c: 1.44-1.51
sys/netipsec/ipsec_netbsd.c: 1.41-1.45
sys/netipsec/ipsec_output.c: 1.49-1.64
sys/netipsec/ipsec_private.h: 1.5
sys/netipsec/key.c: 1.164-1.234
sys/netipsec/key.h: 1.20-1.32
sys/netipsec/key_debug.c: 1.18-1.21
sys/netipsec/key_debug.h: 1.9
sys/netipsec/keydb.h: 1.16-1.20
sys/netipsec/keysock.c: 1.59-1.62
sys/netipsec/keysock.h: 1.10
sys/netipsec/xform.h: 1.9-1.12
sys/netipsec/xform_ah.c: 1.55-1.74
sys/netipsec/xform_esp.c: 1.56-1.72
sys/netipsec/xform_ipcomp.c: 1.39-1.53
sys/netipsec/xform_ipip.c: 1.50-1.54
sys/netipsec/xform_tcp.c: 1.12-1.16
sys/rump/librump/rumpkern/Makefile.rumpkern: 1.170
sys/rump/librump/rumpnet/net_stub.c: 1.27
sys/sys/protosw.h: 1.67-1.68
tests/net/carp/t_basic.sh: 1.7
tests/net/if_gif/t_gif.sh: 1.11
tests/net/if_l2tp/t_l2tp.sh: 1.3
tests/net/ipsec/Makefile: 1.7-1.9
tests/net/ipsec/algorithms.sh: 1.5
tests/net/ipsec/common.sh: 1.4-1.6
tests/net/ipsec/t_ipsec_ah_keys.sh: 1.2
tests/net/ipsec/t_ipsec_esp_keys.sh: 1.2
tests/net/ipsec/t_ipsec_gif.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_l2tp.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_misc.sh: 1.8-1.18
tests/net/ipsec/t_ipsec_sockopt.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tcp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_transport.sh: 1.5-1.6
tests/net/ipsec/t_ipsec_tunnel.sh: 1.9
tests/net/ipsec/t_ipsec_tunnel_ipcomp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tunnel_odd.sh: 1.3
tests/net/mcast/t_mcast.sh: 1.6
tests/net/net/t_ipaddress.sh: 1.11
tests/net/net_common.sh: 1.20
tests/net/npf/t_npf.sh: 1.3
tests/net/route/t_flags.sh: 1.20
tests/net/route/t_flags6.sh: 1.16
usr.bin/netstat/fast_ipsec.c: 1.22
Do m_pullup before mtod
It may fix panicks of some tests on anita/sparc and anita/GuruPlug.
---
KNF
---
Enable DEBUG for babylon5
---
Apply C99-style struct initialization to xformsw
---
Tweak outputs of netstat -s for IPsec
- Get rid of "Fast"
- Use ipsec and ipsec6 for titles to clarify protocol
- Indent outputs of sub protocols
Original outputs were organized like this:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
New outputs are organized like this:
ipsec:
ah:
esp:
ipip:
ipcomp:
ipsec6:
ah:
esp:
ipip:
ipcomp:
---
Add test cases for IPComp
---
Simplify IPSEC_OSTAT macro (NFC)
---
KNF; replace leading whitespaces with hard tabs
---
Introduce and use SADB_SASTATE_USABLE_P
---
KNF
---
Add update command for testing
Updating an SA (SADB_UPDATE) requires that a process issuing
SADB_UPDATE is the same as a process issued SADB_ADD (or SADB_GETSPI).
This means that update command must be used with add command in a
configuration of setkey. This usage is normally meaningless but
useful for testing (and debugging) purposes.
---
Add test cases for updating SA/SP
The tests require newly-added udpate command of setkey.
---
PR/52346: Frank Kardel: Fix checksumming for NAT-T
See XXX for improvements.
---
Remove codes for PACKET_TAG_IPSEC_IN_CRYPTO_DONE
It seems that PACKET_TAG_IPSEC_IN_CRYPTO_DONE is for network adapters
that have IPsec accelerators; a driver sets the mtag to a packet
when its device has already encrypted the packet.
Unfortunately no driver implements such offload features for long
years and seems unlikely to implement them soon. (Note that neither
FreeBSD nor Linux doesn't have such drivers.) Let's remove related
(unused) codes and simplify the IPsec code.
---
Fix usages of sadb_msg_errno
---
Avoid updating sav directly
On SADB_UPDATE a target sav was updated directly, which was unsafe.
Instead allocate another sav, copy variables of the old sav to
the new one and replace the old one with the new one.
---
Simplify; we can assume sav->tdb_xform cannot be NULL while it's valid
---
Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
---
Use explicit_memset to surely zero-clear key_auth and key_enc
---
Make sure to clear keys on error paths of key_setsaval
---
Add missing KEY_FREESAV
---
Make sure a sav is inserted to a sah list after its initialization completes
---
Remove unnecessary zero-clearing codes from key_setsaval
key_setsaval is now used only for a newly-allocated sav. (It was
used to reset variables of an existing sav.)
---
Correct wrong assumption of sav->refcnt in key_delsah
A sav in a list is basically not to be sav->refcnt == 0. And also
KEY_FREESAV assumes sav->refcnt > 0.
---
Let key_getsavbyspi take a reference of a returning sav
---
Use time_mono_to_wall (NFC)
---
Separate sending message routine (NFC)
---
Simplify; remove unnecessary zero-clears
key_freesaval is used only when a target sav is being destroyed.
---
Omit NULL checks for sav->lft_c
sav->lft_c can be NULL only when initializing or destroying sav.
---
Omit unnecessary NULL checks for sav->sah
---
Omit unnecessary check of sav->state
key_allocsa_policy picks a sav of either MATURE or DYING so we
don't need to check its state again.
---
Simplify; omit unnecessary saidx passing
- ipsec_nextisr returns a saidx but no caller uses it
- key_checkrequest is passed a saidx but it can be gotton by
another argument (isr)
---
Fix splx isn't called on some error paths
---
Fix header size calculation of esp where sav is NULL
---
Fix header size calculation of ah in the case sav is NULL
This fix was also needed for esp.
---
Pass sav directly to opencrypto callback
In a callback, use a passed sav as-is by default and look up a sav
only if the passed sav is dead.
---
Avoid examining freshness of sav on packet processing
If a sav list is sorted (by lft_c->sadb_lifetime_addtime) in advance,
we don't need to examine each sav and also don't need to delete one
on the fly and send up a message. Fortunately every sav lists are sorted
as we need.
Added key_validate_savlist validates that each sav list is surely sorted
(run only if DEBUG because it's not cheap).
---
Add test cases for SAs with different SPIs
---
Prepare to stop using isr->sav
isr is a shared resource and using isr->sav as a temporal storage
for each packet processing is racy. And also having a reference from
isr to sav makes the lifetime of sav non-deterministic; such a reference
is removed when a packet is processed and isr->sav is overwritten by
new one. Let's have a sav locally for each packet processing instead of
using shared isr->sav.
However this change doesn't stop using isr->sav yet because there are
some users of isr->sav. isr->sav will be removed after the users find
a way to not use isr->sav.
---
Fix wrong argument handling
---
fix printf format.
---
Don't validate sav lists of LARVAL or DEAD states
We don't sort the lists so the validation will always fail.
Fix PR kern/52405
---
Make sure to sort the list when changing the state by key_sa_chgstate
---
Rename key_allocsa_policy to key_lookup_sa_bysaidx
---
Separate test files
---
Calculate ah_max_authsize on initialization as well as esp_max_ivlen
---
Remove m_tag_find(PACKET_TAG_IPSEC_PENDING_TDB) because nobody sets the tag
---
Restore a comment removed in previous
The comment is valid for the below code.
---
Make tests more stable
sleep command seems to wait longer than expected on anita so
use polling to wait for a state change.
---
Add tests that explicitly delete SAs instead of waiting for expirations
---
Remove invalid M_AUTHIPDGM check on ESP isr->sav
M_AUTHIPDGM flag is set to a mbuf in ah_input_cb. An sav of ESP can
have AH authentication as sav->tdb_authalgxform. However, in that
case esp_input and esp_input_cb are used to do ESP decryption and
AH authentication and M_AUTHIPDGM never be set to a mbuf. So
checking M_AUTHIPDGM of a mbuf on isr->sav of ESP is meaningless.
---
Look up sav instead of relying on unstable sp->req->sav
This code is executed only in an error path so an additional lookup
doesn't matter.
---
Correct a comment
---
Don't release sav if calling crypto_dispatch again
---
Remove extra KEY_FREESAV from ipsec_process_done
It should be done by the caller.
---
Don't bother the case of crp->crp_buf == NULL in callbacks
---
Hold a reference to an SP during opencrypto processing
An SP has a list of isr (ipsecrequest) that represents a sequence
of IPsec encryption/authentication processing. One isr corresponds
to one opencrypto processing. The lifetime of an isr follows its SP.
We pass an isr to a callback function of opencrypto to continue
to a next encryption/authentication processing. However nobody
guaranteed that the isr wasn't freed, i.e., its SP wasn't destroyed.
In order to avoid such unexpected destruction of isr, hold a reference
to its SP during opencrypto processing.
---
Don't make SAs expired on tests that delete SAs explicitly
---
Fix a debug message
---
Dedup error paths (NFC)
---
Use pool to allocate tdb_crypto
For ESP and AH, we need to allocate an extra variable space in addition
to struct tdb_crypto. The fixed size of pool items may be larger than
an actual requisite size of a buffer, but still the performance
improvement by replacing malloc with pool wins.
---
Don't use unstable isr->sav for header size calculations
We may need to optimize to not look up sav here for users that
don't need to know an exact size of headers (e.g., TCP segmemt size
caclulation).
---
Don't use sp->req->sav when handling NAT-T ESP fragmentation
In order to do this we need to look up a sav however an additional
look-up degrades performance. A sav is later looked up in
ipsec4_process_packet so delay the fragmentation check until then
to avoid an extra look-up.
---
Don't use key_lookup_sp that depends on unstable sp->req->sav
It provided a fast look-up of SP. We will provide an alternative
method in the future (after basic MP-ification finishes).
---
Stop setting isr->sav on looking up sav in key_checkrequest
---
Remove ipsecrequest#sav
---
Stop setting mtag of PACKET_TAG_IPSEC_IN_DONE because there is no users anymore
---
Skip ipsec_spi_*_*_preferred_new_timeout when running on qemu
Probably due to PR 43997
---
Add localcount to rump kernels
---
Remove unused macro
---
Fix key_getcomb_setlifetime
The fix adjusts a soft limit to be 80% of a corresponding hard limit.
I'm not sure the fix is really correct though, at least the original
code is wrong. A passed comb is zero-cleared before calling
key_getcomb_setlifetime, so
comb->sadb_comb_soft_addtime = comb->sadb_comb_soft_addtime * 80 / 100;
is meaningless.
---
Provide and apply key_sp_refcnt (NFC)
It simplifies further changes.
---
Fix indentation
Pointed out by knakahara@
---
Use pslist(9) for sptree
---
Don't acquire global locks for IPsec if NET_MPSAFE
Note that the change is just to make testing easy and IPsec isn't MP-safe yet.
---
Let PF_KEY socks hold their own lock instead of softnet_lock
Operations on SAD and SPD are executed via PF_KEY socks. The operations
include deletions of SAs and SPs that will use synchronization mechanisms
such as pserialize_perform to wait for references to SAs and SPs to be
released. It is known that using such mechanisms with holding softnet_lock
causes a dead lock. We should avoid the situation.
---
Make IPsec SPD MP-safe
We use localcount(9), not psref(9), to make the sptree and secpolicy (SP)
entries MP-safe because SPs need to be referenced over opencrypto
processing that executes a callback in a different context.
SPs on sockets aren't managed by the sptree and can be destroyed in softint.
localcount_drain cannot be used in softint so we delay the destruction of
such SPs to a thread context. To do so, a list to manage such SPs is added
(key_socksplist) and key_timehandler_spd deletes dead SPs in the list.
For more details please read the locking notes in key.c.
Proposed on tech-kern@ and tech-net@
---
Fix updating ipsec_used
- key_update_used wasn't called in key_api_spddelete2 and key_api_spdflush
- key_update_used wasn't called if an SP had been added/deleted but
a reply to userland failed
---
Fix updating ipsec_used; turn on when SPs on sockets are added
---
Add missing IPsec policy checks to icmp6_rip6_input
icmp6_rip6_input is quite similar to rip6_input and the same checks exist
in rip6_input.
---
Add test cases for setsockopt(IP_IPSEC_POLICY)
---
Don't use KEY_NEWSP for dummy SP entries
By the change KEY_NEWSP is now not called from softint anymore
and we can use kmem_zalloc with KM_SLEEP for KEY_NEWSP.
---
Comment out unused functions
---
Add test cases that there are SPs but no relevant SAs
---
Don't allow sav->lft_c to be NULL
lft_c of an sav that was created by SADB_GETSPI could be NULL.
---
Clean up clunky eval strings
- Remove unnecessary \ at EOL
- This allows to omit ; too
- Remove unnecessary quotes for arguments of atf_set
- Don't expand $DEBUG in eval
- We expect it's expanded on execution
Suggested by kre@
---
Remove unnecessary KEY_FREESAV in an error path
sav should be freed (unreferenced) by the caller.
---
Use pslist(9) for sahtree
---
Use pslist(9) for sah->savtree
---
Rename local variable newsah to sah
It may not be new.
---
MP-ify SAD slightly
- Introduce key_sa_mtx and use it for some list operations
- Use pserialize for some list iterations
---
Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
---
Remove out-of-date log output
Pointed out by riastradh@
---
Use KDASSERT instead of KASSERT for mutex_ownable
Because mutex_ownable is too heavy to run in a fast path
even for DIAGNOSTIC + LOCKDEBUG.
Suggested by riastradh@
---
Assemble global lists and related locks into cache lines (NFCI)
Also rename variable names from *tree to *list because they are
just lists, not trees.
Suggested by riastradh@
---
Move locking notes
---
Update the locking notes
- Add locking order
- Add locking notes for misc lists such as reglist
- Mention pserialize, key_sp_ref and key_sp_unref on SP operations
Requested by riastradh@
---
Describe constraints of key_sp_ref and key_sp_unref
Requested by riastradh@
---
Hold key_sad.lock on SAVLIST_WRITER_INSERT_TAIL
---
Add __read_mostly to key_psz
Suggested by riastradh@
---
Tweak wording (pserialize critical section => pserialize read section)
Suggested by riastradh@
---
Add missing mutex_exit
---
Fix setkey -D -P outputs
The outputs were tweaked (by me), but I forgot updating libipsec
in my local ATF environment...
---
MP-ify SAD (key_sad.sahlist and sah entries)
localcount(9) is used to protect key_sad.sahlist and sah entries
as well as SPD (and will be used for SAD sav).
Please read the locking notes of SAD for more details.
---
Introduce key_sa_refcnt and replace sav->refcnt with it (NFC)
---
Destroy sav only in the loop for DEAD sav
---
Fix KASSERT(solocked(sb->sb_so)) failure in sbappendaddr that is called eventually from key_sendup_mbuf
If key_sendup_mbuf isn't passed a socket, the assertion fails.
Originally in this case sb->sb_so was softnet_lock and callers
held softnet_lock so the assertion was magically satisfied.
Now sb->sb_so is key_so_mtx and also softnet_lock isn't always
held by callers so the assertion can fail.
Fix it by holding key_so_mtx if key_sendup_mbuf isn't passed a socket.
Reported by knakahara@
Tested by knakahara@ and ozaki-r@
---
Fix locking notes of SAD
---
Fix deadlock between key_sendup_mbuf called from key_acquire and localcount_drain
If we call key_sendup_mbuf from key_acquire that is called on packet
processing, a deadlock can happen like this:
- At key_acquire, a reference to an SP (and an SA) is held
- key_sendup_mbuf will try to take key_so_mtx
- Some other thread may try to localcount_drain to the SP with
holding key_so_mtx in say key_api_spdflush
- In this case localcount_drain never return because key_sendup_mbuf
that has stuck on key_so_mtx never release a reference to the SP
Fix the deadlock by deferring key_sendup_mbuf to the timer
(key_timehandler).
---
Fix that prev isn't cleared on retry
---
Limit the number of mbufs queued for deferred key_sendup_mbuf
It's easy to be queued hundreds of mbufs on the list under heavy
network load.
---
MP-ify SAD (savlist)
localcount(9) is used to protect savlist of sah. The basic design is
similar to MP-ifications of SPD and SAD sahlist. Please read the
locking notes of SAD for more details.
---
Simplify ipsec_reinject_ipstack (NFC)
---
Add per-CPU rtcache to ipsec_reinject_ipstack
It reduces route lookups and also reduces rtcache lock contentions
when NET_MPSAFE is enabled.
---
Use pool_cache(9) instead of pool(9) for tdb_crypto objects
The change improves network throughput especially on multi-core systems.
---
Update
ipsec(4), opencrypto(9) and vlan(4) are now MP-safe.
---
Write known issues on scalability
---
Share a global dummy SP between PCBs
It's never be changed so it can be pre-allocated and shared safely between PCBs.
---
Fix race condition on the rawcb list shared by rtsock and keysock
keysock now protects itself by its own mutex, which means that
the rawcb list is protected by two different mutexes (keysock's one
and softnet_lock for rtsock), of course it's useless.
Fix the situation by having a discrete rawcb list for each.
---
Use a dedicated mutex for rt_rawcb instead of softnet_lock if NET_MPSAFE
---
fix localcount leak in sav. fixed by ozaki-r@n.o.
I commit on behalf of him.
---
remove unnecessary comment.
---
Fix deadlock between pserialize_perform and localcount_drain
A typical ussage of localcount_drain looks like this:
mutex_enter(&mtx);
item = remove_from_list();
pserialize_perform(psz);
localcount_drain(&item->localcount, &cv, &mtx);
mutex_exit(&mtx);
This sequence can cause a deadlock which happens for example on the following
situation:
- Thread A calls localcount_drain which calls xc_broadcast after releasing
a specified mutex
- Thread B enters the sequence and calls pserialize_perform with holding
the mutex while pserialize_perform also calls xc_broadcast
- Thread C (xc_thread) that calls an xcall callback of localcount_drain tries
to hold the mutex
xc_broadcast of thread B doesn't start until xc_broadcast of thread A
finishes, which is a feature of xcall(9). This means that pserialize_perform
never complete until xc_broadcast of thread A finishes. On the other hand,
thread C that is a callee of xc_broadcast of thread A sticks on the mutex.
Finally the threads block each other (A blocks B, B blocks C and C blocks A).
A possible fix is to serialize executions of the above sequence by another
mutex, but adding another mutex makes the code complex, so fix the deadlock
by another way; the fix is to release the mutex before pserialize_perform
and instead use a condvar to prevent pserialize_perform from being called
simultaneously.
Note that the deadlock has happened only if NET_MPSAFE is enabled.
---
Add missing ifdef NET_MPSAFE
---
Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
---
- sanitize key debugging so that we don't print extra newlines or unassociated
debugging messages.
- remove unused functions and make internal ones static
- print information in one line per message
---
humanize printing of ip addresses
---
cast reduction, NFC.
---
Fix typo in comment
---
Pull out ipsec_fill_saidx_bymbuf (NFC)
---
Don't abuse key_checkrequest just for looking up sav
It does more than expected for example key_acquire.
---
Fix SP is broken on transport mode
isr->saidx was modified accidentally in ipsec_nextisr.
Reported by christos@
Helped investigations by christos@ and knakahara@
---
Constify isr at many places (NFC)
---
Include socketvar.h for softnet_lock
---
Fix buffer length for ipsec_logsastr
|
| 1.376.2.7 | 18-Jan-2019 |
pgoyette | Synch with HEAD
|
| 1.376.2.6 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.376.2.5 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.376.2.4 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.376.2.3 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.376.2.2 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.376.2.1 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.384.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.384.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.389.2.3 | 07-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1226):
sys/netinet6/ip6_id.c: revision 1.19-1.21
sys/netinet6/ip6_var.h: revision 1.88
sys/netinet/ip_input.c: revision 1.400
sys/netinet/tcp_subr.c: revision 1.285
sys/netinet/ip6.h: revision 1.30
netinet: Enable random IP fragment ids by default (from riastradh)
netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
netinet6: Mark randomid unused.
Will make merging and bisection easier if anything goes wrong with
flow label or fragment id randomization changes.
(from riastradh)
netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
Replace randomid() by cprng_fast32()
|
| 1.389.2.2 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #238):
sys/netipsec/ipsec_output.c: revision 1.83
sys/net/route.h: revision 1.125
sys/netinet6/ip6_input.c: revision 1.210
sys/netinet6/ip6_input.c: revision 1.211
sys/net/if.c: revision 1.461
sys/net/if_gif.h: revision 1.33
sys/net/route.c: revision 1.220
sys/net/route.c: revision 1.221
sys/net/if.h: revision 1.277
sys/netinet6/ip6_forward.c: revision 1.97
sys/netinet/wqinput.c: revision 1.6
sys/net/if_ipsec.h: revision 1.5
sys/netinet6/in6_l2tp.c: revision 1.18
sys/netinet6/in6_gif.c: revision 1.94
sys/net/if_l2tp.h: revision 1.7
sys/net/if_gif.c: revision 1.149
sys/net/if_l2tp.h: revision 1.8
sys/netinet/in_gif.c: revision 1.95
sys/netinet/in_l2tp.c: revision 1.17
sys/netipsec/ipsecif.c: revision 1.17
sys/net/if_ipsec.c: revision 1.24
sys/net/if_l2tp.c: revision 1.37
sys/netinet/ip_input.c: revision 1.391
sys/net/if_l2tp.c: revision 1.38
sys/netinet/ip_input.c: revision 1.392
sys/net/if_l2tp.c: revision 1.39
Avoid having a rtcache directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
wqinput: avoid having struct wqinput_worklist directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Input handlers of wqinput normally involves sleepable operations so we must
avoid dereferencing a percpu data (struct wqinput_worklist) after executing
an input handler. Address this situation by having just a pointer to the data
in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
Add missing #include <sys/kmem.h>
-
Divide Tx context of l2tp(4) to improve performance.
It seems l2tp(4) call path is too long for instruction cache. So, dividing
l2tp(4) Tx context improves CPU use efficiency.
After this commit, l2tp(4) throughput gains 10% on my machine(Atom C3000).
-
Apply some missing changes lost on the previous commit
-
Avoid having a rtcache directly in a percpu storage for tunnel protocols.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
-
l2tp(4): avoid having struct ifqueue directly in a percpu storage.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Tx processing of l2tp(4) uses normally involves sleepable operations so we
must avoid dereferencing a percpu data (struct ifqueue) after executing Tx
processing. Address this situation by having just a pointer to the data in
a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
|
| 1.389.2.1 | 17-Sep-2019 |
martin | Pull up following revision(s) (requested by bouyer in ticket #208):
sys/netinet6/ip6_input.c: revision 1.209
sys/netinet/ip_input.c: revision 1.390
Packet filters can return an mbuf chain with fragmented headers, so
m_pullup() it if needed and remove the KASSERT()s.
|
| 1.397.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.402.4.2 | 29-Jul-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1140):
sys/netinet/ip_output.c: revision 1.330
sys/netinet/sctp_output.c: revision 1.39
sys/netinet/ip_mroute.c: revision 1.166
sys/netipsec/ipsecif.c: revision 1.24
sys/netipsec/xform_ipip.c: revision 1.80
sys/netinet/ip_output.c: revision 1.327
sys/netinet/ip_output.c: revision 1.328
sys/netinet/ip_input.c: revision 1.406
sys/netinet/ip_output.c: revision 1.329
sys/netinet/in_var.h: revision 1.105
in: get rid of unused argument from ip_newid() and ip_newid_range()
in: take a reference of ifp on IP_ROUTETOIF
The ifp could be released after ia4_release(ia).
in: narrow the scope of ifa in ip_output (NFC)
sctp: follow the recent change of ip_newid()
in: avoid racy ifa_acquire(rt->rt_ifa) in ip_output()
If a rtentry is being destroyed asynchronously, ifa referenced by rt_ifa
can be destructed and taking ifa_acquire(rt->rt_ifa) aborts with a
KASSERT failure. Fortunately, the ifa is not actually freed because of
a reference by rt_ifa, it can be available (except some functions like
psref) so as long the rtentry is held.
PR kern/59527
in: avoid racy ia4_acquire(ifatoia(rt->rt_ifa) in ip_rtaddr()
Same as the case of ip_output(), it's racy and should be avoided.
PR kern/59527
|
| 1.402.4.1 | 14-Jul-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1137):
sys/netinet/ip_input.c: revision 1.405
in: avoid packet looping on incoming packets destining to an initializing
address
The initialization of an IPv4 address is done by adding a connected route and
a local route (if necessary), and then publishing itself by adding it to the
global list (and the global hashtable). Thus, there can exist a route with an
address that is not published. This inconsistent state allows an incoming
packet destining to one of a host address which is not published but has a
local route to be forwarded and routed to a loopback interface. This results
in forwarding the packet back to ip_input, that is, packet looping.
To avoid the situation, prohibit packets being forwarded via a local route.
This is a workaround for "IPv4 address initialization atomicity" in
doc/TODO.smpnet.
|
| 1.403.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.17 | 04-Mar-2002 |
sommerfeld | The "gif*" tunnelling interface does everything ipip does.
Move usage example from ipip.4 to gif.4
Excise ipip and stitch up the scars.
|
| 1.16 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.15 | 13-Apr-2001 |
thorpej | branches: 1.15.2;
Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.14 | 17-Jan-2001 |
thorpej | branches: 1.14.2;
Fix a rather annoying problem where the sockaddr_dl which holds
the link level name for the interface (ifp->if_sadl) is allocated
before ifp->if_addrlen is initialized, which could lead to allocating
too little space for the link level address.
Do this by splitting allocation of the link level name out of
if_attach() and into if_alloc_sadl(), which is normally called
by functions like ether_ifattach(). Network interfaces which
don't have a link-specific attach routine must call if_alloc_sadl()
themselves (example: gif).
Link level names are freed by if_free_sadl(), which can be called
from e.g. ether_ifdetach(). Drivers never need call if_free_sadl()
themselves as if_detach() will do it if it is not already done.
While here, add the ability to pass an AF_LINK address to
SIOCSIFADDR in ether_ioctl() (this is what caused me to notice
the problem that the above fixes).
|
| 1.13 | 18-Dec-2000 |
thorpej | Pull in BPF includes.
|
| 1.12 | 18-Dec-2000 |
thorpej | Fill in if_dlt.
|
| 1.11 | 12-Dec-2000 |
thorpej | Adapt to bpfattach() changes, and further centralize the bpfattach()
and bpfdetach() calls into link-type subroutines where possible.
|
| 1.10 | 19-Apr-2000 |
itojun | introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.9 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.8 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.7 | 26-Aug-1999 |
itojun | branches: 1.7.2; 1.7.8;
clear m->m_pkthdr.rcvif before calling ip_output().
the member is used to pass struct socket to ip{,6}_output for ipsec decisions.
(i agree it is kind of ugly. we need to modify struct mbuf if we are
to do better - which seems to me a bit too much)
|
| 1.6 | 04-Apr-1999 |
tron | - Make sure that interface can't be marked up before a route to the remote
tunnel end was found.
- Flush route and reset MTU if interface is marked down.
|
| 1.5 | 04-Apr-1999 |
tron | Avoid kernel panic if interface is configured before a route to the
remote of the tunnel can be found.
XXX If you manually mark the interface as "UP" and set the MTU later
XXX sending a packet will still cause a kernel panic.
|
| 1.4 | 02-Apr-1999 |
hwr | Setting of source and destination IP is not done by
passing SIOCSIFADDR/SIOIFDSTADDR, but by passing the addresses in
the appropriate structs.
One of the mysteries of ifconfig IMHO...
Should fix kern/6899.
|
| 1.3 | 02-Feb-1999 |
thorpej | branches: 1.3.2;
Set the tunnel destination address correctly. Should fix PR #6899.
|
| 1.2 | 13-Jan-1999 |
thorpej | Use the count supplied to the pseudo-device attach routine to dynamically
allocate (once) the ipip_softc array; don't assume NIPIP contains the count.
|
| 1.1 | 11-Jan-1999 |
thorpej | Separate out the IP-in-IP implementation from the GRE code. This cleans
up the interface to ip_mroute.c somewhat, and properly separates IP-IP
from GRE. (They are similar, but they are different protocols, and should
not be implemented in the same place.)
|
| 1.3.2.1 | 04-Apr-1999 |
tron | branches: 1.3.2.1.2; 1.3.2.1.4;
Pull up revisions 1.4, 1.5 and 1.6 from trunk.
|
| 1.3.2.1.4.1 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.3.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.7.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.7.2.5 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.7.2.4 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.7.2.3 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.7.2.2 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.7.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.14.2.3 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.14.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.14.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.15.2.2 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.15.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.3 | 04-Mar-2002 |
sommerfeld | The "gif*" tunnelling interface does everything ipip does.
Move usage example from ipip.4 to gif.4
Excise ipip and stitch up the scars.
|
| 1.2 | 19-Apr-2000 |
itojun | branches: 1.2.6; 1.2.8;
introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.1 | 11-Jan-1999 |
thorpej | branches: 1.1.8;
Separate out the IP-in-IP implementation from the GRE code. This cleans
up the interface to ip_mroute.c somewhat, and properly separates IP-IP
from GRE. (They are similar, but they are different protocols, and should
not be implemented in the same place.)
|
| 1.1.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.2.8.1 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.2.6.1 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.5 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.4 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.3 | 28-Mar-2004 |
martti | branches: 1.3.2;
Upgraded IPFilter to 4.1.1
|
| 1.2 | 01-Apr-2002 |
jdolecek | branches: 1.2.10;
add __KERNEL_RCSID()
|
| 1.1 | 24-Jan-2002 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.2 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.1 | 24-Jan-2002 |
martti | branches: 1.1.1.1.2; 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 3.4.23
|
| 1.1.1.1.6.3 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.1.1.1.6.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.1.1.1.6.1 | 24-Jan-2002 |
nathanw | file ip_ipsec_pxy.c was added on branch nathanw_sa on 2002-02-28 04:15:08 +0000
|
| 1.1.1.1.4.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.1.1.1.4.3 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.1.1.1.4.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.1.1.1.4.1 | 24-Jan-2002 |
jdolecek | file ip_ipsec_pxy.c was added on branch kqueue on 2002-02-11 20:10:35 +0000
|
| 1.1.1.1.2.4 | 27-Nov-2002 |
itojun | sys/netinet/ip_h323_pxy.c via patch
sys/netinet/ip_ipsec_pxy.c via patch
sys/netinet/ip_netbios_pxy.c via patch
Fix compilation on a.out systems.
(Thorsten Frueauf)
|
| 1.1.1.1.2.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.1.1.1.2.2 | 09-Feb-2002 |
he | Pull up revision 1.1 (new, requested by martti):
Updated IPFilter to 3.4.23
|
| 1.1.1.1.2.1 | 24-Jan-2002 |
he | file ip_ipsec_pxy.c was added on branch netbsd-1-5 on 2002-02-09 16:55:22 +0000
|
| 1.2.10.3 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.2.10.2 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.2.10.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.3.2.1 | 13-Aug-2004 |
jmc | branches: 1.3.2.1.2;
Pullup rev 1.4 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.3.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.5 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 23-Jul-2004 |
martti | branches: 1.1.1.2.2;
Import IPFilter 4.1.3
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.2;
Import IPFilter 4.1.1
|
| 1.1.1.2.2.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.2.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.2.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.2.2.3 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.1.1.2.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.2.2.1 | 23-Jul-2004 |
skrll | file ip_irc_pxy.c was added on branch ktrace-lwp on 2004-08-03 10:54:40 +0000
|
| 1.1.1.1.2.1 | 13-Aug-2004 |
jmc | branches: 1.1.1.1.2.1.2;
Pullup rev 1.1.1.2 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.1.1.1.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.26 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.25 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.24 | 28-Mar-2004 |
martti | branches: 1.24.2;
Upgraded IPFilter to 4.1.1
|
| 1.23 | 25-Sep-2002 |
martti | branches: 1.23.6;
Fix ipmon problems on 64-bit platforms (PR#17403 and PR#17404).
|
| 1.22 | 19-Sep-2002 |
martti | Resync with official IPF
|
| 1.21 | 01-Jul-2002 |
christos | Fix iplog problem on sparc64 [from Tomi Nylund]
1. size_t is 64 bits, so use a u_32_t for iplused
2. microtime() and friends expect a struct timeval,
passing the first of two unsigned longs will not cut it.
|
| 1.20 | 09-Jun-2002 |
itojun | whitespace
|
| 1.19 | 02-May-2002 |
martti | branches: 1.19.2; 1.19.4;
Upgraded IPFilter to 3.4.27
|
| 1.18 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.17 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.16 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.15 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.14 | 28-Sep-2001 |
chs | don't depend on other headers to include sys/proc.h for us.
|
| 1.13 | 26-Mar-2001 |
mike | branches: 1.13.2; 1.13.4;
Resolve conflicts.
|
| 1.12 | 05-Feb-2001 |
chs | branches: 1.12.2;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.11 | 09-Aug-2000 |
veego | Resolve conflicts.
|
| 1.10 | 03-May-2000 |
veego | branches: 1.10.4;
Resolve conflicts.
|
| 1.9 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.8 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.7 | 10-Dec-1998 |
christos | branches: 1.7.2; 1.7.8; 1.7.14;
defopt IPFILTER_LOG
|
| 1.6 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.5 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.4 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.3 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.2 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.1 | 21-Sep-1997 |
veego | branches: 1.1.1;
Initial revision
|
| 1.1.1.15 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.14 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.13 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.12 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.11 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.10 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.9 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.8 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.7 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.6 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.5 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.4 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.3 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.2 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.1 | 21-Sep-1997 |
veego | branches: 1.1.1.1.2; 1.1.1.1.4;
Import ip-filter 3.2beta5
|
| 1.1.1.1.4.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.6 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.1.1.1.4.3 | 22-Jul-1998 |
mellon | Pull up 1.5 (veego)
|
| 1.1.1.1.4.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.1.1.1.4.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.1.1.1.2.2 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.1.1.1.2.1 | 21-Sep-1997 |
thorpej | file ip_log.c was added on branch marc-pcmcia on 1997-09-22 06:34:10 +0000
|
| 1.7.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.7.8.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.7.8.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.7.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.7.2.1 | 20-Dec-1999 |
he | Pull up revision 1.8 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.10.4.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.10.4.2 | 09-Feb-2002 |
he | Pull up revisions 1.12-1.17 (requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.10.4.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.12.2.10 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.9 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.12.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.6 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.12.2.5 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.12.2.4 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.12.2.2 | 08-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.12.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.13.4.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.13.2.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.13.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.13.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.13.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.13.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.19.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.19.2.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.19.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.23.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.23.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.24.2.1 | 13-Aug-2004 |
jmc | branches: 1.24.2.1.2;
Pullup rev 1.25 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.24.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.26 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.4 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.3 | 23-Jul-2004 |
martti | branches: 1.3.2;
Upgraded IPFilter to 4.1.3
|
| 1.2 | 01-Apr-2004 |
martin | A few more ioctl vs. copyin changes, spotted by Bill Studenmund.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.2;
Import IPFilter 4.1.1
|
| 1.1.1.1.2.2 | 13-Aug-2004 |
jmc | branches: 1.1.1.1.2.2.2;
Pullup rev 1.3 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.1.1.1.2.1 | 02-Apr-2004 |
tron | Pull up revision 1.2 (requested by martin in ticket #46):
A few more ioctl vs. copyin changes, spotted by Bill Studenmund.
|
| 1.1.1.1.2.2.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.4 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.3.2.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.3.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.3.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.3.2.3 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.3.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.3.2.1 | 23-Jul-2004 |
skrll | file ip_lookup.c was added on branch ktrace-lwp on 2004-08-03 10:54:40 +0000
|
| 1.2 | 05-Oct-2004 |
yamt | move netinet/ip_lookup.h -> dist/ipf/netinet/ip_lookup.h.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_lookup.h was added on branch ktrace-lwp on 2004-08-03 10:54:40 +0000
|
| 1.166 | 11-Jun-2025 |
ozaki-r | in: get rid of unused argument from ip_newid() and ip_newid_range()
|
| 1.165 | 15-Mar-2022 |
andvar | branches: 1.165.4; 1.165.10;
s/heaader/header/
|
| 1.164 | 12-Nov-2020 |
kardel | PR kern/55779:
restore non-desctructive guarantee of ip_mforward() mbuf
argument. This avoids generation invalid UDP checksums
on multicast packets in ip_output().
XXX the root cause of the misguided fix in 2008 should be
XXX investigated
|
| 1.163 | 14-Sep-2018 |
maxv | branches: 1.163.4; 1.163.12;
Use non-variadic function pointer in protosw::pr_input.
|
| 1.162 | 11-Jul-2018 |
martin | Add missing <netinet/in_offload.h> include.
|
| 1.161 | 11-Jul-2018 |
maxv | Rename
ip_undefer_csum -> in_undefer_cksum
in_delayed_cksum -> in_undefer_cksum_tcpudp
The two previous names were inconsistent and misleading.
Put the two functions into in_offload.c. Add comments to explain what
we're doing.
The same could be done for IPv6.
|
| 1.160 | 21-Jun-2018 |
knakahara | branches: 1.160.2;
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.159 | 14-May-2018 |
maxv | Don't crash if there is no inner IP header.
|
| 1.158 | 07-May-2018 |
maxv | Fix possible buffer overflow. We need to make sure the inner IPv4 packet
doesn't have options, because we validate only an option-less header.
|
| 1.157 | 11-Apr-2018 |
maxv | Add XXX.
|
| 1.156 | 11-Apr-2018 |
maxv | Add XXX.
|
| 1.155 | 21-Mar-2018 |
roy | Sprinkle more soroverflow().
|
| 1.154 | 09-Feb-2018 |
maxv | branches: 1.154.2;
Style (realign everything correctly), and fix a typo.
|
| 1.153 | 07-Feb-2018 |
maxv | Style and constify.
|
| 1.152 | 07-Feb-2018 |
maxv | More style. No functional change.
|
| 1.151 | 07-Feb-2018 |
maxv | Remove parentheses in return statements. No functional change.
|
| 1.150 | 07-Feb-2018 |
maxv | Style and remove unused macros. More to come.
|
| 1.149 | 07-Feb-2018 |
maxv | Remove RSVP_ISI, that's mostly dead code. FreeBSD and OpenBSD too removed
it; FreeBSD kept some pieces but they are mostly no-opts.
Sent on tech-net@, no comment.
|
| 1.148 | 15-Nov-2017 |
knakahara | Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.147 | 23-Jul-2017 |
para | kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
|
| 1.146 | 24-Jan-2017 |
ozaki-r | branches: 1.146.6;
Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.145 | 11-Jan-2017 |
ozaki-r | branches: 1.145.2;
Get rid of unnecessary header inclusions
|
| 1.144 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.143 | 04-Jul-2016 |
knakahara | branches: 1.143.2;
make encap_lock_{enter,exit} interruptable.
|
| 1.142 | 04-Jul-2016 |
knakahara | let gif(4) promise softint(9) contract (2/2) : ip_encap side
The last commit does not care encaptab. This commit fixes encaptab race which
is used not only gif(4).
|
| 1.141 | 21-Jun-2016 |
ozaki-r | Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.140 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.139 | 26-Apr-2016 |
ozaki-r | Sweep unnecessary route.h inclusions
|
| 1.138 | 26-Jan-2016 |
knakahara | eliminate variable argument in encapsw
|
| 1.137 | 26-Jan-2016 |
knakahara | implement encapsw instead of protosw and uniform prototype.
suggested and advised by riastradh@n.o, thanks.
BTW, It seems in_stf_input() had bugs...
|
| 1.136 | 22-Jan-2016 |
riastradh | Back out previous change to introduce struct encapsw.
This change was intended, but Nakahara-san had already made a better
one locally! So I'll let him commit that one, and I'll try not to
step on anyone's toes again.
|
| 1.135 | 22-Jan-2016 |
riastradh | Don't abuse struct protosw for ip_encap -- introduce struct encapsw.
Mostly mechanical change to replace it, culling some now-needless
boilerplate around all the users.
This does not substantively change the ip_encap API or eliminate
abuse of sketchy pointer casts -- that will come later, and will be
easier now that it is not tangled up with struct protosw.
|
| 1.134 | 20-Jan-2016 |
riastradh | Eliminate struct protosw::pr_output.
You can't use this unless you know what it is a priori: the formal
prototype is variadic, and the different instances (e.g., ip_output,
route_output) have different real prototypes.
Convert the only user of it, raw_send in net/raw_cb.c, to take an
explicit callback argument. Convert the only instances of it,
route_output and key_output, to such explicit callbacks for raw_send.
Use assertions to make sure the conversion to explicit callbacks is
warranted.
Discussed on tech-net with no objections:
https://mail-index.netbsd.org/tech-net/2016/01/16/msg005484.html
|
| 1.133 | 17-Jan-2016 |
christos | PR/50670: David Binderman: Tidy up debugging printfs to avoid if else confusion.
|
| 1.132 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.131 | 18-Oct-2014 |
snj | branches: 1.131.2;
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.130 | 05-Jun-2014 |
rmind | - Implement pktqueue interface for lockless IP input queue.
- Replace ipintrq and ip6intrq with the pktqueue mechanism.
- Eliminate kernel-lock from ipintr() and ip6intr().
- Some preparation work to push softnet_lock out of ipintr().
Discussed on tech-net.
|
| 1.129 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.128 | 14-Sep-2013 |
martin | branches: 1.128.2;
Remove unused variable
|
| 1.127 | 05-Jun-2013 |
christos | branches: 1.127.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.126 | 24-Sep-2012 |
msaitoh | Add missing "\n" in log(9)
|
| 1.125 | 01-May-2012 |
martin | branches: 1.125.2;
Explicitly include <sys/kmem.h>
|
| 1.124 | 30-Apr-2012 |
rmind | - Replace some malloc(9) uses with kmem(9).
- G/C M_IPMOPTS, M_IPMADDR and M_BWMETER.
|
| 1.123 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.122 | 19-Dec-2011 |
drochner | branches: 1.122.2;
rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.121 | 19-Oct-2011 |
dyoung | branches: 1.121.2; 1.121.6;
Use if_addr_init() and if_mcast_op() instead of ifp->if_ioctl().
|
| 1.120 | 31-Aug-2011 |
plunky | NULL does not need a cast
|
| 1.119 | 17-Jul-2011 |
joerg | Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.118 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.117 | 19-Dec-2008 |
cegger | branches: 1.117.2;
use M_ZERO on malloc() and remove subsequent bzero().
|
| 1.116 | 01-Oct-2008 |
rmind | branches: 1.116.2;
PR/39664: Dave Huang: ip_mrouter_done: free hash using hashdone(9).
|
| 1.115 | 06-Aug-2008 |
plunky | Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.114 | 22-May-2008 |
dyoung | branches: 1.114.4;
Don't cast to void * unnecessarily.
|
| 1.113 | 08-May-2008 |
taca | Make sure to clear csum_flags before forward the packet.
This change should be fix DIAGNOSTIC kernel's panic when the machine act
as multicast router.
Advised from tls@ and approved by thorpej@.
|
| 1.112 | 05-May-2008 |
ad | branches: 1.112.2;
- Convert hashinit() to use kmem_alloc(). The hash tables can be large
and it's better to not have them in kmem_map.
- Convert a couple of minor items along the way to kmem_alloc().
- Fix some memory leaks.
|
| 1.111 | 06-Feb-2008 |
matt | branches: 1.111.6; 1.111.8; 1.111.10;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.110 | 21-Dec-2007 |
matt | Add fix for ip_id information leakage. Since the leakage information is
primarily used with TCP SYN and RST packets and such packets are less than
the smallest sized packet that an IP stack is allowed to fragment, we simply
set ip_id to 0 for all packets 68 bytes or less.
|
| 1.109 | 27-Nov-2007 |
christos | branches: 1.109.2; 1.109.6;
require that the options argument is the right size, not that it is greater
or equal to the requested size. Suggested by Matt Thomas.
|
| 1.108 | 02-Sep-2007 |
dyoung | branches: 1.108.6;
m_copym(..., 0, M_COPYALL, ...) -> m_copypacket(..., ...).
|
| 1.107 | 02-Sep-2007 |
dyoung | m_copy() was deprecated, apparently, long ago. m_copy(...) ->
m_copym(..., M_DONTWAIT).
|
| 1.106 | 31-Aug-2007 |
dyoung | Fix bug in last: add missing ampersand.
|
| 1.105 | 31-Aug-2007 |
dyoung | Stop sharing a sockaddr_in template among multicast routines,
because that's just going to cause problems down the road. (Suppose
we can have two CPUs in the network stack someday?) Instead, use
sockaddr_in_init() to initialize a sockaddr_in on the stack.
Use ifreq_setaddr() to initialize ifreq.ifr_addr.
|
| 1.104 | 09-Jul-2007 |
ad | branches: 1.104.2; 1.104.6; 1.104.8;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.103 | 13-Jun-2007 |
christos | PR/36484: Pavlin Radoslavov: PIM Register in-kernel encapsulation IP_DF
setting is incorrect
|
| 1.102 | 25-Apr-2007 |
dyoung | Get rid of some gratuitous casts and join some lines.
|
| 1.101 | 04-Mar-2007 |
christos | branches: 1.101.2; 1.101.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.100 | 16-Nov-2006 |
christos | branches: 1.100.2; 1.100.4; 1.100.8;
__unused removal on arguments; approved by core.
|
| 1.99 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.98 | 30-Aug-2006 |
christos | branches: 1.98.2; 1.98.4;
Fix initializers.
|
| 1.97 | 25-Apr-2006 |
liamjfoy | - use MAXTTL
ok christos@
|
| 1.96 | 11-Dec-2005 |
christos | branches: 1.96.4; 1.96.6; 1.96.8; 1.96.10; 1.96.12;
merge ktrace-lwp.
|
| 1.95 | 03-Aug-2005 |
gdt | Restore to working order; this has apparently been nonworking since
the decapsulator dispatch changes in 2001. Problems found and fixed
by Christine Jones of BBN. Specifically:
Check for a packet's protocol to be ENCAP_PROTO, not AF_INET.
Remove one-back cache for last vif, because vif_encapcheck is called
for each vif, rather than being expected to find the appropriate vif.
The cache usage caused packets to be input on the wrong vif and hence
usually dropped.
In vif_encapcheck, verify the local source as well. While mrouted
endeavors not to create multiple tunnels with a peer, a packet
arriving with the wrong local address is still wrong and should not be
accepted. (This is a correctness nit, not a security issue.) Order
checks to fail quickly for packets being checked to see if they match
a vif other than the one they belong on (essentially, check peer
source address in outer header first).
Claim 69 bits of match (32 each from outer src/dst and 5 from checking
that inner dst is within 224/5). This should result in the vif having
a higher priority for multicast packets compared to a parallel gif(4)
tunnel, and that both seems appropriate if both are configured and
seems to match the semantics expected by the decapsulator dispatch
machinery.
(These changes were made in 2.99.15 and about a dozen nodes are
running them with many vifs. ip_mroute.c has not changed
significantly since then (February 2005) and the changes applied
cleanly to current and compile cleanly.)
|
| 1.94 | 06-Jun-2005 |
martin | branches: 1.94.2;
Since we decided "const struct mbuf *" would not do the right thing (tm),
remove ~all const from mbuf pointers.
|
| 1.93 | 06-Jun-2005 |
martin | Adapt to ip_encap.h constification.
|
| 1.92 | 06-Jun-2005 |
christos | make this compile again.
|
| 1.91 | 29-May-2005 |
christos | change casts to proper unconst. mark XXXUNCONST
|
| 1.90 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.89 | 03-Feb-2005 |
perry | ANSIify function declarations
|
| 1.88 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.87 | 15-Jan-2005 |
manu | branches: 1.87.2; 1.87.4;
Duplicate nested if statement in PIM code (from the OpenBSD tree)
|
| 1.86 | 04-Sep-2004 |
manu | IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.85 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.84 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.83 | 21-Apr-2004 |
itojun | kill sprintf, use snprintf
|
| 1.82 | 19-Nov-2003 |
jonathan | Patch back support for (badly) randomized IP ids, by request:
* Include "opt_inet.h" everywhere IP-ids are generated with ip_newid(),
so the RANDOM_IP_ID option is visible. Also in ip_id(), to ensure
the prototype for ip_randomid() is made visible.
* Add new sysctl to enable randomized IP-ids, provided the kernel was
configured with RANDOM_IP_ID. (The sysctl defaults to zero, and is
a read-only zero if RANDOM_IP_ID is not configured).
Note that the implementation of randomized IP ids is still defective,
and should not be enabled at all (even if configured) without
very careful deliberation. Caveat emptor.
|
| 1.81 | 17-Nov-2003 |
jonathan | Revert the (default) ip_id algorithm to the pre-randomid algorithm,
due to demonstrated low-period repeated IDs from the randomized IP_id
code. Consensus is that the low-period repetition (much less than
2^15) is not suitable for general-purpose use.
Allocators of new IPv4 IDs should now call the function ip_newid().
Randomized IP_ids is now a config-time option, "options RANDOM_IP_ID".
ip_newid() can use ip_random-id()_IP_ID if and only if configured
with RANDOM_IP_ID. A sysctl knob should be provided.
This API may be reworked in the near future to support linear ip_id
counters per (src,dst) IP-address pair.
|
| 1.80 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.79 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.78 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.77 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.76 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.75 | 30-Jun-2003 |
itojun | branches: 1.75.2;
better ip_mrouter_detach(). by ono@kame
|
| 1.74 | 26-Jun-2003 |
itojun | fix stats to meet 4.4BSD practice
|
| 1.73 | 26-Jun-2003 |
itojun | expire mrt if mrt_stall goes empty. ono@kame
|
| 1.72 | 26-Jun-2003 |
itojun | cleanup multicast routing stuff on if_detach().
XXX sideeffect to running instance of multicast routing daemon unknown
|
| 1.71 | 14-May-2003 |
itojun | more KNF
|
| 1.70 | 14-May-2003 |
itojun | more KNF
|
| 1.69 | 14-May-2003 |
itojun | wrap multiline macro by do {} while (0)
|
| 1.68 | 14-May-2003 |
itojun | constcond
|
| 1.67 | 14-May-2003 |
itojun | KNF
|
| 1.66 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.65 | 19-Jan-2003 |
simonb | Remove variables that are only assigned too but not referenced.
|
| 1.64 | 05-Nov-2002 |
fair | Add required IPSEC #include files that prevented this from compiling.
|
| 1.63 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.62 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.61 | 31-Jul-2002 |
itojun | bring back old copyright notice lost in rev 1.15 (which is the authors' intent).
|
| 1.60 | 09-Jun-2002 |
itojun | whitespace
|
| 1.59 | 04-Mar-2002 |
sommerfeld | branches: 1.59.6; 1.59.8;
The "gif*" tunnelling interface does everything ipip does.
Move usage example from ipip.4 to gif.4
Excise ipip and stitch up the scars.
|
| 1.58 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.57 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.56 | 22-Jul-2001 |
wiz | branches: 1.56.4;
seperate -> separate
|
| 1.55 | 02-Jun-2001 |
thorpej | branches: 1.55.2;
Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.54 | 08-May-2001 |
itojun | pull encapsulated packet for vif* via ip_encap framework.
|
| 1.53 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.52 | 24-Jan-2001 |
itojun | branches: 1.52.2;
- record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.51 | 08-Nov-2000 |
ad | Update for hashinit() change.
|
| 1.50 | 19-Apr-2000 |
itojun | branches: 1.50.4;
introduce sys/netinet/ip_encap.c, to dispatch inbound packets
to protocol handlers, based on src/dst (for ip proto #4/41).
see comment in ip_encap.c for details of the problem we have.
there are too many protocol specs for ip proto #4/41.
backward compatibility with MROUTING case is now provided in ip_encap.c.
fix ipip to work with gif (using ip_encap.c). sorry for breakage.
gif now uses ip_encap.c.
introduce stf pseudo interface (implements 6to4, another IPv6-over-IPv4 code
with ip proto #41).
|
| 1.49 | 16-Apr-2000 |
chs | remove an LBL ifdef that we can't turn on anyway.
|
| 1.48 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.47 | 23-Mar-2000 |
thorpej | New callout mechanism with two major improvements over the old
timeout()/untimeout() API:
- Clients supply callout handle storage, thus eliminating problems of
resource allocation.
- Insertion and removal of callouts is constant time, important as
this facility is used quite a lot in the kernel.
The old timeout()/untimeout() API has been removed from the kernel.
|
| 1.46 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.45 | 01-Feb-2000 |
thorpej | Fix a couple of whitespace glitches.
|
| 1.44 | 09-Jul-1999 |
thorpej | branches: 1.44.2;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.43 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.42 | 27-Mar-1999 |
nathanw | branches: 1.42.4; 1.42.6;
Alpha printf format fixes.
Closes PR kern/7258.
|
| 1.41 | 02-Feb-1999 |
marc | remove gre_softc declaration; the symbol is no longer used in this
file.
|
| 1.40 | 01-Feb-1999 |
mycroft | Clear mfchashtbl after it's deallocated, to kill a stray pointer. Fixes PR
5400.
|
| 1.39 | 11-Jan-1999 |
thorpej | Adjust for the new IP-IP input path. mrt_ipip_input() is called from
ipip_input(), and returns non-zero if mrt_ipip_input() handled the
packet.
XXX Eventually, the multicast code should probably use regular IP-IP
XXX `interfaces', but mrouted knows about the VIF table, etc.
|
| 1.38 | 22-Dec-1998 |
thorpej | ipip_input() -> mrt_ipip_input().
|
| 1.37 | 13-Sep-1998 |
hwr | branches: 1.37.4;
Add a gre tunnel pseudo network device. Gre = generic route encapsulation.
This device shows up like any other network interface and can be used to
tunnel L3 protocols as e.g. IP over IP.
|
| 1.36 | 07-Feb-1998 |
chs | add flags arg to hashinit(), to pass to malloc().
|
| 1.35 | 14-Aug-1997 |
mycroft | branches: 1.35.4;
Make sure we install the route returned by the upcall before trying to
forward any queued packets. From Bill Fenner, via Brad Karp.
|
| 1.34 | 13-Oct-1996 |
christos | branches: 1.34.10;
backout previous kprintf changes
|
| 1.33 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.32 | 14-Sep-1996 |
mrg | remove an unused variable.
|
| 1.31 | 09-Sep-1996 |
mycroft | Rework the token bucket filter to use a list of packets rather than a static
array. Also, fix several memory leaks. From Bill Fenner.
|
| 1.30 | 09-Sep-1996 |
mycroft | Cosmetic changes, some from Bill Fenner.
|
| 1.29 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.28 | 23-Jun-1996 |
mycroft | Return ENOPROTOOPT rather than picking pseudo-random error values.
Don't allow SIOCGET{VIF,SG}CNT from sockets other than the multicast router.
Restructure rip_ctloutput() like ip_ctloutput(), and fix memory leaks.
|
| 1.27 | 07-May-1996 |
thorpej | branches: 1.27.4;
Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
|
| 1.26 | 16-Mar-1996 |
christos | Fix printf format args.
|
| 1.25 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.24 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.23 | 12-Jun-1995 |
mycroft | Clear the MFC entry's statistical counters when doing an upcall.
|
| 1.22 | 04-Jun-1995 |
mycroft | Simplify ipip_input() a bit. Don't blow away the vif cache if someone sends
us a bogus packet.
|
| 1.21 | 04-Jun-1995 |
mycroft | Simply tbf_control() a bit.
|
| 1.20 | 04-Jun-1995 |
mycroft | Eliminate compiler warnings.
|
| 1.19 | 04-Jun-1995 |
mycroft | For consistency, set sin_len for SIOC{ADD,DEL}MULTI.
|
| 1.18 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.17 | 02-Jun-1995 |
mycroft | Dynamically allocate the deencapsulation interfaces. Abstract the code to
reset a vif into a separate function.
|
| 1.16 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.15 | 31-May-1995 |
mycroft | Integrate multicast 3.5 distribution, with several bugs fixed and general
cleanup. This is a (working) snapshot of work in progress.
|
| 1.14 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.13 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.12 | 09-Jun-1994 |
brezak | Update to version 2 mrouting; from pre-4.4lite NetBSD + 4.4 mods
|
| 1.11 | 04-Jun-1994 |
mycroft | Remove a spurious splx().
|
| 1.10 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.9 | 10-Feb-1994 |
mycroft | Format police.
|
| 1.8 | 10-Feb-1994 |
mycroft | Deprecate af.h.
|
| 1.7 | 29-Jan-1994 |
brezak | Fix some cases of NOT dealing with m_pkthdr's. This code is still suspect though, at least this fixes some panics.
|
| 1.6 | 18-Jan-1994 |
brezak | Fix some prototype detected warnings/errors.
|
| 1.5 | 18-Jan-1994 |
brezak | Patch for ip-multicast bugs from mccanne@ee.lbl.gov (Steven McCanne)
|
| 1.4 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.3 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.2 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.1 | 06-Dec-1993 |
hpeyerl | branches: 1.1.1;
multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.1.1.1 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.27.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.34.10.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.35.4.1 | 03-Feb-1999 |
cgd | pull up rev 1.40 from trunk (mycroft)
|
| 1.37.4.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.42.6.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.42.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.42.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.42.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.44.2.4 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.44.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.44.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.44.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.50.4.4 | 08-Feb-2004 |
msaitoh | Apply patch (requested by Klaus Heinz in ticket #121):
make it compile with "options MROUTING".
|
| 1.50.4.3 | 30-Nov-2003 |
he | Pull up revisions 1.73,1.75 via patch, requested by itojun in ticket #54:
Clean up multicast routing when an interface is detached.
|
| 1.50.4.2 | 04-Sep-2002 |
itojun | pullup sys/netinet/ip_mroute.c 1.61 (itojun)
bring back old copyright notice lost in rev 1.15 (which is the authors' intent).
|
| 1.50.4.1 | 06-Apr-2001 |
he | Pull up revision 1.52 (requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.52.2.8 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.52.2.7 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.52.2.6 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.52.2.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.52.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.52.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.52.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.52.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.55.2.5 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.55.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.55.2.3 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.55.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.55.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.56.4.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.59.8.4 | 11-Jul-2003 |
tron | Pull up revision 1.75 (requested by itojun in ticket #1363):
better ip_mrouter_detach(). by ono@kame
|
| 1.59.8.3 | 30-Jun-2003 |
grant | Pull up revision 1.73 (requested by itojun in ticket #1345):
expire mrt if mrt_stall goes empty. ono@kame
|
| 1.59.8.2 | 30-Jun-2003 |
grant | Pull up revision 1.72 (requested by itojun in ticket #1342):
cleanup multicast routing stuff on if_detach().
|
| 1.59.8.1 | 02-Aug-2002 |
lukem | Pull up revision 1.61 (requested by itojun in ticket #595):
bring back old copyright notice lost in rev 1.15 (which is the authors' intent).
|
| 1.59.6.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.59.6.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.75.2.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.75.2.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.75.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.75.2.4 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.75.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.75.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.75.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.87.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.87.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.87.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.94.2.6 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.94.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.94.2.4 | 07-Dec-2007 |
yamt | sync with head
|
| 1.94.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.94.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.94.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.96.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.96.10.1 | 11-May-2006 |
elad | sync with head
|
| 1.96.8.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.96.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.96.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.96.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.98.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.98.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.98.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.100.8.1 | 03-Jun-2008 |
skrll | Sync with netbsd-4.
|
| 1.100.4.2 | 07-May-2007 |
yamt | sync with head.
|
| 1.100.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.100.2.1 | 17-May-2008 |
bouyer | Pull up following revision(s) (requested by taca in ticket #1141):
sys/netinet/ip_mroute.c: revision 1.113
Make sure to clear csum_flags before forward the packet.
This change should be fix DIAGNOSTIC kernel's panic when the machine act
as multicast router.
Advised from tls@ and approved by thorpej@.
|
| 1.101.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.101.2.4 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.101.2.3 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.101.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.101.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.104.8.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.104.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.104.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.104.6.2 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.104.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.104.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.108.6.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.108.6.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.108.6.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.109.6.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.109.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.111.10.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.111.10.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.111.8.2 | 04-Jun-2008 |
yamt | sync with head
|
| 1.111.8.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.111.6.4 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.111.6.3 | 05-Oct-2008 |
mjf | Sync with HEAD.
|
| 1.111.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.111.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.112.2.3 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.112.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.112.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.114.4.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.116.2.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.116.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.117.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.121.6.3 | 02-Jun-2012 |
mrg | sync to latest -current.
|
| 1.121.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.121.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.121.2.4 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.121.2.3 | 30-Oct-2012 |
yamt | sync with head
|
| 1.121.2.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.121.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.122.2.1 | 23-Oct-2012 |
riz | Pull up following revision(s) (requested by msaitoh in ticket #616):
sys/netinet/if_atm.c: revision 1.33
sys/net/if_arcsubr.c: revision 1.64
sys/netinet/ip_mroute.c: revision 1.126
Add missing "\n" in log(9)
|
| 1.125.2.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.125.2.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.125.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.125.2.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.127.2.2 | 18-May-2014 |
rmind | sync with head
|
| 1.127.2.1 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.128.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.131.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.131.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.131.2.5 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.131.2.4 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.131.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.131.2.2 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.131.2.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.143.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.143.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.145.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.146.6.5 | 07-Dec-2020 |
martin | Pull up following revision(s) (requested by kardel in ticket #1632):
sys/netinet/ip_mroute.c: revision 1.164 (patch)
PR kern/55779:
restore non-desctructive guarantee of ip_mforward() mbuf
argument. This avoids generation invalid UDP checksums
on multicast packets in ip_output().
XXX the root cause of the misguided fix in 2008 should be
XXX investigated
|
| 1.146.6.4 | 13-Jul-2018 |
martin | Pull up following revision(s) via patch (requested by knakahara in ticket #905):
sys/netinet/ip_mroute.c: revision 1.160
sys/netinet6/in6_l2tp.c: revision 1.16
sys/net/if.h: revision 1.263
sys/netinet/in_l2tp.c: revision 1.15
sys/netinet/ip_icmp.c: revision 1.172
sys/netinet/igmp.c: revision 1.68
sys/netinet/ip_encap.c: revision 1.69
sys/netinet6/ip6_mroute.c: revision 1.129
sbappendaddr() is required any lock. Currently, softnet_lock is appropriate.
When rip_input() is called as inetsw[].pr_input, rip_iput() is always called
with holding softnet_lock, that is, in case of !defined(NET_MPSAFE) it is
acquired in ipintr(), otherwise(defined(NET_MPSAFE)) it is acquire in
PR_WRAP_INPUT macro.
However, some function calls rip_input() directly without holding softnet_lock.
That causes assertion failure in sbappendaddr().
rip6_input() and icmp6_rip6_input() are also required softnet_lock for the same
reason.
|
| 1.146.6.3 | 09-Apr-2018 |
bouyer | Pull up following revision(s) (requested by roy in ticket #724):
tests/net/icmp/t_ping.c: revision 1.19
sys/netinet6/raw_ip6.c: revision 1.166
sys/netinet6/ip6_input.c: revision 1.195
sys/net/raw_usrreq.c: revision 1.59
sys/sys/socketvar.h: revision 1.151
sys/kern/uipc_socket2.c: revision 1.128
tests/lib/libc/sys/t_recvmmsg.c: revision 1.2
lib/libc/sys/recv.2: revision 1.38
sys/net/rtsock.c: revision 1.239
sys/netinet/udp_usrreq.c: revision 1.246
sys/netinet6/icmp6.c: revision 1.224
tests/net/icmp/t_ping.c: revision 1.20
sys/netipsec/keysock.c: revision 1.63
sys/netinet/raw_ip.c: revision 1.172
sys/kern/uipc_socket.c: revision 1.260
tests/net/icmp/t_ping.c: revision 1.22
sys/kern/uipc_socket.c: revision 1.261
tests/net/icmp/t_ping.c: revision 1.23
sys/netinet/ip_mroute.c: revision 1.155
sbin/route/route.c: revision 1.159
sys/netinet6/ip6_mroute.c: revision 1.123
sys/netatalk/ddp_input.c: revision 1.31
sys/netcan/can.c: revision 1.3
sys/kern/uipc_usrreq.c: revision 1.184
sys/netinet6/udp6_usrreq.c: revision 1.138
tests/net/icmp/t_ping.c: revision 1.18
socket: report receive buffer overflows
Add soroverflow() which increments the overflow counter, sets so_error
to ENOBUFS and wakes the receive socket up.
Replace all code that manually increments this counter with soroverflow().
Add soroverflow() to raw_input().
This allows userland to detect route(4) overflows so it can re-sync
with the current state.
socket: clear error even when peeking
The error has already been reported and it's pointless requiring another
recv(2) call just to clear it.
socket: remove now incorrect comment that so_error is only udp
As it can be affected by route(4) sockets which are raw.
rtsock: log dropped messages that we cannot report to userland
Handle ENOBUFS when receiving messages.
Don't send messages if the receiver has died.
Sprinkle more soroverflow().
Handle ENOBUFS in recv
Handle ENOBUFS in sendto
Note value received. Harden another sendto for ENOBUFS.
Handle the routing socket overflowing gracefully.
Allow a valid sendto .... duh
Handle errors better.
Fix test for checking we sent all the data we asked to.
|
| 1.146.6.2 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.146.6.1 | 10-Dec-2017 |
snj | Pull up following revision(s) (requested by knakahara in ticket #419):
sys/net/if_stf.c: revision 1.103
sys/net/if_stf.h: revision 1.8
sys/netinet/in_gif.c: revision 1.89
sys/netinet/in_gif.h: revision 1.17
sys/netinet/in_l2tp.c: revision 1.4
sys/netinet/ip_encap.c: revision 1.66
sys/netinet/ip_encap.h: revision 1.23
sys/netinet/ip_mroute.c: revision 1.148
sys/netinet6/in6_gif.c: revision 1.87
sys/netinet6/in6_gif.h: revision 1.16
sys/netinet6/in6_l2tp.c: revision 1.7
sys/netipsec/xform.h: revision 1.13
sys/netipsec/xform_ipip.c: revision 1.55
Add argument to encapsw->pr_input() instead of m_tag.
|
| 1.154.2.6 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.154.2.5 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.154.2.4 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.154.2.3 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.154.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.154.2.1 | 22-Mar-2018 |
pgoyette | Synch with HEAD, resolve conflicts
|
| 1.160.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.163.12.1 | 14-Dec-2020 |
thorpej | Sync w/ HEAD.
|
| 1.163.4.1 | 07-Dec-2020 |
martin | Pull up following revision(s) (requested by kardel in ticket #1143):
sys/netinet/ip_mroute.c: revision 1.164
PR kern/55779:
restore non-desctructive guarantee of ip_mforward() mbuf
argument. This avoids generation invalid UDP checksums
on multicast packets in ip_output().
XXX the root cause of the misguided fix in 2008 should be
XXX investigated
|
| 1.165.10.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.165.4.1 | 29-Jul-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1140):
sys/netinet/ip_output.c: revision 1.330
sys/netinet/sctp_output.c: revision 1.39
sys/netinet/ip_mroute.c: revision 1.166
sys/netipsec/ipsecif.c: revision 1.24
sys/netipsec/xform_ipip.c: revision 1.80
sys/netinet/ip_output.c: revision 1.327
sys/netinet/ip_output.c: revision 1.328
sys/netinet/ip_input.c: revision 1.406
sys/netinet/ip_output.c: revision 1.329
sys/netinet/in_var.h: revision 1.105
in: get rid of unused argument from ip_newid() and ip_newid_range()
in: take a reference of ifp on IP_ROUTETOIF
The ifp could be released after ia4_release(ia).
in: narrow the scope of ifa in ip_output (NFC)
sctp: follow the recent change of ip_newid()
in: avoid racy ifa_acquire(rt->rt_ifa) in ip_output()
If a rtentry is being destroyed asynchronously, ifa referenced by rt_ifa
can be destructed and taking ifa_acquire(rt->rt_ifa) aborts with a
KASSERT failure. Fortunately, the ifa is not actually freed because of
a reference by rt_ifa, it can be available (except some functions like
psref) so as long the rtentry is held.
PR kern/59527
in: avoid racy ia4_acquire(ifatoia(rt->rt_ifa) in ip_rtaddr()
Same as the case of ip_output(), it's racy and should be avoided.
PR kern/59527
|
| 1.36 | 28-Jul-2025 |
kim | Don't hide struct igmpmsg under _KERNEL
The struct is documented as a communication structure with userland.
Fixes PR kern/59561.
|
| 1.35 | 03-Feb-2021 |
roy | branches: 1.35.18; 1.35.24;
CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.34 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.33 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.32 | 07-Feb-2018 |
maxv | branches: 1.32.16;
Remove RSVP_ISI, that's mostly dead code. FreeBSD and OpenBSD too removed
it; FreeBSD kept some pieces but they are mostly no-opts.
Sent on tech-net@, no comment.
|
| 1.31 | 07-Aug-2008 |
cegger | make this compile as proposed by dholland@
|
| 1.30 | 06-Aug-2008 |
plunky | Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.29 | 25-Dec-2007 |
perry | branches: 1.29.6; 1.29.10; 1.29.12; 1.29.16;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.28 | 09-Jul-2007 |
ad | branches: 1.28.8; 1.28.14; 1.28.16; 1.28.20;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.27 | 04-Mar-2007 |
christos | branches: 1.27.2; 1.27.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.26 | 10-Dec-2005 |
elad | branches: 1.26.26;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.25 | 26-Feb-2005 |
perry | branches: 1.25.4;
nuke trailing whitespace
|
| 1.24 | 04-Sep-2004 |
manu | branches: 1.24.4; 1.24.6;
IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.23 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.22 | 18-Apr-2004 |
matt | De __P()
|
| 1.21 | 26-Jun-2003 |
itojun | branches: 1.21.2;
cleanup multicast routing stuff on if_detach().
XXX sideeffect to running instance of multicast routing daemon unknown
|
| 1.20 | 09-Jun-2002 |
itojun | whitespace
|
| 1.19 | 08-May-2001 |
itojun | branches: 1.19.2; 1.19.14; 1.19.16;
pull encapsulated packet for vif* via ip_encap framework.
|
| 1.18 | 23-Mar-2000 |
thorpej | branches: 1.18.4; 1.18.6;
New callout mechanism with two major improvements over the old
timeout()/untimeout() API:
- Clients supply callout handle storage, thus eliminating problems of
resource allocation.
- Insertion and removal of callouts is constant time, important as
this facility is used quite a lot in the kernel.
The old timeout()/untimeout() API has been removed from the kernel.
|
| 1.17 | 20-Nov-1999 |
thorpej | Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.16 | 01-Jul-1999 |
itojun | branches: 1.16.2; 1.16.8;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.15 | 11-Jan-1999 |
thorpej | branches: 1.15.4; 1.15.6;
Adjust for the new IP-IP input path. mrt_ipip_input() is called from
ipip_input(), and returns non-zero if mrt_ipip_input() handled the
packet.
XXX Eventually, the multicast code should probably use regular IP-IP
XXX `interfaces', but mrouted knows about the VIF table, etc.
|
| 1.14 | 22-Dec-1998 |
thorpej | ipip_input() -> mrt_ipip_input().
|
| 1.13 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.12 | 09-Sep-1996 |
mycroft | Rework the token bucket filter to use a list of packets rather than a static
array. Also, fix several memory leaks. From Bill Fenner.
|
| 1.11 | 23-Jun-1996 |
mycroft | Return ENOPROTOOPT rather than picking pseudo-random error values.
Don't allow SIOCGET{VIF,SG}CNT from sockets other than the multicast router.
Restructure rip_ctloutput() like ip_ctloutput(), and fix memory leaks.
|
| 1.10 | 13-Feb-1996 |
christos | branches: 1.10.4;
netinet prototypes
|
| 1.9 | 31-May-1995 |
mycroft | Integrate multicast 3.5 distribution, with several bugs fixed and general
cleanup. This is a (working) snapshot of work in progress.
|
| 1.8 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.7 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.6 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.5 | 09-Jun-1994 |
brezak | Update to version 2 mrouting; from pre-4.4lite NetBSD + 4.4 mods
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.2 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.1 | 06-Dec-1993 |
hpeyerl | branches: 1.1.1;
multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.1 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.10.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.15.6.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.15.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.15.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.16.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.16.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.18.6.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.18.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.18.4.1 | 30-Nov-2003 |
he | Pull up revision 1.21 via patch, requested by itojun in ticket #54:
Clean up multicast routing when an interface is detached.
|
| 1.19.16.1 | 30-Jun-2003 |
grant | Pull up revision 1.21 (requested by itojun in ticket #1342):
cleanup multicast routing stuff on if_detach().
|
| 1.19.14.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.19.2.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.21.2.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.21.2.4 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.21.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.21.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.21.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.24.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.24.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.25.4.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.25.4.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.25.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.26.26.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.27.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.27.2.1 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.28.20.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.28.16.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.28.14.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.28.8.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.29.16.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.29.12.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.29.10.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.29.6.1 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.32.16.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.35.24.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.35.18.1 | 29-Jul-2025 |
martin | Pull up following revision(s) (requested by kim in ticket #1141):
sys/netinet/ip_mroute.h: revision 1.36
Don't hide struct igmpmsg under _KERNEL
The struct is documented as a communication structure with userland.
Fixes PR kern/59561.
|
| 1.63 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.62 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.61 | 08-Jul-2004 |
christos | Bring in flags from 4.1.2 to make things compile.
|
| 1.60 | 29-Jun-2004 |
christos | PR/25999: Jeff Rizzo: ipf: ipnat is corrupting "bimap" translations in 2.0_BETA and -current
|
| 1.59 | 20-May-2004 |
christos | PR/25646: Perry Metzger: Commit a patch that compiles awaiting feedback.
|
| 1.58 | 10-May-2004 |
christos | PR/25103: Martin Husemann: IP Filter 4.4.1 breaks some connections when NATing
patch from Darren applied.
|
| 1.57 | 10-May-2004 |
christos | PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.56 | 28-Mar-2004 |
martti | branches: 1.56.2;
Upgraded IPFilter to 4.1.1
|
| 1.55 | 04-Dec-2003 |
christos | fix unused variable warnings when LARGE_NAT is defined.
|
| 1.54 | 24-Sep-2002 |
sommerfeld | branches: 1.54.6;
Relax overly-conservative TCP option parsing used by ipnat when
hunting for an MSS option to clamp. The previous code assumed that at least
one more byte of options (such as a TCPOPT_EOL) would follow the MSS
option; now, we allow the MSS option to end on the last byte of the
TCP header.
Packets have been observed "in the wild" with a TCP header length of
'6' (24 bytes.. 20 bytes fixed header, 4 bytes options) with a 4-byte
MSS option exactly filling the 4 bytes of options payload and no
following TCPOPT_EOL.
RFC793 is quite explicit that the EOL byte:
" .. need only be used if the end of the options would not
otherwise coincide with the end of the TCP header."
|
| 1.53 | 19-Sep-2002 |
martti | Resync with official IPF
|
| 1.52 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.51 | 09-Jun-2002 |
itojun | whitespace
|
| 1.50 | 05-Jun-2002 |
itojun | typo/bound check fix from YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>
|
| 1.49 | 04-Jun-2002 |
itojun | in mss clamping code, do not go past TCPOPT_EOL. enforce stricter
boundary checking. discussed on tech-net
|
| 1.48 | 02-May-2002 |
martti | branches: 1.48.2; 1.48.4;
Fix compilation problems
|
| 1.47 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.46 | 14-Mar-2002 |
martin | Add MSS clamping to the IP Filter NAT subsystem.
Configured by a new option "mssclamp" in NAT rules, like:
map pppoe0 192.168.1.0/24 -> 0/32 mssclamp 1452
This is based on work by Xiaodan Tang <xtang@qnx.com>.
|
| 1.45 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.44 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.43 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.42 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.41 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.40 | 20-May-2001 |
martin | branches: 1.40.2;
Remove tests for IPN_FRAG bits.
There is no place in the source where this bit could ever be set (or I'm
to blind to find it).
This fixes PR 12671.
If someone thinks this is the wrong solution, please make sure to (a) reopen
the PR and (b) explain to me how the tested bits would ever get set. I'll
be glad to then look further for the real cause (i.e. the flags not getting
set in the case described in the PR).
|
| 1.39 | 06-Apr-2001 |
darrenr | fix fragment cache security hole
|
| 1.38 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.37 | 05-Feb-2001 |
chs | branches: 1.37.2;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.36 | 12-Aug-2000 |
veego | Apply fix from IWAMOTO Toshihiro in pr#10813:
rev 1.35 of ip_nat.c checks if packets are too short.
For ICMP packets, this packet length checking double counts
the length of an IP header contained in ICMP messages.
So, unless ICMP packets are long enough (such as echo-reply),
packets are mistakingly considered too short and are dropped.
|
| 1.35 | 09-Aug-2000 |
veego | Resolve conflicts.
|
| 1.34 | 12-Jun-2000 |
veego | branches: 1.34.2;
Resolve conflicts.
|
| 1.33 | 21-May-2000 |
veego | branches: 1.33.2;
Resolve conflicts.
|
| 1.32 | 11-May-2000 |
veego | Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.31 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.30 | 16-Apr-2000 |
chs | remove ifdefs to skip htons() on some big-endian platforms.
|
| 1.29 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.28 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.27 | 28-Dec-1999 |
darrenr | update ipfilter code to 3.3.6
|
| 1.26 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.25 | 05-Mar-1999 |
mycroft | branches: 1.25.2; 1.25.8; 1.25.14;
Minor cleanup to use LONG_SUM() and CALC_SUMD() more.
|
| 1.24 | 02-Feb-1999 |
cjs | Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.23 | 22-Nov-1998 |
mrg | branches: 1.23.2;
merge ipf 3.2.10
|
| 1.22 | 15-Nov-1998 |
drochner | fix the previous: "securelevel" in kernel only
|
| 1.21 | 14-Nov-1998 |
tls | In 'highly secure' mode (securelevel >= 2), the filter lists may not be tampered with. It might be desirable to allow enabling of preset filter lists, but it seems too good a candidate for a denial-of-service attack, so we don't.
|
| 1.20 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.19 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.18 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.17 | 29-Mar-1998 |
scottr | Change from IP-Filter 3.2.3: avoid infinite loop in nat_new() when
NAT'ing to a single IP address.
|
| 1.16 | 25-Nov-1997 |
mrg | fixes for memory leaks in proxying, and byte ordering problems. from darren reed.
|
| 1.15 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.14 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.13 | 21-Sep-1997 |
veego | branches: 1.13.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.12 | 21-Jul-1997 |
kleink | branches: 1.12.2;
Nuke an `#ifdef sparc' conditional around ntohs() usage: this (1) is incomplete
and (2) makes no difference anyway. Also, minor KNF.
|
| 1.11 | 16-Jul-1997 |
kleink | Fix a misplaced brace which caused NAT list corruption; from Dave Huang
<khym@bga.com> in PR kern/3872.
|
| 1.10 | 06-Jul-1997 |
thorpej | Restore original RCS IDs.
|
| 1.9 | 06-Jul-1997 |
thorpej | Deal with a bogus warning from -Wuninitialized.
|
| 1.8 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.7 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.6 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures again:
- Add prototypes.
- Get arguments to ioctl right (cmd is a u_long in NetBSD)
|
| 1.5 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.4 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.3 | 29-Jan-1997 |
thorpej | ioctl cmd arguments are u_long, not int. Pointed out by
Fred L. Templin <templin@nas.nasa.gov>
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.27 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.26 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.25 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.24 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.23 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.22 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.21 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.20 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.19 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.18 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.17 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.16 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.15 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.14 | 28-Dec-1999 |
darrenr | update DARRENR branch of netinet to 3.3.6
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.12.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.13.2.5 | 24-Nov-1998 |
cgd | pull up rev(s) 1.23 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.13.2.4 | 22-Jul-1998 |
mellon | Pull up 1.20 (veego)
|
| 1.13.2.3 | 25-Nov-1997 |
mrg | pull up from trunk: fixes for memory leaks in proxying, and byte ordering problems. from darren reed.
|
| 1.13.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.13.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.23.2.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.25.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.25.8.4 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.25.8.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.25.8.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.25.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.25.2.3 | 14-Apr-2001 |
he | Pull up revision 1.39 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.25.2.2 | 08-Jan-2000 |
he | Pull up revision 1.27 (requested by darrenr):
Update IPF to version 3.3.6.
|
| 1.25.2.1 | 20-Dec-1999 |
he | Pull up revision 1.26 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.33.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.34.2.6 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.34.2.5 | 26-Feb-2002 |
he | Apply patch (requested by darrenr):
Prevent panic when packets are received in a certain order.
|
| 1.34.2.4 | 09-Feb-2002 |
he | Pull up revisions 1.37-1.38,1.40-1.44 (via patch, requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.34.2.3 | 14-Apr-2001 |
he | Pull up revision 1.39 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.34.2.2 | 31-Aug-2000 |
veego | Pull up revision 1.36 (requested by veego). Approved by releng-1-5.
>Apply fix from IWAMOTO Toshihiro in pr#10813:
> rev 1.35 of ip_nat.c checks if packets are too short.
> For ICMP packets, this packet length checking double counts
> the length of an IP header contained in ICMP messages.
> So, unless ICMP packets are long enough (such as echo-reply),
> packets are mistakingly considered too short and are dropped.
|
| 1.34.2.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.37.2.10 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.37.2.9 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.37.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.37.2.7 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.37.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.37.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.37.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.37.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.37.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.37.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.40.2.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.40.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.40.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.40.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.48.4.6 | 30-Dec-2003 |
jmc | Backout changes from #1568 as too many people are reporting problems
with "out of the box" configurations.
|
| 1.48.4.5 | 26-Nov-2003 |
cyber | Patch (requested by jklos ticket #1564):
Change to ip filter"s NAT code to keep excessive NAT entries from
causing the kernel to panic.
|
| 1.48.4.4 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.48.4.3 | 13-Oct-2002 |
lukem | Pull up revision 1.54 (requested by sommerfeld in ticket #884):
Relax overly-conservative TCP option parsing used by ipnat when
hunting for an MSS option to clamp. The previous code assumed that at least
one more byte of options (such as a TCPOPT_EOL) would follow the MSS
option; now, we allow the MSS option to end on the last byte of the
TCP header.
Packets have been observed "in the wild" with a TCP header length of
'6' (24 bytes.. 20 bytes fixed header, 4 bytes options) with a 4-byte
MSS option exactly filling the 4 bytes of options payload and no
following TCPOPT_EOL.
RFC793 is quite explicit that the EOL byte:
" .. need only be used if the end of the options would not
otherwise coincide with the end of the TCP header."
|
| 1.48.4.2 | 05-Jun-2002 |
lukem | Pull up revision 1.50 (requested by itojun in ticket #177):
typo/bound check fix from YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>
|
| 1.48.4.1 | 05-Jun-2002 |
lukem | Pull up revision 1.49 (requested by itojun in ticket #154):
in mss clamping code, do not go past TCPOPT_EOL. enforce stricter
boundary checking. discussed on tech-net
|
| 1.48.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.54.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.54.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.56.2.5 | 13-Aug-2004 |
jmc | branches: 1.56.2.5.2;
Pullup rev 1.61-1.62 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.56.2.4 | 02-Jul-2004 |
he | Pull up revision 1.60 (requested by christos in ticket #571):
Fix problem where ipnat is corrupting ``bimap'' translations.
Fixes PR#25999.
|
| 1.56.2.3 | 30-May-2004 |
tron | Pull up revision 1.59 (requested by christos in ticket #416):
PR/25646: Perry Metzger: Commit a patch that compiles awaiting feedback.
|
| 1.56.2.2 | 30-May-2004 |
tron | Pull up revision 1.58 (requested by christos in ticket #416):
PR/25103: Martin Husemann: IP Filter 4.4.1 breaks some connections when NATing
patch from Darren applied.
|
| 1.56.2.1 | 30-May-2004 |
tron | Pull up revision 1.57 (requested by christos in ticket #416):
PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.56.2.5.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.63 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.35 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.34 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.33 | 08-Jul-2004 |
christos | Bring in flags from 4.1.2 to make things compile.
|
| 1.32 | 10-May-2004 |
christos | PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.31 | 28-Mar-2004 |
martti | branches: 1.31.2;
Upgraded IPFilter to 4.1.1
|
| 1.30 | 16-Jan-2004 |
abs | Allow DEF_NAT_AGE to be set in kernel config.
|
| 1.29 | 03-Jan-2004 |
tron | Remove extra tokens at end of #undef directive.
|
| 1.28 | 16-Dec-2003 |
abs | Comment out #undef LARGE_NAT so LARGE_NAT can be set in a kernel config file
without having to edit this file as well.
|
| 1.27 | 19-Sep-2002 |
martti | branches: 1.27.6;
Upgraded IPFilter to 3.4.29
|
| 1.26 | 02-May-2002 |
martti | branches: 1.26.4;
Upgraded IPFilter to 3.4.27
|
| 1.25 | 14-Mar-2002 |
martin | Add MSS clamping to the IP Filter NAT subsystem.
Configured by a new option "mssclamp" in NAT rules, like:
map pppoe0 192.168.1.0/24 -> 0/32 mssclamp 1452
This is based on work by Xiaodan Tang <xtang@qnx.com>.
|
| 1.24 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.23 | 06-Apr-2001 |
darrenr | branches: 1.23.2;
fix fragment cache security hole
|
| 1.22 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.21 | 09-Aug-2000 |
veego | branches: 1.21.2;
Resolve conflicts.
|
| 1.20 | 12-Jun-2000 |
veego | branches: 1.20.2;
Resolve conflicts.
|
| 1.19 | 21-May-2000 |
veego | branches: 1.19.2;
Resolve conflicts.
|
| 1.18 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.17 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.16 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.15 | 22-Nov-1998 |
mrg | branches: 1.15.4; 1.15.10; 1.15.16;
merge ipf 3.2.10
|
| 1.14 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.13 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.12 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.11 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.10 | 21-Sep-1997 |
veego | branches: 1.10.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.9 | 06-Jul-1997 |
thorpej | branches: 1.9.2;
Restore original RCS IDs.
|
| 1.8 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.7 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.6 | 27-May-1997 |
thorpej | Make this compile on 32-bit architecutres again:
- Get arguments to ioctl right (cmd is a u_long in NetBSD)
|
| 1.5 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.4 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.3 | 29-Jan-1997 |
thorpej | ioctl cmd arguments are u_long, not int. Pointed out by
Fred L. Templin <templin@nas.nasa.gov>
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.24 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.23 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.22 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.21 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.20 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.19 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.18 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.17 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.16 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.15 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.14 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.9.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.10.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.15 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.10.2.3 | 22-Jul-1998 |
mellon | Pull up 1.14 (veego)
|
| 1.10.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.10.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.15.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.15.10.3 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.15.10.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.15.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.15.4.2 | 14-Apr-2001 |
he | Pull up revision 1.23 (via patch, requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.15.4.1 | 20-Dec-1999 |
he | Pull up revision 1.16 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.19.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.20.2.4 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.20.2.3 | 09-Feb-2002 |
he | Pull up revisions 1.22,1.24 (requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.20.2.2 | 14-Apr-2001 |
he | Pull up revision 1.23 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.20.2.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.21.2.5 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.21.2.4 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.21.2.3 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.21.2.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.21.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.23.2.3 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.23.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.23.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.26.4.3 | 30-Dec-2003 |
jmc | Backout changes from #1568 as too many people are reporting problems
with "out of the box" configurations.
|
| 1.26.4.2 | 26-Nov-2003 |
cyber | Patch (requested by jklos ticket #1564):
Change to ip filter"s NAT code to keep excessive NAT entries from
causing the kernel to panic.
|
| 1.26.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.27.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.27.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.31.2.2 | 13-Aug-2004 |
jmc | branches: 1.31.2.2.2;
Pullup rev 1.33-1.34 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.31.2.1 | 30-May-2004 |
tron | Pull up revision 1.32 (requested by christos in ticket #416):
PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.31.2.2.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.35 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.6 | 02-Oct-2004 |
christos | These are ipfilter files, although they don't have the same copyright.
Thanks jaromir.
|
| 1.5 | 28-Mar-2004 |
martti | branches: 1.5.4;
Upgraded IPFilter to 4.1.1
|
| 1.4 | 19-Sep-2002 |
martti | branches: 1.4.6;
Resync with official IPF
|
| 1.3 | 09-Jun-2002 |
itojun | whitespace
|
| 1.2 | 01-Apr-2002 |
jdolecek | branches: 1.2.2;
add __KERNEL_RCSID()
|
| 1.1 | 24-Jan-2002 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.1 | 24-Jan-2002 |
martti | branches: 1.1.1.1.2; 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 3.4.23
|
| 1.1.1.1.6.5 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.1.1.1.6.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.1.1.1.6.3 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.1.1.1.6.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.1.1.1.6.1 | 24-Jan-2002 |
nathanw | file ip_netbios_pxy.c was added on branch nathanw_sa on 2002-02-28 04:15:09 +0000
|
| 1.1.1.1.4.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.1.1.1.4.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.1.1.1.4.3 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.1.1.1.4.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.1.1.1.4.1 | 24-Jan-2002 |
jdolecek | file ip_netbios_pxy.c was added on branch kqueue on 2002-02-11 20:10:36 +0000
|
| 1.1.1.1.2.4 | 27-Nov-2002 |
itojun | sys/netinet/ip_h323_pxy.c via patch
sys/netinet/ip_ipsec_pxy.c via patch
sys/netinet/ip_netbios_pxy.c via patch
Fix compilation on a.out systems.
(Thorsten Frueauf)
|
| 1.1.1.1.2.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.1.1.1.2.2 | 09-Feb-2002 |
he | Pull up revision 1.1 (new, requested by martti):
Updated IPFilter to 3.4.23
|
| 1.1.1.1.2.1 | 24-Jan-2002 |
he | file ip_netbios_pxy.c was added on branch netbsd-1-5 on 2002-02-09 16:55:22 +0000
|
| 1.2.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.4.6.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.4.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.4.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.4.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.5.4.1 | 06-Feb-2005 |
jmc | Pull up revision 1.6 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.330 | 17-Jul-2025 |
ozaki-r | in: avoid racy ifa_acquire(rt->rt_ifa) in ip_output()
If a rtentry is being destroyed asynchronously, ifa referenced by rt_ifa
can be destructed and taking ifa_acquire(rt->rt_ifa) aborts with a
KASSERT failure. Fortunately, the ifa is not actually freed because of
a reference by rt_ifa, it can be available (except some functions like
psref) so as long the rtentry is held.
PR kern/59527
|
| 1.329 | 11-Jun-2025 |
ozaki-r | in: narrow the scope of ifa in ip_output (NFC)
|
| 1.328 | 11-Jun-2025 |
ozaki-r | in: take a reference of ifp on IP_ROUTETOIF
The ifp could be released after ia4_release(ia).
|
| 1.327 | 11-Jun-2025 |
ozaki-r | in: get rid of unused argument from ip_newid() and ip_newid_range()
|
| 1.326 | 19-Apr-2023 |
mlelstv | branches: 1.326.6;
Again allow multicast packets to be sent from unnumbered interfaces.
|
| 1.325 | 19-Apr-2023 |
ozaki-r | Revert "Fix panic on packet sending via a route with rt_ifa of AF_LINK."
The fix is mistakenly upstreamed.
|
| 1.324 | 21-Nov-2022 |
knakahara | branches: 1.324.2;
Fix panic on packet sending via a route with rt_ifa of AF_LINK.
A route with rt_ifa of AF_LINK can be set by some routing daemons when
it adds a route that has a gateway of AF_LINK. If there is no address on
a target interface, the kernel sets an AF_LINK address of the interface to
rt_ifa of the route. In that case, a variable of a local address in
ip_output (ia) can be NULL and we need more NULL-checks of it.
|
| 1.323 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.322 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.321 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.320 | 08-Sep-2020 |
christos | Add IP_BINDANY, IPV6_BINDANY which can be used to bind to any address in
order to implement transparent proxies.
|
| 1.319 | 28-Aug-2020 |
christos | Don't cache the sa, because we are dealing with multiple mbufs (from ozaki-r)
|
| 1.318 | 28-Aug-2020 |
ozaki-r | inet: reduce silent packet discards
|
| 1.317 | 28-Aug-2020 |
ozaki-r | inet: reduce indents of a normal path to improve readability (NFCI)
|
| 1.316 | 28-Aug-2020 |
ozaki-r | inet, inet6: count packets dropped by IPsec
The counters count packets dropped due to security policy checks.
|
| 1.315 | 27-Dec-2019 |
msaitoh | s/referece/reference/ in comment.
|
| 1.314 | 05-Jun-2019 |
knakahara | The packets which will be esp-fragmented should not be applied pfil. Pointed out by ohishi@IIJ, thanks.
|
| 1.313 | 05-Jun-2019 |
knakahara | Fix rtcache cannot be released once an esp-fragmented packet is sent. Pointed out by ohishi@IIJ, thanks.
|
| 1.312 | 15-May-2019 |
ozaki-r | Get rid of IFNET_LOCK for if_mcast_op to avoid a deadlock
The IFNET_LOCK was added to avoid data races on if_flags for IFF_ALLMULTI.
Unfortunatetly it caused a deadlock instead. A known scenario causing a
deadlock is to occur the following two operations concurrently: (a) a removal of
an IP adddres assigned to an interface and (b) a manipulation of multicast
groups to the interface. The resource dependency graph is like this:
softnet_lock => IFNET_LOCK => psref_target_destroy => softint => softnet_lock
Thanks to the previous commit that avoids data races on if_flags for
IFF_ALLMULTI by another approach, we can remove IFNET_LOCK and defuse the
deadlock.
PR kern/54189
|
| 1.311 | 13-May-2019 |
ozaki-r | Count packets dropped by pfil
|
| 1.310 | 04-Feb-2019 |
mrg | rework the #ifdef IPSEC code to not use fallthru.
same number of lines with more local context.
|
| 1.309 | 22-Dec-2018 |
maxv | Replace: M_MOVE_PKTHDR -> m_move_pkthdr. No functional change, since the
former is a macro to the latter.
|
| 1.308 | 12-Dec-2018 |
rin | Simplify logic in ip{,6}_output().
Now, we have M_CSUM_TSOv[46] bit in ifp->if_csum_flags_tx when
TSO[46] is enabled for the interface. So we can simply check
whether TSO[46] is required in a packet but missing in the
interface by (sw_csum & M_CSUM_TSOv[46]).
Note that this is a very rare case where TSO[46] is suddenly
turned off during a packet passing b/w TCP and IP.
part of PR kern/53562
OK msaitoh
|
| 1.307 | 11-Jul-2018 |
maxv | Rename
ip_undefer_csum -> in_undefer_cksum
in_delayed_cksum -> in_undefer_cksum_tcpudp
The two previous names were inconsistent and misleading.
Put the two functions into in_offload.c. Add comments to explain what
we're doing.
The same could be done for IPv6.
|
| 1.306 | 02-Jun-2018 |
maxv | branches: 1.306.2;
Copy more mbuf flags.
|
| 1.305 | 29-May-2018 |
maxv | Fix an XXX of mine, be clearer about what we're doing. Basically we want to
preserve the fragment offset and flags. That's necessary if the packet
we're fragmenting is itself a fragment.
|
| 1.304 | 29-Apr-2018 |
maxv | Remove unused and misleading argument from ipsec_set_policy.
|
| 1.303 | 21-Apr-2018 |
maxv | Remove #ifndef __vax__.
The check enforces a 4-byte-aligned size for the option mbuf. If the size
is not multiple of 4, the computation of ip_hl gets truncated in the
output path. There is no reason for this check not to be present on VAX.
While here add a KASSERT in ip_insertoptions to enforce the assumption.
Discussed briefly on tech-net@
|
| 1.302 | 13-Apr-2018 |
maxv | Remove useless comment and style.
|
| 1.301 | 13-Apr-2018 |
maxv | Reduce the diff between similar blocks.
|
| 1.300 | 13-Apr-2018 |
maxv | Reorder a few instructions to clarify. Replace two bcopy by memcpy.
|
| 1.299 | 30-Mar-2018 |
maya | correct typo: and and -> and (comments only)
heads up on this being a common typo from chris28.
|
| 1.298 | 03-Mar-2018 |
maxv | branches: 1.298.2;
Add KASSERTs, we don't want m_nextpkt in ipsec{4/6}_process_packet.
|
| 1.297 | 27-Feb-2018 |
maxv | Dedup: merge ipsec4_set_policy and ipsec6_set_policy. The content of the
original ipsec_set_policy function is inlined into the new one.
|
| 1.296 | 27-Feb-2018 |
maxv | Dedup: merge
ipsec4_get_policy and ipsec6_get_policy
ipsec4_delete_pcbpolicy and ipsec6_delete_pcbpolicy
The already-existing ipsec_get_policy() function is inlined in the new
one.
|
| 1.295 | 12-Feb-2018 |
christos | Keep a pointer to the interface of the multicast membership, because the
multicast element itself might go away in in_delmulti (but the interface
can't because we hold the lock). From ozaki-r@
|
| 1.294 | 07-Feb-2018 |
mrg | ip_add_membership() has an missing {} issue, but solve it by
dropping the "goto out" that would have happened immediately
next anyway, ie, should be NFC.
|
| 1.293 | 06-Feb-2018 |
maxv | Several changes, mostly cosmetic:
* Add a KASSERT in ip_output(), we expect (at least) the IP header to be
here.
* In ip_fragment(), declare two variables instead of recomputing the
values each time. Add an XXX for ipoff, it seems to me we should also
remove IP_RF.
* Rename the arguments of ip_optcopy().
* Style: use NULL for pointers, remove ()s for return statements, and
add whitespaces for clarity.
No real functional change.
|
| 1.292 | 10-Jan-2018 |
christos | from ozaki-r: use the proper ifp.
XXX: perhaps push the lock in in_delmulti()?
|
| 1.291 | 10-Jan-2018 |
christos | - this is not python, we need braces
- protect ifp locking against NULL
|
| 1.290 | 01-Jan-2018 |
christos | Remove comment now that the getsockopt code passes the size.
|
| 1.289 | 01-Jan-2018 |
christos | 1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
|
| 1.288 | 22-Dec-2017 |
ozaki-r | Fix usage of curlwp_bind in ip_output
curlwp_bindx must be called in LIFO order, i.e., we can't call curlwp_bind
and curlwp_bindx like this:
bound1 = curlwp_bind();
bound2 = curlwp_bind();
curlwp_bindx(bound1);
curlwp_bindx(bound2);
ip_outout did so if NET_MPSAFE. Fix it.
|
| 1.287 | 15-Dec-2017 |
ozaki-r | Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
|
| 1.286 | 11-Dec-2017 |
ryo | As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.285 | 17-Nov-2017 |
ozaki-r | Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
|
| 1.284 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.283 | 23-Jul-2017 |
para | kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
|
| 1.282 | 04-Jul-2017 |
roy | Rename u to udst, .dst to .sa and .dst4 to sin.
Create sockaddr for the source address in usrc so it won't stamp on udst.
This fixes a regression caused in r1.280
|
| 1.281 | 03-Jul-2017 |
khorben | Typo
|
| 1.280 | 03-Jul-2017 |
roy | When outputting, search for the sending address on the sending interface
rather than blindly picking the first matcing address from any interface
when testing source address validity.
This allows another interface to have the same address, but be detached.
|
| 1.279 | 12-May-2017 |
ryo | branches: 1.279.2;
replace in_fmtaddr() by IN_PRINT(), and delete function in_fmtaddr()
|
| 1.278 | 10-May-2017 |
ozaki-r | Stop ipsec4_output returning SP to the caller
SP isn't used by the caller (ip_output) and also holding its
reference looks unnecessary.
|
| 1.277 | 07-May-2017 |
christos | PR/52074: Frank Kardel: current npf map directive broken
Don't filter packets that can't be resolved to source interfaces because
they could have been generated by a packet filter.
|
| 1.276 | 05-Mar-2017 |
ozaki-r | branches: 1.276.4;
Fix the position of curlwp_bindx; it should be after if_put
|
| 1.275 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.274 | 02-Mar-2017 |
ozaki-r | Make sure imo_membership is protected by inp's lock (solock)
|
| 1.273 | 02-Mar-2017 |
ozaki-r | Make usages of ifp MP-safe in some functions of IP multicast
|
| 1.272 | 22-Feb-2017 |
ozaki-r | Add assertions and comments for lock states of socket and pcb
|
| 1.271 | 17-Feb-2017 |
ozaki-r | Make NOMPSAFE comments informative
|
| 1.270 | 13-Feb-2017 |
ozaki-r | Use IFQ_LOCK instead of splnet for if_snd
|
| 1.269 | 16-Jan-2017 |
christos | rename arplog -> ARPLOG to make it clear that it is a macro and tuck-in the
buffer used for address formatting.
|
| 1.268 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.267 | 11-Jan-2017 |
ozaki-r | branches: 1.267.2;
Get rid of unnecessary header inclusions
|
| 1.266 | 10-Jan-2017 |
knakahara | avoid double rtcache_unref().
reviewed by ozaki-r@n.o.
|
| 1.265 | 12-Dec-2016 |
ozaki-r | Make the routing table and rtcaches MP-safe
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
|
| 1.264 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.263 | 20-Sep-2016 |
roy | Drop UDP packets as well as TCP without error when sending from detached or
tentative addresses.
|
| 1.262 | 18-Sep-2016 |
christos | Dealing with arplog is a bit more complicated...
|
| 1.261 | 15-Sep-2016 |
roy | Ensure that packets are sent from a valid address.
If the packet is TCP and the address is detached or tentative then
it's just dropped, otherwise an error is returned.
This is needed because you can bind to a valid address and it can then
become invalid.
This satisfies RFC 4862 section 5.5.4.
|
| 1.260 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.259 | 08-Jul-2016 |
ozaki-r | branches: 1.259.2;
Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
|
| 1.258 | 21-Jun-2016 |
ozaki-r | Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.257 | 20-Jun-2016 |
knakahara | apply if_output_lock() to L3 callers which call ifp->if_output() of L2(or L3 tunneling).
|
| 1.256 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.255 | 09-May-2016 |
ozaki-r | Fix compilation for ppc
|
| 1.254 | 04-May-2016 |
christos | fix compilation for ppc.
|
| 1.253 | 28-Apr-2016 |
ozaki-r | Constify rtentry of if_output
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
|
| 1.252 | 26-Apr-2016 |
ozaki-r | Stop using rt_gwroute on packet sending paths
rt_gwroute of rtentry is a reference to a rtentry of the gateway
for a rtentry with RTF_GATEWAY. That was used by L2 (arp and ndp)
to look up L2 addresses. By separating L2 nexthop caches, we don't
need a route for the purpose and we can stop using rt_gwroute.
By doing so, we can reduce referencing and modifying rtentries,
which makes it easy to apply a lock (and/or psref) to the
routing table and rtentries.
One issue to do this is to keep RTF_REJECT behavior. It seems it
was broken when we moved rtalloc1 things from L2 output routines
(e.g., ether_output) to ip_hresolv_output, but (fortunately?)
it works unexpectedly. What we mistook are:
- RTF_REJECT was checked for any routes in L2 output routines,
but in ip_hresolv_output it is checked only when the route
is RTF_GATEWAY
- The RTF_REJECT check wasn't copied to IPv6 (nd6_output)
It seems that rt_gwroute checks hid the mistakes and it looked
work (unexpectedly) and removing rt_gwroute checks unveil the
issue. So we need to fix RTF_REJECT checks in ip_hresolv_output
and also add them to nd6_output.
One more point we have to care is returning an errno; we need
to mimic looutput behavior. Originally RTF_REJECT check was
done either in L2 output routines or in looutput. The latter is
applied when a reject route directs to a loopback interface.
However, now RTF_REJECT check is done before looutput so to keep
the original behavior we need to return an errno which looutput
chooses. Added rt_check_reject_route does such tweaks.
|
| 1.251 | 19-Apr-2016 |
ozaki-r | Fix error path
|
| 1.250 | 19-Apr-2016 |
ozaki-r | Separate MPLS-related routines from ip_hresolv_output
No functional changes.
|
| 1.249 | 18-Apr-2016 |
ozaki-r | Get rid of meaningless RTF_UP check from ip_hresolv_output
The check is meaningless because
- An obtained rtentry is ensured that it's always RTF_UP by rtcache,
rtalloc1 and rtlookup. If the rtentry isn't changed (i.e., RTF_UP gets
dropped) during processing, the check should be unnecessary
- Even if not, i.e., an obtained rtentry can be changed during processing,
checking only at the point doesn't help; the rtentry can be changed after
the check
Instead we have to ensure that RTF_UP isn't dropped if someone is using it
somehow. Note that we already ensure that a rtentry being used isn't freed
by rt_refcnt.
Proposed on tech-kern and tech-net.
|
| 1.248 | 20-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.247 | 02-Sep-2015 |
ozaki-r | Do rt_refcnt++ when set a rtentry to another rtentry's rt_gwroute
And also do rtfree when deref a rtentry from rt_gwroute.
|
| 1.246 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.245 | 07-Aug-2015 |
ozaki-r | Use time_uptime instead of time_second to avoid time leaps
Some codes in sys/net* use time_second to manage time periods such as
cache expirations. However, time_second doesn't increase monotonically
and can leap by say settimeofday(2) according to time_second(9). We
should use time_uptime instead of it to avoid such time leaps.
This change replaces time_second with time_uptime. Additionally it
converts a time based on time_uptime to a time based on time_second
when the kernel passes the time to userland programs that expect
the latter, and vice versa.
Note that we shouldn't leak time_uptime to other hosts over the
netowrk. My investigation shows there is no such leak:
http://mail-index.netbsd.org/tech-net/2015/08/06/msg005332.html
Discussed on tech-kern and tech-net.
|
| 1.244 | 17-Jul-2015 |
ozaki-r | Reform use of rt_refcnt
rt_refcnt of rtentry was used in bad manners, for example, direct rt_refcnt++
and rt_refcnt-- outside route.c, "rt->rt_refcnt++; rtfree(rt);" idiom, and
touching rt after rt->rt_refcnt--.
These abuses seem to be needed because rt_refcnt manages only references
between rtentry and doesn't take care of references during packet processing
(IOW references from local variables). In order to reduce the above abuses,
the latter cases should be counted by rt_refcnt as well as the former cases.
This change improves consistency of use of rt_refcnt:
- rtentry is always accessed with rt_refcnt incremented
- rtentry's rt_refcnt is decremented after use (rtfree is always used instead
of rt_refcnt--)
- functions returning rtentry increment its rt_refcnt (and caller rtfree it)
Note that rt_refcnt prevents rtentry from being freed but doesn't prevent
rtentry from being updated. Toward MP-safe, we need to provide another
protection for rtentry, e.g., locks. (Or introduce a better data structure
allowing concurrent readers during updates.)
|
| 1.243 | 14-Jul-2015 |
ozaki-r | Move rt_gwroute operation out of stripoutput
We should do it in ip_hresolv_needed.
|
| 1.242 | 01-Jul-2015 |
ozaki-r | Use ip_hresolv_output for if_token as well
I thought we cannot apply ip_hresolv_output to if_token because
rt0 looked being needed by arpresolve in token_output. However,
rt0 is actually not used by arpresolve in NetBSD (see obsolete
ARPRESOLVE macro).
|
| 1.241 | 08-Jun-2015 |
roy | errno -> error, spotted by the hawk skrll
|
| 1.240 | 08-Jun-2015 |
roy | It's possible we could not have any ready addresses.
|
| 1.239 | 04-Jun-2015 |
ozaki-r | Pull out route lookups from L2 output routines
Route lookups for routes of RTF_GATEWAY were done in L2 output
routines such as ether_output, but they should be done in L3
i.e., before L2 output routines. This change places the lookups
between L3 output routines (say ip_output) and the L2 output
routines.
The change is based on dyoung's patch submitted in the thread:
https://mail-index.netbsd.org/tech-net/2013/02/01/msg003847.html
You can find out detailed investigations by dyoung about the
issue in there.
Note that the change introduces a workaround for MPLS. ether_output
knew that it needs to fill the ethertype of a frame as MPLS,
based on a tag of an original route (rtentry), but now we don't
pass it to ehter_output. So we have to tell that in another way.
We use mtag to do so for now, which introduces some overhead.
We should fix it somehow in the future.
Discussed on tech-kern and tech-net.
|
| 1.238 | 27-Apr-2015 |
ozaki-r | Add missing error checks on rtcache_setdst
It can fail with ENOMEM.
|
| 1.237 | 24-Apr-2015 |
ozaki-r | KNF
|
| 1.236 | 03-Apr-2015 |
ozaki-r | Don't grab KERNEL_LOCK during if_output when NET_MPSAFE
The change makes L3 MP-safe work easy. At this point
we deal with only IP forwarding.
No functional change when NET_MPSAFE isn't enabled.
|
| 1.235 | 31-Mar-2015 |
ozaki-r | Add missing ifdef IPSEC
|
| 1.234 | 23-Mar-2015 |
roy | Add RTF_BROADCAST to mark routes used for the broadcast address when
they are created on the fly. This makes it clear what the route is for
and allows an optimisation in ip_output() by avoiding a call to
in_broadcast() because most of the time we do talk to a host.
It also avoids a needless allocation for the storage of llinfo_arp and
thus vanishes from arp(8) - it showed as incomplete anyway so this
is a nice side effect.
Guard against this and routes marked with RTF_BLACKHOLE in
ip_fastforward().
While here, guard against routes marked with RTF_BLACKHOLE in
ip6_fastforward().
RTF_BROADCAST is IPv4 only, so don't bother checking that here.
|
| 1.233 | 26-Nov-2014 |
ozaki-r | branches: 1.233.2;
Call looutput with holding KERNEL_LOCK
This fixes diagnostic assertion "KERNEL_LOCKED_P()" in if_loop.c.
PR kern/49410
|
| 1.232 | 12-Oct-2014 |
christos | Refactor the multicast membership code so that we can handle v4 mapped
addresses using the v6 membership ioctls.
|
| 1.231 | 11-Oct-2014 |
christos | exposet multicast option functions which are used by the v6 code now.
|
| 1.230 | 06-Jun-2014 |
rmind | branches: 1.230.2;
ip_output: zero iproute structure only when needed; reduce the scope
of some variables.
|
| 1.229 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.228 | 29-May-2014 |
rmind | Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.227 | 23-May-2014 |
rmind | Fix the assert in the previous commit.
|
| 1.226 | 22-May-2014 |
rmind | - Make ip_setmoptions(), ip_getmoptions() and ip_pcbopts() static.
- ip_output: eliminate 7th variadic argument; IP_RETURNMTU is flag
always used to store MTU size into struct inpcb::inp_errormtu.
- Clean up these routines: reduce #ifdefs, variable scopes, etc.
|
| 1.225 | 17-May-2014 |
rmind | Replace open-coded access (and boundary checking) of ifindex2ifnet with
if_byindex() function.
|
| 1.224 | 29-Jun-2013 |
rmind | branches: 1.224.4;
- Rewrite parts of pfil(9): use array to store hooks and thus be more cache
friendly (there are only few hooks in the system). Make the structures
opaque and the interface more strict.
- Remove PFIL_HOOKS option by making pfil(9) mandatory.
|
| 1.223 | 27-Jun-2013 |
christos | branches: 1.223.2;
implement IP_PKTINFO and IP_RECVPKTINFO.
|
| 1.222 | 08-Jun-2013 |
rmind | Split IPsec code in ip_input() and ip_forward() into the separate routines
ipsec4_input() and ipsec4_forward(). Tested by christos@.
|
| 1.221 | 08-Jun-2013 |
rmind | Split IPSec logic from ip_output() into a separate routine - ipsec4_output().
No change to the mechanism intended. Tested by christos@.
|
| 1.220 | 05-Jun-2013 |
christos | IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.219 | 04-Jun-2013 |
christos | PR/47886: Dr. Wolfgang Stukenbrock: IPSEC_NAT_T enabled kernels may access
outdated pointers and pass ESP data to UPD-sockets.
While here, simplify the code and remove the IPSEC_NAT_T option; always
compile nat-traversal in so that it does not bitrot.
|
| 1.218 | 02-Feb-2013 |
kefren | get rid of ip_len local variable. Use ntohs(ip->ip_len) like the rest
of the code in the two places this variable was used
|
| 1.217 | 25-Jun-2012 |
christos | branches: 1.217.2;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.216 | 22-Jun-2012 |
christos | PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.215 | 30-Apr-2012 |
rmind | - Replace some malloc(9) uses with kmem(9).
- G/C M_IPMOPTS, M_IPMADDR and M_BWMETER.
|
| 1.214 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.213 | 15-Feb-2012 |
drochner | fix for IPSEC tunnel + NAT-T + esp_frag:
Output packets larger than "esp_frag" are fragmented first
and then reinjected into ip_output for encapsulation
and transfer. The problem was that each packet got a new
ip_id value assigned, so that fragments couldn't be matched
by the receiver. Offset information was overwritten too.
approved by releng
|
| 1.212 | 31-Dec-2011 |
christos | - fix offsetof usage, and redundant defines
- kill pointer casts to 0
|
| 1.211 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.210 | 31-Oct-2011 |
yamt | branches: 1.210.2; 1.210.6;
redo ip_output.c rev.1.206 and 1.207 differently. PR/43664.
ok'ed by martin@
|
| 1.209 | 17-Jul-2011 |
joerg | Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.208 | 14-Apr-2011 |
yamt | after ip_input.c rev.1.285 and 1.286, restore kernel_lock for if_output.
|
| 1.207 | 09-Apr-2011 |
martin | PR kern/43664:
mlelstv pointed out that we sometimes may use checksums on loopback
interfaces. Make the test consistent with the code path selecting
the checksum operation before invoking fragmentation.
|
| 1.206 | 09-Apr-2011 |
martin | We do not do checksums on loopback interfaces, not even if fragmenting.
Fixes PR kern/43664.
|
| 1.205 | 17-Jul-2009 |
minskim | branches: 1.205.4; 1.205.6;
Add the IP_MINTTL socket option.
The IP_MINTTL option may be used on SOCK_STREAM sockets to discard
packets with a TTL lower than the option value. This can be used to
implement the Generalized TTL Security Mechanism (GTSM) according to
RFC 3682.
OK'ed by christos@.
|
| 1.204 | 16-Jul-2009 |
minskim | Add the IP_RECVTTL option support.
If the IP_RECVTTL option is enabled on a SOCK_DGRAM socket, the
recvmsg(2) call will return the TTL of the received datagram. The
msg_control field in the msghdr structure points to a buffer that
contains a cmsghdr structure followed by the TTL value.
Modeled after FreeBSD implementation.
|
| 1.203 | 01-Jul-2009 |
martin | From Wolfgang Stukenbrock in PR kern/41659: add missing splx().
|
| 1.202 | 06-May-2009 |
elad | Remove some usage of "priv" and "privileged" variables and instead pass
around credentials. Also push down kauth(9) calls closer to where the
operation is done.
Mailing list reference:
http://mail-index.netbsd.org/tech-net/2009/04/30/msg001270.html
|
| 1.201 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.200 | 12-Oct-2008 |
plunky | branches: 1.200.2; 1.200.4; 1.200.8; 1.200.10;
update ip_pcbopts() to use sockopt(9) API.
cleans up function and one small fix is that we now stop copying user
options to the mbuf when the _EOL is given, previously this function
would continue to copy options.
|
| 1.199 | 12-Oct-2008 |
plunky | do not sleep while allocating memory here as socket lock is held
|
| 1.198 | 16-Aug-2008 |
plunky | constify sockopt in the PRCO_SETOPT path
|
| 1.197 | 06-Aug-2008 |
plunky | Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.196 | 28-Apr-2008 |
martin | branches: 1.196.2; 1.196.6;
Remove clause 3 and 4 from TNF licenses
|
| 1.195 | 23-Apr-2008 |
thorpej | branches: 1.195.2;
Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.194 | 12-Apr-2008 |
thorpej | branches: 1.194.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.193 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.192 | 06-Feb-2008 |
matt | branches: 1.192.6;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.191 | 14-Jan-2008 |
dyoung | Use rtcache_validate() instead of rtcache_getrt(). Shorten staircase
in in_losing().
|
| 1.190 | 12-Jan-2008 |
dyoung | Good-bye, rtcache_check(). Call both rtcache_validate() and
rtcache_update(,1) instead of rtcache_check().
|
| 1.189 | 29-Dec-2007 |
degroote | Restore correctly the sp level in case of FAST_IPSEC + IPSEC_NAT_T
|
| 1.188 | 29-Dec-2007 |
degroote | Simplify the FAST_IPSEC output path
Only record an IPSEC_OUT_DONE tag when we have finished the processing
In ip{,6}_output, check this tag to know if we have already processed this
packet.
Remove some dead code (IPSEC_PENDING_TDB is not used in NetBSD)
Fix pr/36870
|
| 1.187 | 21-Dec-2007 |
matt | Add fix for ip_id information leakage. Since the leakage information is
primarily used with TCP SYN and RST packets and such packets are less than
the smallest sized packet that an IP stack is allowed to fragment, we simply
set ip_id to 0 for all packets 68 bytes or less.
|
| 1.186 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.185 | 28-Nov-2007 |
dyoung | branches: 1.185.2; 1.185.6;
Move IN_NEED_CHECKSUM() to in_offload.h for re-use.
|
| 1.184 | 19-Sep-2007 |
dyoung | branches: 1.184.6;
1) Introduce a new socket option, (SOL_SOCKET, SO_NOHEADER), that
tells a socket that it should both add a protocol header to tx'd
datagrams and remove the header from rx'd datagrams:
int onoff = 1, s = socket(...);
setsockopt(s, SOL_SOCKET, SO_NOHEADER, &onoff);
2) Add an implementation of (SOL_SOCKET, SO_NOHEADER) for raw IPv4
sockets.
3) Reorganize the protocols' pr_ctloutput implementations a bit.
Consistently return ENOPROTOOPT when an option is unsupported,
and EINVAL if a supported option's arguments are incorrect.
Reorganize the flow of code so that it's more clear how/when
options are passed down the stack until they are handled.
Shorten some pr_ctloutput staircases for readability.
4) Extract common mbuf code into subroutines, add new sockaddr
methods, and introduce a new subroutine, fsocreate(), for reuse
later; use it first in sys_socket():
struct mbuf *m_getsombuf(struct socket *so)
Create an mbuf and make its owner the socket `so'.
struct mbuf *m_intopt(struct socket *so, int val)
Create an mbuf, make its owner the socket `so', put the
int `val' into it, and set its length to sizeof(int).
int fsocreate(..., int *fd)
Create a socket, a la socreate(9), put the socket into the
given LWP's descriptor table, return the descriptor at `fd'
on success.
void *sockaddr_addr(struct sockaddr *sa, socklen_t *slenp)
const void *sockaddr_const_addr(const struct sockaddr *sa, socklen_t *slenp)
Extract a pointer to the address part of a sockaddr. Write
the length of the address part at `slenp', if `slenp' is
not NULL.
socklen_t sockaddr_getlen(const struct sockaddr *sa)
Return the length of a sockaddr. This just evaluates to
sa->sa_len. I only add this for consistency with code that
appears in a portable userland library that I am going to
import.
const struct sockaddr *sockaddr_any(const struct sockaddr *sa)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
const void *sockaddr_anyaddr(const struct sockaddr *sa, socklen_t *slenp)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
|
| 1.183 | 02-Sep-2007 |
dyoung | m_copym(..., 0, M_COPYALL, ...) -> m_copypacket(..., ...).
|
| 1.182 | 02-Sep-2007 |
dyoung | m_copy() was deprecated, apparently, long ago. m_copy(...) ->
m_copym(..., M_DONTWAIT).
|
| 1.181 | 28-Aug-2007 |
cube | Fix ipv4 multicast that could sometimes send packets with the wrong
Ethernet multicast address.
Reported by jmcneill@, fix discussed with dyoung@, _very_ light testing by
myself, some more money for my dealer of anxiolytics after reading
ip_output()'s twisted code maze.
|
| 1.180 | 02-May-2007 |
dyoung | branches: 1.180.2; 1.180.6; 1.180.8;
Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.179 | 04-Mar-2007 |
christos | branches: 1.179.2; 1.179.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.178 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.177 | 17-Feb-2007 |
dyoung | branches: 1.177.2;
Join lines.
|
| 1.176 | 29-Jan-2007 |
dyoung | bzero -> memset.
|
| 1.175 | 29-Jan-2007 |
dyoung | In ip_setmoptions(), don't leave a route cache (struct route) on
the stack if we exit with EADDRNOTAVAIL.
|
| 1.174 | 13-Jan-2007 |
joerg | Unconditionally zero and free iproute. Before IPsec tunnel packets e.g.
from ICMP could end up in leaking the reference in iproute, as
ipsec4_output would overwrite the ro pointer in state.
Tested by Juraj Hercek and supposed to fix PR kern/35273 and kern/35318.
|
| 1.173 | 08-Jan-2007 |
yamt | ip_output: reload ip_len after running pfil_run_hooks.
pf "fragment reassemble" rule can change it, at least.
|
| 1.172 | 04-Jan-2007 |
elad | Consistent usage of KAUTH_GENERIC_ISSUSER.
|
| 1.171 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.170 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.169 | 06-Dec-2006 |
dyoung | Remove stray curly brace. Thanks, yamt!
|
| 1.168 | 06-Dec-2006 |
dyoung | KNF.
|
| 1.167 | 25-Nov-2006 |
yamt | branches: 1.167.2;
move tso-by-software code to their own files. no functional changes.
|
| 1.166 | 13-Nov-2006 |
dyoung | Add a source-address selection policy mechanism to the kernel.
Also, add ioctls SIOCGIFADDRPREF/SIOCSIFADDRPREF to get/set preference
numbers for addresses. Make ifconfig(8) set/display preference
numbers.
To activate source-address selection policies in your kernel, add
'options IPSELSRC' to your kernel configuration.
Miscellaneous changes in support of source-address selection:
1 Factor out some common code, producing rt_replace_ifa().
2 Abbreviate a for-loop with TAILQ_FOREACH().
3 Add the predicates on IPv4 addresses IN_LINKLOCAL() and
IN_PRIVATE(), that are true for link-local unicast
(169.254/16) and RFC1918 private addresses, respectively.
Add the predicate IN_ANY_LOCAL() that is true for link-local
unicast and multicast.
4 Add IPv4-specific interface attach/detach routines,
in_domifattach and in_domifdetach, which build #ifdef
IPSELSRC.
See in_getifa(9) for a more thorough description of source-address
selection policy.
|
| 1.165 | 23-Jul-2006 |
ad | branches: 1.165.4; 1.165.6;
Use the LWP cached credentials where sane.
|
| 1.164 | 12-Jul-2006 |
tron | Remove test for M_CSUM_TSOv6 flag which is not (yet) defined in
NetBSD-current.
|
| 1.163 | 12-Jul-2006 |
tron | Add diagnostic checks for hardware-assisted checksum related flags in
the mbuf which supposed to get sent out:
- Complain in ip_output() if any of the IPv6 related flags are set.
- Complain in ip6_output() if any of the IPv4 related flags are set.
- Complain in both functions if the flags indicate that both a TCP and
UCP checksum should be calculated by the hardware.
|
| 1.162 | 15-May-2006 |
christos | branches: 1.162.4;
kauth fallout
|
| 1.161 | 14-May-2006 |
elad | integrate kauth.
|
| 1.160 | 23-Feb-2006 |
christos | branches: 1.160.2; 1.160.4; 1.160.6;
Handle IPSEC_NAT_T in the FAST_IPSEC case.
XXX: need to fix the FAST_IPSEC code now.
|
| 1.159 | 11-Dec-2005 |
christos | branches: 1.159.2; 1.159.4; 1.159.6;
merge ktrace-lwp.
|
| 1.158 | 19-Sep-2005 |
dyoung | People have to read this code, so I am removing the double-negative
tautology, #ifndef notdef, which is not only superfluous, but easily
misread as #ifdef notyet.
|
| 1.157 | 11-Sep-2005 |
seb | Replace plain 255 by MAXTTL.
|
| 1.156 | 11-Sep-2005 |
christos | Allow the multicast_ttl and the multicast_loop options to be set with both
u_char and u_int option variables. Original patch from seb.
|
| 1.155 | 18-Aug-2005 |
yamt | - introduce M_MOVE_PKTHDR and use it where appropriate.
intended to be mostly API compatible with openbsd/freebsd.
- remove a glue #define in netipsec/ipsec_osdep.h.
|
| 1.154 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.153 | 29-May-2005 |
christos | branches: 1.153.2;
- add const
- remove bogus casts
- avoid nested variables
|
| 1.152 | 18-Apr-2005 |
yamt | ip_output: handle the case M_CSUM_TSOv4 but !IFCAP_TSOv4.
|
| 1.151 | 18-Apr-2005 |
yamt | fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.150 | 07-Apr-2005 |
yamt | when doing TSO, avoid to use duplicated ip_id heavily.
XXX ip_randomid
|
| 1.149 | 11-Mar-2005 |
matt | branches: 1.149.2;
Set ip_len to 0 in the wm driver when TSO is being used.
|
| 1.148 | 10-Mar-2005 |
thorpej | In ip_fragment():
- Use the correct IP header length variable for other-than-first packets.
- Remove redundant setting of the original IP header length in the first
packet's csum_data. (It's already set before ip_fragment() is called
in 1.147.)
|
| 1.147 | 09-Mar-2005 |
matt | Move all the hardware-assisted checksum/segment offload code together.
|
| 1.146 | 06-Mar-2005 |
matt | Add IPv4/TCP hooks for TCP Segment Offload on transmit.
|
| 1.145 | 05-Mar-2005 |
briggs | Fix checksum offload for fragmented packets. From John Heasley
on gnats-bugs in PR kern/29544.
Tested with an NFS client using default rwsize on an NFS server
with wm(4) interface configured IP4CSUM,TCP4CSUM,UDP4CSUM.
Prior revision required the server to have checksum offload disabled.
|
| 1.144 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.143 | 18-Feb-2005 |
heas | My last change for pseudo-header checksums was flawed. The pseudo-header
checksum is always in the L4 header by the time we get to this point. It
was occasionally not there due to a bug in tcp_respond, which has since
been fixed.
So, instead just stash the length of the L3 header in the high 16 bits of
csum_data.
|
| 1.142 | 12-Feb-2005 |
heas | For controllers (eg: hme & gem) that can only perform linear hardware checksums
(from an offset to the end of the packet), the pseudo-header checksum must be
calculated by software. So, provide it in the TCP/UDP header when
M_CSUM_NO_PSEUDOHDR is set in the interface's if_csum_flags_tx.
The start offset, the end of the IP header, is also provided in the high 16
bits of pkthdr.csum_data. Such that the driver need not examine the packet
at all.
XXX At the request of Jonathan Stone, note that sharing of if_csum_flags_tx &
pkthdr.csum_flags for checksum quirks should be re-evaluated.
|
| 1.141 | 12-Feb-2005 |
manu | Add support for IPsec Network Address Translator traversal (NAT-T), as
described by RFC 3947 and 3948.
|
| 1.140 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.139 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.138 | 15-Dec-2004 |
thorpej | branches: 1.138.2; 1.138.4;
Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.137 | 04-Dec-2004 |
peter | Convert lo(4) to a clonable device.
This also removes the loif array and changes all code to use the new
lo0ifp pointer which points to the lo0 ifnet structure.
Approved by christos.
|
| 1.136 | 06-Oct-2004 |
thorpej | Slight simplification to IFA_STATS handling.
|
| 1.135 | 04-Sep-2004 |
manu | IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.134 | 06-Jul-2004 |
minoura | Remove broken code for now: getsockopt(s, IPPROTO_IP, IP_IPSEC_POLICY,...).
It returned EINVAL, now returns ENOPROTOOPT.
Ok'd by itojun.
|
| 1.133 | 01-Jun-2004 |
itojun | update mtu value if outgoing interface changes with ipsec ops
(draft-touch-vpn case only?) iij seil team
|
| 1.132 | 18-May-2004 |
christos | Fix buffer overrun in in_pcbopts() (FreeBSD PR/66386)
|
| 1.131 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.130 | 02-Mar-2004 |
thorpej | Use the new IPSEC_PCB_SKIP_IPSEC() to bypass a socket policy lookup
when possible. This shaves several cycles from the output path for
non-IPsec connections, even if the policy is cached in the PCB.
|
| 1.129 | 10-Dec-2003 |
itojun | use if_indexlim (instead of if_index) and ifindex2ifnet[x] != NULL
to check if interface exists, as (1) if_index has different meaning
(2) ifindex2ifnet could become NULL when interface gets destroyed,
since when we have introduced dynamically-created interfaces. from kame
|
| 1.128 | 19-Nov-2003 |
jonathan | Patch back support for (badly) randomized IP ids, by request:
* Include "opt_inet.h" everywhere IP-ids are generated with ip_newid(),
so the RANDOM_IP_ID option is visible. Also in ip_id(), to ensure
the prototype for ip_randomid() is made visible.
* Add new sysctl to enable randomized IP-ids, provided the kernel was
configured with RANDOM_IP_ID. (The sysctl defaults to zero, and is
a read-only zero if RANDOM_IP_ID is not configured).
Note that the implementation of randomized IP ids is still defective,
and should not be enabled at all (even if configured) without
very careful deliberation. Caveat emptor.
|
| 1.127 | 17-Nov-2003 |
jonathan | Revert the (default) ip_id algorithm to the pre-randomid algorithm,
due to demonstrated low-period repeated IDs from the randomized IP_id
code. Consensus is that the low-period repetition (much less than
2^15) is not suitable for general-purpose use.
Allocators of new IPv4 IDs should now call the function ip_newid().
Randomized IP_ids is now a config-time option, "options RANDOM_IP_ID".
ip_newid() can use ip_random-id()_IP_ID if and only if configured
with RANDOM_IP_ID. A sysctl knob should be provided.
This API may be reworked in the near future to support linear ip_id
counters per (src,dst) IP-address pair.
|
| 1.126 | 17-Oct-2003 |
enami | Increment stats when packet is dropped since there is no room
to put all fragments in the interfaces's send queue. Some large
UDP packets are dropped here and administrator may want to bump ifqmaxlen.
|
| 1.125 | 14-Oct-2003 |
itojun | more correction to ip_fragment; free mbuf correctly if ENOBUFS is raised
during fragmenting.
|
| 1.124 | 14-Oct-2003 |
itojun | avoid mbuf leak on ip_fragment(); obey 4.4bsd mbuf passing rule (mbuf passed
to a function must be freed by the called function on error).
pointed out by enami
|
| 1.123 | 03-Oct-2003 |
itojun | when dropping M_PKTHDR, need to free m_tag associated with it.
|
| 1.122 | 01-Oct-2003 |
itojun | correct ip_fragment() wrt ip->ip_off handling.
do not send out incomplete fragment due to ENOBUFS (behavior change from 4.4BSD)
|
| 1.121 | 19-Sep-2003 |
jonathan | Fast-ipsec can call ip_output() with a null 'struct socket *so'
argument. So check so is non-NULL before doing the pointer-chasing
dance to find the PCB. (Unless and until we rework fast-ipsec and
KAME, to pass a struct in_pcbhdr * instead of the struct socket *).
|
| 1.120 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.119 | 27-Aug-2003 |
itojun | don't intiialize m by m0, m0 is not initialized (by introduction of ip_fragment)
|
| 1.118 | 23-Aug-2003 |
itojun | need sys/domain.h for FAST_IPSEC case; jonathan
|
| 1.117 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.116 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.115 | 22-Aug-2003 |
jonathan | Change KAME code for ip_output()/ip6_output() to obtain struct socket*
from the explicit inpcb*/in6pcb* argument. set_socket() becomes redundant.
|
| 1.114 | 19-Aug-2003 |
itojun | remove unneeded #ifdef __NetBSD__
|
| 1.113 | 19-Aug-2003 |
itojun | make ip_fragment public (it is for coming PF integration)
|
| 1.112 | 19-Aug-2003 |
christos | make ip_fragment static and add prototype.
|
| 1.111 | 19-Aug-2003 |
itojun | correct ip_multicast_if fix to always set ifp (tnx Shiva)
|
| 1.110 | 18-Aug-2003 |
itojun | fix problem we can't drop membership on !IFF_UP interface.
reported by Shiva Shenoy
while we're here, fix another problem when the same interface address is
assigned to !IFF_MULTICAST and IFF_MULTICAST interface. if ip_multicast_if()
returns the first one, join/leave will fail, which is not an desired effect.
|
| 1.109 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.108 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.107 | 30-Jun-2003 |
itojun | branches: 1.107.2;
freebsd code somehow crept in
|
| 1.106 | 30-Jun-2003 |
itojun | after pfil_run_hooks, need to fix hlen as well
|
| 1.105 | 26-Jun-2003 |
itojun | tabify
|
| 1.104 | 26-May-2003 |
yamt | - don't pass mbufs with M_CSUM_* flags which isn't supported by the interface
to if_output.
- offload ip-checksumming for each fragmented packets as well.
|
| 1.103 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.102 | 17-Sep-2002 |
darrenr | From FreeBSD (1.164) courtesy of Maxim Konovalov:
"In rare cases when there is no room for ip options ip_insertoptions()
can fail and corrupt a header length. Initialize len and check what
ip_insertoptions() returns."
|
| 1.101 | 11-Sep-2002 |
itojun | KNF - return is not a function. sync w/kame.
|
| 1.100 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.99 | 24-Jun-2002 |
itojun | set ia as well
|
| 1.98 | 24-Jun-2002 |
itojun | do not consult routing table under the following condition:
- the destination is IPv4 multicast or 255.255.255.255, and
- outgoing interface is specified via socket option
this simplifies operation of routed
(no longer reqiure 224.0.0.0/4 to be set up)
|
| 1.97 | 09-Jun-2002 |
itojun | whitespace
|
| 1.96 | 31-May-2002 |
itojun | since if_mtu is u_long, use u_long for mtu.
|
| 1.95 | 07-Feb-2002 |
thorpej | branches: 1.95.8; 1.95.10;
IFF_POINTTOPOINT interfaces can also transmit packets to broadcast
destinations.
|
| 1.94 | 06-Feb-2002 |
thorpej | ip_mloopback(): process the delayed checksum on the copy, not
the original mbuf.
|
| 1.93 | 31-Jan-2002 |
itojun | correct bad ip checksum on multicast loopback packet. PR14597
|
| 1.92 | 22-Jan-2002 |
itojun | make sure to check address family on route cache. with IPv4 mapped
address we can see both AF_INET/INET6.
|
| 1.91 | 08-Jan-2002 |
itojun | don't panic when there's no interface address exist for the specified multicast
outgoing interface (ia == NULL after IFP_TO_IA).
historic behavior (up to revision 1.43) was to use 0.0.0.0 as source address,
but it seems like a mistake according to RFC1112/1122.
|
| 1.90 | 21-Nov-2001 |
itojun | update outgoing ifp, only if tunnel mode ipsec is used. this is to
honor IP_MULTICAST_IF setsockopt on ipsec-over-multicast. sync with kame
|
| 1.89 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.88 | 17-Sep-2001 |
thorpej | Split the pre-computed ifnet checksum flags into Tx and Rx directions.
Add capabilities bits that indicate an interface can only perform
in-bound TCPv4 or UDPv4 checksums. There is at least one Gig-E chip
for which this is true (Level One LXT-1001), and this is also the
case for the Intel i82559 10/100 Ethernet chips.
|
| 1.87 | 11-Aug-2001 |
yamt | branches: 1.87.2;
fix cksum error of udp and tcp packet with ip options
|
| 1.86 | 02-Jun-2001 |
thorpej | branches: 1.86.2;
Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.85 | 26-May-2001 |
ragge | Remove one #ifdef vax, bugfix another. Should probably be #ifdef i386 also.
|
| 1.84 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.83 | 27-Feb-2001 |
itojun | branches: 1.83.2;
remove obsolete #if 0'ed section
(IPsec and DF bit interaction - the code was incorrect anyways)
|
| 1.82 | 24-Jan-2001 |
itojun | - record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.81 | 13-Jan-2001 |
itojun | allow IP_MULTICAST_IF and IP_ADD/DROP_MEMBERSHIP to specify interface
by interface index. if the interface address specified is in 0.0.0.0/8
it will be considered as interface index in network byteorder.
getsockopt(IP_MULTICAST_IF) preserves old behavior if
setsockopt(IP_MULTICAST_IF) was done with interface address, and
returns interface index if setsockopt(IP_MULTICAST_IF) was done with
interface index (again using the form in 0.0.0.0/8).
Suggested by Dave Thaler, based on RIPv2 MIB spec (RFC1724 section 3.3).
http://mail-index.netbsd.org/tech-net/2001/01/13/0003.html
|
| 1.80 | 13-Jan-2001 |
itojun | on getsockopt(IP_IPSEC_POLICY), make sure to initialize len
|
| 1.79 | 11-Nov-2000 |
thorpej | Actually, our local ip_off variable isn't needed.
|
| 1.78 | 11-Nov-2000 |
thorpej | Restructure the PFIL_HOOKS mechanism a bit:
- All packets are passed to PFIL_HOOKS as they come off the wire, i.e.
fields in protocol headers in network order, etc.
- Allow for multiple hooks to be registered, using a "key" and a "dlt".
The "dlt" is a BPF data link type, indicating what type of header is
present.
- INET and INET6 register with key == AF_INET or AF_INET6, and
dlt == DLT_RAW.
- PFIL_HOOKS now take an argument for the filter hook, and mbuf **,
an ifnet *, and a direction (PFIL_IN or PFIL_OUT), thus making them
less IP (really, IP Filter) centric.
Maintain compatibility with IP Filter by adding wrapper functions for
IP Filter.
|
| 1.77 | 23-Oct-2000 |
itojun | fix IFA_STATS.
- use hashed in_ifaddr lookup.
- correct endianness.
|
| 1.76 | 17-Oct-2000 |
thorpej | Add an IP_MTUDISC flag to the flags that can be passed to
ip_output(). This flag, if set, causes ip_output() to set
DF in the IP header if the MTU in the route is not locked.
This allows a bunch of redundant code, which I was never
really all that happy about adding in the first place, to
be eliminated.
Inspired by a similar change made by provos@openbsd.org when
he integrated NetBSD's Path MTU Discovery code into OpenBSD.
|
| 1.75 | 28-Jun-2000 |
mrg | remove include of <vm/vm.h>
|
| 1.74 | 10-May-2000 |
itojun | branches: 1.74.4;
add missing boundary checks to ip options processing.
correct timestamp option validation (len and ptr upper/lower bound
based on RFC791).
fill "pointer" field for parameter problem in timestamp option processing.
|
| 1.73 | 13-Apr-2000 |
is | Copy M_BCAST and M_MCAST flags when fragmenting a packet (else
Multicast packets won't be send to the correct link layer address
by the interface driver).
By Artur Grabowski, PR 9772.
|
| 1.72 | 31-Mar-2000 |
jdolecek | Since last duplicate prototype cleanup, we need to include
<netinet/ip_mroute.h> to get ip_mforward() prototype if MROUTING
is defined.
|
| 1.71 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.70 | 22-Mar-2000 |
itojun | tabify a line.
|
| 1.69 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.68 | 20-Feb-2000 |
darrenr | pass "struct pfil_head *" to pfil_add_hook and pfil_remove hook rather
than "struct protosw *".
|
| 1.67 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.66 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.65 | 20-Dec-1999 |
itojun | avoid shared cluster mbuf overwrite on multicast packet loopback.
(bsdi and freebsd fixed this a long time ago...)
PR: 9020
From: pavlin@catarina.usc.edu
|
| 1.64 | 13-Dec-1999 |
is | Handle packets to 255.255.255.255 like multicast packets. Fixes PR 7682 by
Darren Reed.
|
| 1.63 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.62 | 09-Jul-1999 |
thorpej | branches: 1.62.2; 1.62.8;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.61 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.60 | 07-Jun-1999 |
mrg | oops. move sendit: above the PFIL_HOOKS so that multicast traffic is filtered. from darren reed.
|
| 1.59 | 04-May-1999 |
hwr | Don't let packets with a Class-D source address escape the host.
Fixes second half of kern/7003 by Jonathan Stone <jonathan@DSG.Stanford.EDU>.
|
| 1.58 | 27-Mar-1999 |
aidan | branches: 1.58.2; 1.58.4; 1.58.6;
Added per-addr input/output statistics. Currently just support netatalk
and netinet, currently only tested under netinet.
Disabled by default, enabled by compiling the kernel with option
IFA_STATS. Enabling this feature seems to make the ip_output function
take 13% longer than before, which should be OK for people that need
this feature.
|
| 1.57 | 12-Mar-1999 |
perry | exterminate ovbcopy. patches provided by Erik Bertelsen, pr-7145
|
| 1.56 | 19-Jan-1999 |
mycroft | There's just no plausible reason to byte-swap ip_id internally. It's opaque.
|
| 1.55 | 11-Jan-1999 |
thorpej | Fix byte order and ip_len inconsistencies in ICMP reply code. Also, fix
some formatting and HTONS(foo) vs. foo = htons(foo) inconsistencies.
PR #6602, Darren Reed.
|
| 1.54 | 19-Dec-1998 |
thorpej | Reverse the copyright-notice-swap. It went against existing practice.
|
| 1.53 | 26-Oct-1998 |
ws | branches: 1.53.4;
Fix a buglet when looking up an interface for multicast:
Zero out the routing structure before calling the route lookup code
in order to correctly match addresses.
|
| 1.52 | 20-Oct-1998 |
matt | vax -> __vax__ (and mips to __mips__ in ultrix_misc.c)
|
| 1.51 | 30-Sep-1998 |
tls | Switch order of TNF and UCB copyrights so UCB copyright is first; this seems more appropriate since UCB wrote the original code, after all.
|
| 1.50 | 09-Aug-1998 |
mrg | defopt PFIL_HOOKS.
|
| 1.49 | 17-Jul-1998 |
sommerfe | Fix PR5508: ipfil cut-through forwarding causes panic
|
| 1.48 | 28-Apr-1998 |
matt | Only transmit fragments if the send queue of interface can actually hold
all of the fragments. Use the mtu of route in preference of the MTU of the
interface when doing fragmentation decisions. (ie. Fragment to the path
mtu if it is available).
|
| 1.47 | 24-Mar-1998 |
kml | Ensure that we take the IP option length into account when we calculate
the effective maximum send size for TCP. ip_optlen() and tcp_optlen()
should probably be inlined for efficiency.
|
| 1.46 | 19-Mar-1998 |
mrg | convert pfil(9) in and out lists from <sys/queue.h> LISTs to TAILQs, and
change pfil_add_hook to put output filters at the tail of the queue,
while continuing to place input filters at the head of the queue. update
the two users of these functions, and document these changes.
fixes PR#4593.
|
| 1.45 | 15-Feb-1998 |
tls | Add correct copyright notice for IP address hash change. This code is donated to TNF by the original copyright holder, Panix.
|
| 1.44 | 13-Feb-1998 |
tls | Change list of interface IP addresses to a hash. Improves performance on hosts with a large number of IP addresses significantly.
|
| 1.43 | 13-Feb-1998 |
kleink | Fix variable declarations: register -> register int.
|
| 1.42 | 12-Jan-1998 |
scottr | Use option header file for MROUTING
|
| 1.41 | 07-Jan-1998 |
lukem | add the following, derived from FreeBSD:
* IP_PORTRANGE socket option, which controls how the ephemeral ports
are allocated. it takes the following settings:
IP_PORTRANGE_DEFAULT use anonportmin (49152) -> anonportmax (65535)
IP_PORTRANGE_HIGH as IP_PORTRANGE_DEFAULT (retained for FreeBSD
compat reasons, where these are separate)
IP_PORTRANGE_LOW use 600 -> 1023. only works if uid==0.
* in_pcb flag INP_ANONPORT. set if port was allocated ephmerally
|
| 1.40 | 14-Oct-1997 |
matt | branches: 1.40.2;
Add support for returning maximum supported MTU when ip_output fails with
EMSGSIZE.
|
| 1.39 | 15-Apr-1997 |
christos | branches: 1.39.4;
Move the mtod calls *after* we've made sure that the packet has passed the
filter successfully. Otherwise it can be NULL if the filter blocked it,
and we die. How did this ever work?
|
| 1.38 | 18-Feb-1997 |
mrg | pseudo-device ipfilter brings in PFIL_HOOKS.
|
| 1.37 | 11-Jan-1997 |
thorpej | branches: 1.37.4;
Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.36 | 20-Dec-1996 |
mrg | always reassign ip after calling function.
|
| 1.35 | 20-Dec-1996 |
mrg | in pfil_hooks: always reassign ip after calling hook.
|
| 1.34 | 22-Oct-1996 |
veego | Fix a panic from the pfil_hooks.
|
| 1.33 | 11-Oct-1996 |
is | Fix a mbuf leak in ip_output().
Scenario: If ip_insertoptions() prepends a new mbuf to the chain, the
bad: label's m_freem(m0) still would free only the original mbuf chain
if the transmission failed for, e.g., no route to host; resulting in
one lost mbuf per failed packet. (The original posting included a
demonstration program).
Original report of this bug was by jinmei@isl.rdc.toshiba.co.jp
(JINMEI Tatuya) on comp.bugs.4bsd.
|
| 1.32 | 14-Sep-1996 |
mrg | move the packet filter hooks in to a saner location. while i'm here, rename
PACKET_FILTER to PFIL_HOOKS.
|
| 1.31 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.30 | 06-Sep-1996 |
mrg | add packet filter interface code. see pfil(9) for more details. you
need the PACKET_FILTER option to enable this code. currently, ipfilter
version 3.1.1-beta has been converted to use this new interface.
|
| 1.29 | 26-Feb-1996 |
mrg | branches: 1.29.4;
two more local addr changes, all done differently now (idea from charles)
|
| 1.28 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.27 | 01-Jul-1995 |
cgd | null mbuf pointer could cause system crash; avoid it. From
Torsten Duwe <duwe@immd4.informatik.uni-erlangen.de>.
|
| 1.26 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.25 | 04-Jun-1995 |
mycroft | Don't cast things unnecessarily.
|
| 1.24 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.23 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.22 | 15-May-1995 |
cgd | simplify ip_output() out-of-memory condition slightly, and style nits.
|
| 1.21 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.20 | 11-Apr-1995 |
mycroft | Remove some explicit references to loif.
|
| 1.19 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.18 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.17 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.16 | 19-Jan-1994 |
brezak | Fix arguments to ip_getmoptions.
|
| 1.15 | 18-Jan-1994 |
brezak | Fix some prototype detected warnings/errors.
|
| 1.14 | 18-Jan-1994 |
brezak | Patch for ip-multicast bugs from mccanne@ee.lbl.gov (Steven McCanne)
|
| 1.13 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.12 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.11 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.10 | 07-Jan-1994 |
cgd | kill COMPAT_OLDSOCKOPT
|
| 1.9 | 06-Jan-1994 |
ws | Apparently noone ever tested the COMPAT_OLDSOCKOPT flag...
|
| 1.8 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.7 | 06-Dec-1993 |
cgd | oops; fix that last...
|
| 1.6 | 06-Dec-1993 |
cgd | the ugliest compatibility hack i think i've ever seen...
define COMPAT_OLDSOCKOPT to get new kernels to work with the
old args to [sg]sockopt. this is going to go away "soon".
note that this option only has effect if MULTICAST is not defined.
|
| 1.5 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.4 | 05-Nov-1993 |
cgd | fix from david greenman, davidg@freefall.cdrom.com:
fixed bug where large amounts of unidirectional UDP traffic would fill
the interface output queue and further udp packets would be fragmented
and only partially sent - keeping the output queue full and jamming the
network, but not actually getting any real work done (because you can't
send just 'part' of a udp packet - if you fragment it, you must send
the whole thing). The fix involves adding a check to make sure that the
output queue has sufficient space for all of the fragments.
|
| 1.3 | 22-May-1993 |
cgd | branches: 1.3.4;
add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.3.4.2 | 06-Nov-1993 |
mycroft | Merge changes from trunk.
|
| 1.3.4.1 | 16-Oct-1993 |
mycroft | Nuke references to machine/mtpr.h.
|
| 1.29.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Fix a mbuf leak when fragmentation fails due to lack of memory.
|
| 1.37.4.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.39.4.1 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.40.2.4 | 29-Oct-1998 |
cgd | pull up rev 1.53 from trunk (ws)
|
| 1.40.2.3 | 01-Oct-1998 |
cgd | pull up revisions 1.44-1.45, 1.51 (via patch) from trunk. (tls)
|
| 1.40.2.2 | 22-Jul-1998 |
mellon | Pull up 1.46 and 1.49 (veego)
|
| 1.40.2.1 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.53.4.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.58.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.58.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.58.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.58.4.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.58.4.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.58.4.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.58.2.3 | 30-Apr-2000 |
he | Pull up revision 1.73 (requested by is):
Pass M_BCAST and M_MCAST flags to fragments. Fixes PR#9772.
|
| 1.58.2.2 | 20-Dec-1999 |
he | Pull up revision 1.65 (requested by itojun):
Avoid panic caused by shared cluster mbuf overwrite on multicast
packet loopback for packets with certain sizes. Fixes PR#9020.
|
| 1.58.2.1 | 22-Jun-1999 |
perry | pullup 1.59->1.60 (mrg): ipfilter should filter multicast traffic...
|
| 1.62.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.62.2.6 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.62.2.5 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.62.2.4 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.62.2.3 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.62.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.62.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.74.4.4 | 04-Aug-2003 |
msaitoh | Pull up revision 1.106-1.107 (requested by itojun in ticket #53):
after pfil_run_hooks, need to fix hlen as well.
freebsd code somehow crept in.
|
| 1.74.4.3 | 15-Dec-2002 |
he | Pull up revision 1.102 (requested by darrenr):
Initialize len and check what ip_insertoptions() returns.
In some rare cases there might not be sufficient room for
the options.
|
| 1.74.4.2 | 14-Jan-2002 |
he | Pull up revision 1.91 (requested by itojun):
Avoid kernel panic on IPv4 multicast packet transmission if there
is no IPv4 address assigned to the specified outgoing interface.
|
| 1.74.4.1 | 06-Apr-2001 |
he | Pull up revision 1.82 (via patch, requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.83.2.16 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.83.2.15 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.83.2.14 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.83.2.13 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.83.2.12 | 12-Jul-2002 |
nathanw | No longer need to pull in lwp.h; proc.h pulls it in for us.
|
| 1.83.2.11 | 24-Jun-2002 |
nathanw | Curproc->curlwp renaming.
Change uses of "curproc->l_proc" back to "curproc", which is more like the
original use. Bare uses of "curproc" are now "curlwp".
"curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL)
so that it is always safe to reference curproc (*de*referencing curproc
is another story, but that's always been true).
|
| 1.83.2.10 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.83.2.9 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.83.2.8 | 11-Jan-2002 |
nathanw | More catchup.
|
| 1.83.2.7 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.83.2.6 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.83.2.5 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.83.2.4 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.83.2.3 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.83.2.2 | 13-Mar-2001 |
nathanw | Be more careful not to dereference curproc when there might not be
a process context.
|
| 1.83.2.1 | 05-Mar-2001 |
nathanw | Initial commit of scheduler activations and lightweight process support.
|
| 1.86.2.7 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.86.2.6 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.86.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.86.2.4 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.86.2.3 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.86.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.86.2.1 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.87.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.95.10.3 | 30-Jun-2003 |
grant | Pull up revisions 1.106-1.107 (requested by itojun in ticket #1358):
after pfil_run_hooks, need to fix hlen as well
freebsd code somehow crept in
|
| 1.95.10.2 | 01-Nov-2002 |
tron | Pull up revision 1.98-1.99 (requested by itojun in ticket #356):
do not consult routing table under the following condition:
- - the destination is IPv4 multicast or 255.255.255.255, and
- - outgoing interface is specified via socket option
this simplifies operation of routed
(no longer require 224.0.0.0/4 to be set up)
|
| 1.95.10.1 | 30-Sep-2002 |
lukem | Pull up revision 1.102 (requested by darrenr in ticket #842):
From FreeBSD (1.164) courtesy of Maxim Konovalov:
"In rare cases when there is no room for ip options ip_insertoptions()
can fail and corrupt a header length. Initialize len and check what
ip_insertoptions() returns."
|
| 1.95.8.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.95.8.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.95.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.107.2.11 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.107.2.10 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.107.2.9 | 08-Mar-2005 |
skrll | Sync with HEAD.
|
| 1.107.2.8 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.107.2.7 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.107.2.6 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.107.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.107.2.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.107.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.107.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.107.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.138.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.138.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.138.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.149.2.5 | 31-Mar-2007 |
bouyer | Pull up following revision(s) (requested by joerg in ticket #1734):
sys/netinet/ip_output.c: revision 1.167.2.2
Unconditionally zero and free iproute. Before IPsec tunnel packets e.g.
from ICMP could end up in leaking the reference in iproute, as
ipsec4_output would overwrite the ro pointer in state.
Tested by Juraj Hercek and supposed to fix PR kern/35273 and kern/35318.
|
| 1.149.2.4 | 28-Jan-2007 |
tron | Pull up following revision(s) (requested by yamt in ticket #1656):
sys/netinet/ip_output.c: revision 1.173
ip_output: reload ip_len after running pfil_run_hooks.
pf "fragment reassemble" rule can change it, at least.
|
| 1.149.2.3 | 21-Oct-2005 |
riz | branches: 1.149.2.3.2; 1.149.2.3.4;
Pull up following revision(s) (requested by seb in ticket #903):
sys/netinet/ip_output.c: revisions 1.156 - 1.157
Allow the multicast_ttl and the multicast_loop options to be set with both
u_char and u_int option variables. Original patch from seb.
|
| 1.149.2.2 | 06-May-2005 |
tron | Pull up revision 1.151 (requested by yamt in ticket #251):
fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.149.2.1 | 13-Apr-2005 |
tron | Pull up revision 1.150 (requested by yamt in ticket #145):
when doing TSO, avoid to use duplicated ip_id heavily.
XXX ip_randomid
|
| 1.149.2.3.4.1 | 28-Jan-2007 |
tron | Pull up following revision(s) (requested by yamt in ticket #1656):
sys/netinet/ip_output.c: revision 1.173
ip_output: reload ip_len after running pfil_run_hooks.
pf "fragment reassemble" rule can change it, at least.
|
| 1.149.2.3.2.1 | 28-Jan-2007 |
tron | Pull up following revision(s) (requested by yamt in ticket #1656):
sys/netinet/ip_output.c: revision 1.173
ip_output: reload ip_len after running pfil_run_hooks.
pf "fragment reassemble" rule can change it, at least.
|
| 1.153.2.8 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.153.2.7 | 21-Jan-2008 |
yamt | sync with head
|
| 1.153.2.6 | 07-Dec-2007 |
yamt | sync with head
|
| 1.153.2.5 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.153.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.153.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.153.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.153.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.159.6.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.159.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.159.4.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.159.4.1 | 07-Feb-2006 |
rpaulo | sotoinpcb_hdr -> sotoinpcb.
|
| 1.159.2.1 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.160.6.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.160.4.2 | 10-Mar-2006 |
elad | generic_authorize() -> kauth_authorize_generic().
|
| 1.160.4.1 | 08-Mar-2006 |
elad | Adapt to kernel authorization KPI.
|
| 1.160.2.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.160.2.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.162.4.1 | 13-Jul-2006 |
gdamore | Merge from HEAD.
|
| 1.165.6.2 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.165.6.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.165.4.3 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.165.4.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.165.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.167.2.2 | 28-Mar-2007 |
jdc | Pull up revision 1.174 (requested by joerg in ticket #524).
Unconditionally zero and free iproute. Before IPsec tunnel packets e.g.
from ICMP could end up in leaking the reference in iproute, as
ipsec4_output would overwrite the ro pointer in state.
Tested by Juraj Hercek and supposed to fix PR kern/35273 and kern/35318.
|
| 1.167.2.1 | 18-Jan-2007 |
tron | Pull up following revision(s) (requested by yamt in ticket #361):
sys/netinet/ip_output.c: revision 1.173
ip_output: reload ip_len after running pfil_run_hooks.
pf "fragment reassemble" rule can change it, at least.
|
| 1.177.2.4 | 07-May-2007 |
yamt | sync with head.
|
| 1.177.2.3 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.177.2.2 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.177.2.1 | 17-Feb-2007 |
yamt | file ip_output.c was added on branch yamt-idlelwp on 2007-02-27 16:54:56 +0000
|
| 1.179.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.179.2.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.179.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.180.8.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.180.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.180.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.180.6.3 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.180.6.2 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.180.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.180.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.184.6.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.184.6.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.184.6.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.185.6.2 | 19-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.185.6.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.185.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.192.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.192.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.192.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.194.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.195.2.5 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.195.2.4 | 18-Jul-2009 |
yamt | sync with head.
|
| 1.195.2.3 | 16-May-2009 |
yamt | sync with head
|
| 1.195.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.195.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.196.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.196.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.200.10.1 | 09-Jul-2009 |
snj | branches: 1.200.10.1.2;
Pull up following revision(s) (requested by martin in ticket #847):
sys/netinet/ip_output.c: revision 1.203
From Wolfgang Stukenbrock in PR kern/41659: add missing splx().
|
| 1.200.10.1.2.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.200.8.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.200.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.200.4.1 | 09-Jul-2009 |
snj | Pull up following revision(s) (requested by martin in ticket #847):
sys/netinet/ip_output.c: revision 1.203
From Wolfgang Stukenbrock in PR kern/41659: add missing splx().
|
| 1.200.2.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.205.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.205.4.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.210.6.3 | 02-Jun-2012 |
mrg | sync to latest -current.
|
| 1.210.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.210.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.210.2.4 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.210.2.3 | 30-Oct-2012 |
yamt | sync with head
|
| 1.210.2.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.210.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.217.2.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.217.2.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.217.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.217.2.1 | 25-Feb-2013 |
tls | resync with head
|
| 1.223.2.3 | 17-Oct-2013 |
rmind | Eliminate some of the splsoftnet() calls, misc clean up.
|
| 1.223.2.2 | 28-Aug-2013 |
rmind | sync with head
|
| 1.223.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.224.4.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.230.2.1 | 01-Dec-2014 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #277):
sys/netinet/ip_output.c: revision 1.233
Call looutput with holding KERNEL_LOCK
This fixes diagnostic assertion "KERNEL_LOCKED_P()" in if_loop.c.
PR kern/49410
|
| 1.233.2.10 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.233.2.9 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.233.2.8 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.233.2.7 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.233.2.6 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.233.2.5 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.233.2.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.233.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.233.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.233.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.259.2.4 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.259.2.3 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.259.2.2 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.259.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.267.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.276.4.2 | 19-May-2017 |
pgoyette | Resolve conflicts from previous merge (all resulting from $NetBSD
keywork expansion)
|
| 1.276.4.1 | 11-May-2017 |
pgoyette | Sync with HEAD
|
| 1.279.2.7 | 18-Mar-2018 |
martin | Pull up following revision(s) (requested by tih in ticket #639):
sys/kern/uipc_socket.c: revision 1.258
sys/kern/uipc_socket.c: revision 1.259
sys/netinet/ip_input.c: revision 1.364 (via patch)
sys/netinet/ip_output.c: revision 1.289
sys/netinet/in.h: revision 1.102
sys/netinet/in_pcb.c: revision 1.181
share/man/man9/sockopt.9: revision 1.11
sys/netinet/in_pcb.h: revision 1.65
sys/sys/socketvar.h: revision 1.146
sys/kern/uipc_syscalls.c: revision 1.189
sys/netinet/ip_output.c: revision 1.290
share/man/man4/ip.4: revision 1.41
share/man/man4/ip.4: revision 1.42
sys/kern/uipc_syscalls.c: revision 1.190
pass valsize for getsockopt like we do for setsockopt
make sure that we have enough space, don't require the exact size
(Tom Ivar Helbekkmo)
1) "#define ipi_spec_dst ipi_addr" in <netinet/in.h>
2) Change the IP_RECVPKTINFO option to control the generation of
IP_PKTINFO control messages, the way it's done in Solaris.
3) Remove the superfluous IP_RECVPKTINFO control message.
4) Change the IP_PKTINFO option to do different things depending on
the parameter it's supplied with:
- If it's sizeof(int), assume it's being used as in Linux:
- If it's non-zero, turn on the IP_RECVPKTINFO option.
- If it's zero, turn off the IP_RECVPKTINFO option.
- If it's sizeof(struct in_pktinfo), assume it's being used as in
Solaris, to set a default for the source interface and/or
source address for outgoing packets on the socket.
5) Return what Linux or Solaris compatible code expects, depending
on data size, and just added a fallback to a Linux (and current NetBSD)
compatible value if the size is unknown (as it is now), or,
in the future, if the calling application specifies a receiving
buffer that doesn't match either data item.
From: Tom Ivar Helbekkmo
new sentence-new line
Remove comment now that the getsockopt code passes the size.
Add a new sockopt member to keep track of the actual size of the option
that should be returned to the caller in getsockopt(2).
(Tom Ivar Helbekkmo)
|
| 1.279.2.6 | 19-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #557):
sys/netinet/ip_output.c: 1.295
Keep a pointer to the interface of the multicast membership, because the
multicast element itself might go away in in_delmulti (but the interface
can't because we hold the lock). From ozaki-r@
|
| 1.279.2.5 | 13-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #494):
sys/netinet/ip_output.c: revision 1.291-1.292
- this is not python, we need braces
- protect ifp locking against NULL
--
from ozaki-r: use the proper ifp.
XXX: perhaps push the lock in in_delmulti()?
|
| 1.279.2.4 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #463):
sys/netinet/in.c: revision 1.212
sys/netinet/ip_output.c: revision 1.288
sys/netinet6/in6.c: revision 1.256
sys/netinet6/in6_pcb.c: revision 1.163
sys/sys/lwp.h: revision 1.176
Add missing curlwp_bindx
--
Add missing curlwp_bindx
--
Check LP_BOUND is surely set in curlwp_bindx
This may find an extra call of curlwp_bindx.
--
Fix usage of curlwp_bind in ip_output
curlwp_bindx must be called in LIFO order, i.e., we can't call curlwp_bind
and curlwp_bindx like this:
bound1 = curlwp_bind();
bound2 = curlwp_bind();
curlwp_bindx(bound1);
curlwp_bindx(bound2);
ip_outout did so if NET_MPSAFE. Fix it.
--
Fix wrong usage of psref_held
We can't use it for checking if a caller does NOT hold a given target.
If you want to do it you should have psref_not_held or something.
|
| 1.279.2.3 | 02-Jan-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #456):
sys/arch/arm/sunxi/sunxi_emac.c: 1.9
sys/dev/ic/dwc_gmac.c: 1.43-1.44
sys/dev/pci/if_iwm.c: 1.75
sys/dev/pci/if_wm.c: 1.543
sys/dev/pci/ixgbe/ixgbe.c: 1.112
sys/dev/pci/ixgbe/ixv.c: 1.74
sys/kern/sys_socket.c: 1.75
sys/net/agr/if_agr.c: 1.43
sys/net/bpf.c: 1.219
sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
sys/net/if_etherip.c: 1.40
sys/net/if_ethersubr.c: 1.243, 1.246
sys/net/if_faith.c: 1.57
sys/net/if_gif.c: 1.132
sys/net/if_l2tp.c: 1.15, 1.17
sys/net/if_loop.c: 1.98-1.101
sys/net/if_media.c: 1.35
sys/net/if_pppoe.c: 1.131-1.132
sys/net/if_spppsubr.c: 1.176-1.177
sys/net/if_tun.c: 1.142
sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
sys/net/npf/npf_ifaddr.c: 1.3
sys/net/npf/npf_os.c: 1.8-1.9
sys/net/rtsock.c: 1.230
sys/netcan/if_canloop.c: 1.3-1.5
sys/netinet/if_arp.c: 1.255
sys/netinet/igmp.c: 1.65
sys/netinet/in.c: 1.210-1.211
sys/netinet/in_pcb.c: 1.180
sys/netinet/ip_carp.c: 1.92, 1.94
sys/netinet/ip_flow.c: 1.81
sys/netinet/ip_input.c: 1.362
sys/netinet/ip_mroute.c: 1.147
sys/netinet/ip_output.c: 1.283, 1.285, 1.287
sys/netinet6/frag6.c: 1.61
sys/netinet6/in6.c: 1.251, 1.255
sys/netinet6/in6_pcb.c: 1.162
sys/netinet6/ip6_flow.c: 1.35
sys/netinet6/ip6_input.c: 1.183
sys/netinet6/ip6_output.c: 1.196
sys/netinet6/mld6.c: 1.90
sys/netinet6/nd6.c: 1.239-1.240
sys/netinet6/nd6_nbr.c: 1.139
sys/netinet6/nd6_rtr.c: 1.136
sys/netipsec/ipsec_output.c: 1.65
sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.
|
| 1.279.2.2 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.279.2.1 | 07-Jul-2017 |
martin | Pull up following revision(s) (requested by roy in ticket #100):
sys/netinet/ip_output.c: revision 1.280
sys/netinet/ip_output.c: revision 1.282
When outputting, search for the sending address on the sending interface
rather than blindly picking the first matcing address from any interface
when testing source address validity.
This allows another interface to have the same address, but be detached.
Rename u to udst, .dst to .sa and .dst4 to sin.
Create sockaddr for the source address in usrc so it won't stamp on udst.
This fixes a regression caused in r1.280
|
| 1.298.2.7 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.298.2.6 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.298.2.5 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.298.2.4 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.298.2.3 | 22-Apr-2018 |
pgoyette | Sync with HEAD
|
| 1.298.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.298.2.1 | 07-Apr-2018 |
pgoyette | Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
|
| 1.306.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.306.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.324.2.3 | 29-Jul-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1140):
sys/netinet/ip_output.c: revision 1.330
sys/netinet/sctp_output.c: revision 1.39
sys/netinet/ip_mroute.c: revision 1.166
sys/netipsec/ipsecif.c: revision 1.24
sys/netipsec/xform_ipip.c: revision 1.80
sys/netinet/ip_output.c: revision 1.327
sys/netinet/ip_output.c: revision 1.328
sys/netinet/ip_input.c: revision 1.406
sys/netinet/ip_output.c: revision 1.329
sys/netinet/in_var.h: revision 1.105
in: get rid of unused argument from ip_newid() and ip_newid_range()
in: take a reference of ifp on IP_ROUTETOIF
The ifp could be released after ia4_release(ia).
in: narrow the scope of ifa in ip_output (NFC)
sctp: follow the recent change of ip_newid()
in: avoid racy ifa_acquire(rt->rt_ifa) in ip_output()
If a rtentry is being destroyed asynchronously, ifa referenced by rt_ifa
can be destructed and taking ifa_acquire(rt->rt_ifa) aborts with a
KASSERT failure. Fortunately, the ifa is not actually freed because of
a reference by rt_ifa, it can be available (except some functions like
psref) so as long the rtentry is held.
PR kern/59527
in: avoid racy ia4_acquire(ifatoia(rt->rt_ifa) in ip_rtaddr()
Same as the case of ip_output(), it's racy and should be avoided.
PR kern/59527
|
| 1.324.2.2 | 21-Sep-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #903):
sys/netinet/ip_output.c: revision 1.326
Again allow multicast packets to be sent from unnumbered interfaces.
|
| 1.324.2.1 | 25-Apr-2023 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #150):
sys/netinet/ip_output.c: revision 1.325
Revert "Fix panic on packet sending via a route with rt_ifa of AF_LINK."
The fix is mistakenly upstreamed.
|
| 1.326.6.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.5 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.4 | 23-Jul-2004 |
martti | branches: 1.4.2;
Upgraded IPFilter to 4.1.3
|
| 1.3 | 21-Apr-2004 |
itojun | kill some strcpy
|
| 1.2 | 28-Mar-2004 |
martti | branches: 1.2.2;
Sync with official IPFilter
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.1 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.2.2.1 | 13-Aug-2004 |
jmc | branches: 1.2.2.1.2;
Pullup rev 1.3-1.4 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.2.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.5 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.4.2.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.4.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.4.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.4.2.3 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.4.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.4.2.1 | 23-Jul-2004 |
skrll | file ip_pool.c was added on branch ktrace-lwp on 2004-08-03 10:54:41 +0000
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_pool.h was added on branch ktrace-lwp on 2004-08-03 10:54:41 +0000
|
| 1.2 | 02-Oct-2004 |
christos | These are ipfilter files, although they don't have the same copyright.
Thanks jaromir.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 23-Jul-2004 |
martti | branches: 1.1.1.2.2;
Import IPFilter 4.1.3
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.2;
Import IPFilter 4.1.1
|
| 1.1.1.2.2.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.2.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.2.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.2.2.3 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.1.1.2.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.2.2.1 | 23-Jul-2004 |
skrll | file ip_pptp_pxy.c was added on branch ktrace-lwp on 2004-08-03 10:54:41 +0000
|
| 1.1.1.1.2.1 | 13-Aug-2004 |
jmc | branches: 1.1.1.1.2.1.2;
Pullup rev 1.1.1.2 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.1.1.1.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.6 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.5 | 17-Feb-2021 |
christos | - pass the alignment instead of the mask (as Roy asked and to match the
other macro)
- use alignof to determine that alignment and CTASSERT what we expect
- remove unused macros
|
| 1.4 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.3 | 28-Apr-2008 |
martin | branches: 1.3.4; 1.3.102;
Remove clause 3 and 4 from TNF licenses
|
| 1.2 | 23-Apr-2008 |
thorpej | branches: 1.2.2;
Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.1 | 12-Apr-2008 |
thorpej | branches: 1.1.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.1.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.2.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.3.102.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.3.4.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.3.4.1 | 28-Apr-2008 |
mjf | file ip_private.h was added on branch mjf-devfs2 on 2008-06-02 13:24:24 +0000
|
| 1.39 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.38 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.37 | 28-Mar-2004 |
martti | branches: 1.37.2;
Upgraded IPFilter to 4.1.1
|
| 1.36 | 19-Sep-2002 |
martti | branches: 1.36.6;
Resync with official IPF
|
| 1.35 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.34 | 09-Jun-2002 |
itojun | whitespace
|
| 1.33 | 02-May-2002 |
martti | branches: 1.33.2; 1.33.4;
Fix compilation problems
|
| 1.32 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.31 | 01-Apr-2002 |
jdolecek | Disable the H.323 proxy again - it's too buggy to be supported option
for now. Suggested by Matthew Green and Bernd Ernesti.
|
| 1.30 | 01-Apr-2002 |
jdolecek | put back ip_h323_pxy.c - the QNX licence seems to be okay upon
further examination
|
| 1.29 | 14-Mar-2002 |
martti | Removed unused proxy file
|
| 1.28 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.27 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.26 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.25 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.24 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.23 | 05-Feb-2001 |
chs | branches: 1.23.2; 1.23.4;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.22 | 11-May-2000 |
veego | branches: 1.22.4;
Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.21 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.20 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.19 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.18 | 24-Aug-1999 |
bouyer | branches: 1.18.2; 1.18.8;
Fix for kern/7831 from Darren Reed after discussion on tech-net 2 weeks ago:
check that the packet if of the rigth protocol before giving it to the
proxy module, otherwise let the ipnat code handle it.
What happens in kern/7831 is that a router sends back a icmp message for
a TCP SYN, and ip_proxy.c forwards it to ip_ftp_pxy.c which can only
handle TCP packets. The icmp message is properly handled by ipnat, no need to
go to ip_ftp_pxy.c.
|
| 1.17 | 02-Feb-1999 |
cjs | branches: 1.17.2;
Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.16 | 23-Jan-1999 |
mycroft | Fix problems with fr_tcpsum() that prevented the FTP proxy from working.
|
| 1.15 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.14 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.13 | 29-May-1998 |
veego | Fix some compiler warnings: Missing prototype and ()'s.
|
| 1.12 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.11 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.10 | 28-Nov-1997 |
darrenr | don't free pointer to static struct. please pullup.
|
| 1.9 | 25-Nov-1997 |
mrg | fixes for memory leaks in proxying, and byte ordering problems. from darren reed.
|
| 1.8 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.7 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.6 | 21-Sep-1997 |
veego | branches: 1.6.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.5 | 06-Jul-1997 |
thorpej | branches: 1.5.2;
Restore original RCS IDs.
|
| 1.4 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.3 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.2 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures:
- Deal with lame gcc -Wuninitialized warning (which is incorrect)
- Add parens around assignments within conditionals.
|
| 1.1 | 26-May-1997 |
darrenr | branches: 1.1.1;
Initial revision
|
| 1.1.1.19 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.18 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.17 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.16 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.15 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.14 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.13 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.12 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.11 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.10 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.9 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.8 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.7 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.6 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.5 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.4 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.3 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.2 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.1 | 26-May-1997 |
darrenr | Import new sources for 3.2alpha7
(blah, someone want to clean away /cvsroot/sys/netinet ?)
|
| 1.5.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.6.2.6 | 24-Nov-1998 |
cgd | pull up rev(s) 1.15 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.6.2.5 | 22-Jul-1998 |
mellon | Pull up 1.14 (veego)
|
| 1.6.2.4 | 28-Nov-1997 |
mellon | Pull rev 1.10 up from trunk (darren)
|
| 1.6.2.3 | 25-Nov-1997 |
mrg | pull up from trunk: fixes for memory leaks in proxying, and byte ordering problems. from darren reed.
|
| 1.6.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.6.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.17.2.2 | 20-Dec-1999 |
he | Pull up revision 1.19 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.17.2.1 | 24-Aug-1999 |
he | Pull up revision 1.18:
Check the protocol before forwarding to proxy module. Fixes
PR#7831. (bouyer/darrenr)
|
| 1.18.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.18.2.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.18.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.22.4.2 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.22.4.1 | 09-Feb-2002 |
he | Pull up revisions 1.23-1.27 (requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.23.4.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.23.4.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.23.4.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.23.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.23.2.8 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.23.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.23.2.6 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.23.2.5 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.23.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.23.2.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.23.2.2 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.23.2.1 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.33.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.33.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.36.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.36.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.37.2.1 | 13-Aug-2004 |
jmc | branches: 1.37.2.1.2;
Pullup rev 1.38 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.37.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.39 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.20 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.19 | 28-Mar-2004 |
martti | branches: 1.19.4;
Upgraded IPFilter to 4.1.1
|
| 1.18 | 19-Sep-2002 |
martti | branches: 1.18.6;
Upgraded IPFilter to 3.4.29
|
| 1.17 | 24-Jan-2002 |
martti | branches: 1.17.10;
Upgraded IPFilter to 3.4.23
|
| 1.16 | 26-Mar-2001 |
mike | branches: 1.16.2;
Resolve conflicts.
|
| 1.15 | 11-May-2000 |
veego | branches: 1.15.4; 1.15.6;
Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.14 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.13 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.12 | 22-Nov-1998 |
mrg | branches: 1.12.4; 1.12.10; 1.12.16;
merge ipf 3.2.10
|
| 1.11 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.10 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.9 | 25-Nov-1997 |
mrg | fixes for memory leaks in proxying, and byte ordering problems. from darren reed.
|
| 1.8 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.7 | 21-Sep-1997 |
veego | branches: 1.7.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.6 | 06-Jul-1997 |
thorpej | branches: 1.6.2;
Restore original RCS IDs.
|
| 1.5 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.4 | 28-May-1997 |
thorpej | Change the aps_tout member of struct ap_session from time_t to u_long
so that it can be passed to the filter rule aging functions, which
expect a pointer to a u_long. (time_t is an int on the alpha.)
|
| 1.3 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.2 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures:
- Add prototypes.
- Add a forward-decl to avoid a cyclic dependency graph.
|
| 1.1 | 26-May-1997 |
darrenr | branches: 1.1.1;
Initial revision
|
| 1.1.1.16 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.15 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.14 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.13 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.12 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.11 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.10 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.9 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.8 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.7 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.6 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.5 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.4 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.3 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.2 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.1 | 26-May-1997 |
darrenr | Import new sources for 3.2alpha7
(blah, someone want to clean away /cvsroot/sys/netinet ?)
|
| 1.6.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.7.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.12 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.7.2.3 | 22-Jul-1998 |
mellon | Pull up 1.11 (veego)
|
| 1.7.2.2 | 25-Nov-1997 |
mrg | pull up from trunk: fixes for memory leaks in proxying, and byte ordering problems. from darren reed.
|
| 1.7.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.12.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.12.10.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.12.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.12.4.1 | 20-Dec-1999 |
he | Pull up revision 1.13 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.15.6.3 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.15.6.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.15.6.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.15.4.2 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.15.4.1 | 09-Feb-2002 |
he | Pull up revisions 1.16-1.17 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.16.2.2 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.16.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.17.10.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.18.6.4 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.18.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.18.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.18.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.19.4.1 | 06-Feb-2005 |
jmc | Pull up revision 1.20 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.11 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.10 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.9 | 28-Mar-2004 |
martti | branches: 1.9.2;
Upgraded IPFilter to 4.1.1
|
| 1.8 | 24-Jan-2002 |
martti | branches: 1.8.16;
Upgraded IPFilter to 3.4.23
|
| 1.7 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.6 | 26-Mar-2001 |
mike | branches: 1.6.2;
Resolve conflicts.
|
| 1.5 | 11-May-2000 |
veego | branches: 1.5.4; 1.5.6; 1.5.8;
Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.4 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.3 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.2 | 28-Dec-1999 |
darrenr | update ipfilter code to 3.3.6
|
| 1.1 | 12-Dec-1999 |
veego | branches: 1.1.1;
Initial revision
|
| 1.1.1.9 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.8 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.7 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.6 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.5 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.4 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.3 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.2 | 28-Dec-1999 |
darrenr | update DARRENR branch of netinet to 3.3.6
|
| 1.1.1.1 | 12-Dec-1999 |
veego | branches: 1.1.1.1.2; 1.1.1.1.4;
Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.1.4.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.1.1.1.2.3 | 08-Jan-2000 |
he | Pull up revision 1.2 (requested by darrenr):
Update IPF to version 3.3.6.
|
| 1.1.1.1.2.2 | 20-Dec-1999 |
he | Pull up revision 1.1.1.1 (new) (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.1.1.1.2.1 | 12-Dec-1999 |
he | file ip_raudio_pxy.c was added on branch netbsd-1-4 on 1999-12-20 21:02:06 +0000
|
| 1.5.8.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.5.8.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.5.8.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.5.6.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.5.6.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.5.6.1 | 11-May-2000 |
bouyer | file ip_raudio_pxy.c was added on branch thorpej_scsipi on 2000-11-20 18:10:33 +0000
|
| 1.5.4.2 | 13-Feb-2002 |
he | Apply patch (requested by he):
Bump first argument to __KERNEL_RCSID for these files, to allow
kernels with IPF to build on a.out systems. Fixes PR#15589.
|
| 1.5.4.1 | 09-Feb-2002 |
he | Pull up revisions 1.6-1.8 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.6.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.6.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.8.16.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.8.16.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.9.2.1 | 13-Aug-2004 |
jmc | branches: 1.9.2.1.2;
Pullup rev 1.10 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.9.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.11 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.12 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.11 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.10 | 28-Mar-2004 |
martti | branches: 1.10.2;
Upgraded IPFilter to 4.1.1
|
| 1.9 | 24-Jan-2002 |
martti | branches: 1.9.16;
Re-sync with IPFilter
|
| 1.8 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.7 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.6 | 26-Mar-2001 |
mike | branches: 1.6.2;
Resolve conflicts.
|
| 1.5 | 09-Aug-2000 |
veego | branches: 1.5.2; 1.5.4;
Resolve conflicts.
|
| 1.4 | 11-May-2000 |
veego | branches: 1.4.4;
Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.3 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.2 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.1 | 12-Dec-1999 |
veego | branches: 1.1.1;
Initial revision
|
| 1.1.1.8 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.7 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.6 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.5 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.4 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.3 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.2 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.1 | 12-Dec-1999 |
veego | branches: 1.1.1.1.2; 1.1.1.1.4;
Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.1.4.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.1.1.1.2.2 | 20-Dec-1999 |
he | Pull up revision 1.1.1.1 (new) (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.1.1.1.2.1 | 12-Dec-1999 |
he | file ip_rcmd_pxy.c was added on branch netbsd-1-4 on 1999-12-20 21:02:07 +0000
|
| 1.4.4.3 | 13-Feb-2002 |
he | Apply patch (requested by he):
Bump first argument to __KERNEL_RCSID for these files, to allow
kernels with IPF to build on a.out systems. Fixes PR#15589.
|
| 1.4.4.2 | 09-Feb-2002 |
he | Pull up revisions 1.6-1.9 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.4.4.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.5.4.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.5.4.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.5.4.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.5.2.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.5.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.5.2.1 | 09-Aug-2000 |
bouyer | file ip_rcmd_pxy.c was added on branch thorpej_scsipi on 2000-11-20 18:10:33 +0000
|
| 1.6.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.6.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.9.16.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.9.16.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.10.2.1 | 13-Aug-2004 |
jmc | branches: 1.10.2.1.2;
Pullup rev 1.11 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.10.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.12 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.23 | 31-May-2022 |
andvar | fix various typos in comments, documentation and messages.
|
| 1.22 | 16-Feb-2022 |
andvar | fix various typos, mainly in comments.
|
| 1.21 | 12-Oct-2018 |
maxv | Force ip_off to zero when the reassembly is complete. This was lost in my
rev1.19 - before that the IP struct was clobbered for the reassembly, but
it actually implicitly guaranteed that the first fragment of the packet
would end up with ip_off = 0, and this was a desired behavior.
|
| 1.20 | 17-Sep-2018 |
maxv | Kick fragments that would introduce several !MFFs in a reassembly chain.
The problem arises if we receive three fragments of the kind
3. A -> has MFF
1. B -> doesn't have MFF
2. C -> doesn't have MFF
Because of the received order B->C->A, we don't see that B is !MFF, and
therefore that there is a problem in this chain.
Now we do two checks, and drop us if:
* there is a fragment preceding us, and this fragment is !MFF, or
* there is a fragment following us, and we are !MFF
Spotted a long time ago.
|
| 1.19 | 17-Sep-2018 |
maxv | Hold ip_off and ip_len in the fragment entry, instead of always reading
the associated mbuf (and converting to host order). This reduces the
cache/TLB misses when processing long lists.
|
| 1.18 | 10-Jul-2018 |
maxv | Remove the second argument from ip_reass_packet(). We want the IP header
on the mbuf, not elsewhere. Simplifies the NPF reassembly code a little.
No real functional change.
|
| 1.17 | 15-May-2018 |
maxv | branches: 1.17.2;
When reassembling IPv4/IPv6 packets, ensure each fragment has been subject
to the same IPsec processing. That is to say, that all fragments are ESP,
or AH, or AH+ESP, or none.
The reassembly mechanism can be used both on the wire and inside an IPsec
tunnel, so we need to make sure all fragments of a packet were received
on only one side.
Even though I haven't tried, I believe there are configurations where it
would be possible for an attacker to inject an unencrypted fragment into a
legitimate stream of already-decrypted-and-authenticated fragments.
Typically on IPsec gateways with ESP tunnels, where we can encapsulate
fragments (as opposed to the general case, where we fragment encapsulated
data).
Note, for the record: a funnier thing, under IPv4, would be to send a
zero-sized !MFF fragment at the head of the packet, and manage to trigger
an ICMP error; M_DECRYPTED gets lost by the reassembly, and ICMP will reply
with the packet in clear (not encrypted).
|
| 1.16 | 03-May-2018 |
maxv | Rename m_pkthdr_remove -> m_remove_pkthdr, to match the existing naming
convention, eg m_copy_pkthdr and m_move_pkthdr.
|
| 1.15 | 11-Apr-2018 |
maxv | Add 'static', like the prototype.
|
| 1.14 | 09-Mar-2018 |
maxv | Remove M_PKTHDR from secondary mbufs when reassembling packets.
This is a real problem, because I found at least one component that relies
on the fact that only the first mbuf has M_PKTHDR: far from here, in
m_splithdr, we don't update m->m_pkthdr.len if M_PKTHDR is found in a
secondary mbuf. (The initial intention there was to avoid updating
m_pkthdr.len twice, the assumption was that if M_PKTHDR is set then we're
dealing with the first mbuf.) Therefore, when handling fragmented IPsec
packets (in particular IPv6, IPv4 is a bit more complicated), we may end
up with an incorrect m_pkthdr.len after authentication or decryption. In
the case of ESP, this can lead to a remote crash on this instruction:
m_copydata(m, m->m_pkthdr.len - 3, 3, lastthree);
m_pkthdr.len is bigger than the actual mbuf chain.
It seems possible to me to trigger this bug even if you don't have the ESP
key, because the fragmentation part is outside of the encrypted ESP
payload.
So if you MITM the target, and intercept an incoming ESP packet (which you
can't decrypt), you should be able to forge a new specially-crafted,
fragmented packet and stuff the ESP payload (still encrypted, as you
intercepted it) into it. The decryption succeeds and the target crashes.
|
| 1.13 | 08-Feb-2018 |
maxv | branches: 1.13.2;
Change the error stat from IP_STAT_BADFRAGS to IP_STAT_TOOLONG. The
ping_of_death ATF test expects this counter to get increased.
|
| 1.12 | 06-Feb-2018 |
maxv | Add one more check in ip_reass_packet(): make sure that the end of each
fragment does not exceed IP_MAXPACKET.
In ip_reass(), we only check the final length of the reassembled packet
against IP_MAXPACKET.
But there is an integer overflow that can happen a little earlier. We
are doing:
i = ntohs(p->ipqe_ip->ip_off) + ntohs(p->ipqe_ip->ip_len) -
ntohs(ip->ip_off);
[...]
ip->ip_off = htons(ntohs(ip->ip_off) + i);
It is possible that
ntohs(p->ipqe_ip->ip_off) + ntohs(p->ipqe_ip->ip_len) > 65535
so the computation of ip_off wraps to zero. This breaks an assumption in
the reassembler - it expects the list of fragments to be ordered by
offset, and here it's not ordered anymore. (Un)Fortunately I couldn't
turn this into anything exploitable.
With the new check, it is guaranteed that ip_off+ip_len<=65535.
|
| 1.11 | 11-Jan-2017 |
ozaki-r | branches: 1.11.8;
Get rid of unnecessary header inclusions
|
| 1.10 | 26-Apr-2016 |
ozaki-r | branches: 1.10.2;
Sweep unnecessary route.h inclusions
|
| 1.9 | 25-Feb-2014 |
pooka | branches: 1.9.4; 1.9.6; 1.9.8; 1.9.12;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.8 | 27-Jun-2011 |
enami | branches: 1.8.2; 1.8.12; 1.8.16;
Don't increment ip_nfragpackets when failed to allocate fragment queue.
No one will decrement it on such case.
|
| 1.7 | 05-Nov-2010 |
rmind | branches: 1.7.6;
ip_reass_packet: finish abstraction; some clean-up.
Discussed some time ago with matt@.
|
| 1.6 | 07-Oct-2010 |
yamt | make ipfr_lock IPL_VM as ip_reass_drain is called in interrupts via
the drain hook for mbuf pools.
|
| 1.5 | 06-Oct-2010 |
enami | Don't free memory still in use. Fixes nfs root problem reported
by Christoph Egger on source-changes-d.
|
| 1.4 | 03-Oct-2010 |
rmind | Re-structure IPv4 reassembly code to make it more MP-friendly and simplify
some code fragments while here. Also, use pool_cache(9) and mutex(9).
IPv4 reassembly mechanism is MP-safe now.
|
| 1.3 | 25-Aug-2010 |
rmind | Use own IPv4 reassembly queue entry structure and leave struct ipqent only
for TCP. Now both struct ipfr_qent, struct ipfr_queue and hashed fragment
queue are abstracted and no longer public.
|
| 1.2 | 19-Jul-2010 |
rmind | branches: 1.2.2; 1.2.4;
Abstract IP reassembly into single generic routine - ip_reass_packet().
Make struct ipq private and struct ipqent not visible to userland.
Push ip_len adjustment into reassembly layer.
OK matt@
|
| 1.1 | 13-Jul-2010 |
rmind | Split-off IPv4 re-assembly mechanism into a separate module. Abstract
into ip_reass_init(), ip_reass_lookup(), etc (note: abstraction is not
yet complete). No functional changes to the actual mechanism.
OK matt@
|
| 1.2.4.4 | 06-Nov-2010 |
uebayasi | Sync with HEAD.
|
| 1.2.4.3 | 22-Oct-2010 |
uebayasi | Sync with HEAD (-D20101022).
|
| 1.2.4.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.2.4.1 | 19-Jul-2010 |
uebayasi | file ip_reass.c was added on branch uebayasi-xip on 2010-08-17 06:47:46 +0000
|
| 1.2.2.3 | 09-Oct-2010 |
yamt | sync with head
|
| 1.2.2.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.2.2.1 | 19-Jul-2010 |
yamt | file ip_reass.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:56 +0000
|
| 1.7.6.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.7.6.1 | 05-Nov-2010 |
rmind | file ip_reass.c was added on branch rmind-uvmplock on 2011-03-05 20:55:58 +0000
|
| 1.8.16.1 | 18-May-2014 |
rmind | sync with head
|
| 1.8.12.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.8.12.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.8.2.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.9.12.1 | 05-Apr-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1594):
sys/kern/uipc_mbuf.c: revision 1.182
sys/netinet6/frag6.c: revision 1.67
sys/netinet/ip_reass.c: revision 1.14
sys/sys/mbuf.h: revision 1.179
Remove M_PKTHDR from secondary mbufs when reassembling packets.
This is a real problem, because I found at least one component that relies
on the fact that only the first mbuf has M_PKTHDR: far from here, in
m_splithdr, we don't update m->m_pkthdr.len if M_PKTHDR is found in a
secondary mbuf. (The initial intention there was to avoid updating
m_pkthdr.len twice, the assumption was that if M_PKTHDR is set then we're
dealing with the first mbuf.) Therefore, when handling fragmented IPsec
packets (in particular IPv6, IPv4 is a bit more complicated), we may end
up with an incorrect m_pkthdr.len after authentication or decryption. In
the case of ESP, this can lead to a remote crash on this instruction:
m_copydata(m, m->m_pkthdr.len - 3, 3, lastthree);
m_pkthdr.len is bigger than the actual mbuf chain.
It seems possible to me to trigger this bug even if you don't have the ESP
key, because the fragmentation part is outside of the encrypted ESP
payload.
So if you MITM the target, and intercept an incoming ESP packet (which you
can't decrypt), you should be able to forge a new specially-crafted,
fragmented packet and stuff the ESP payload (still encrypted, as you
intercepted it) into it. The decryption succeeds and the target crashes.
|
| 1.9.8.1 | 05-Apr-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1594):
sys/kern/uipc_mbuf.c: revision 1.182
sys/netinet6/frag6.c: revision 1.67
sys/netinet/ip_reass.c: revision 1.14
sys/sys/mbuf.h: revision 1.179
Remove M_PKTHDR from secondary mbufs when reassembling packets.
This is a real problem, because I found at least one component that relies
on the fact that only the first mbuf has M_PKTHDR: far from here, in
m_splithdr, we don't update m->m_pkthdr.len if M_PKTHDR is found in a
secondary mbuf. (The initial intention there was to avoid updating
m_pkthdr.len twice, the assumption was that if M_PKTHDR is set then we're
dealing with the first mbuf.) Therefore, when handling fragmented IPsec
packets (in particular IPv6, IPv4 is a bit more complicated), we may end
up with an incorrect m_pkthdr.len after authentication or decryption. In
the case of ESP, this can lead to a remote crash on this instruction:
m_copydata(m, m->m_pkthdr.len - 3, 3, lastthree);
m_pkthdr.len is bigger than the actual mbuf chain.
It seems possible to me to trigger this bug even if you don't have the ESP
key, because the fragmentation part is outside of the encrypted ESP
payload.
So if you MITM the target, and intercept an incoming ESP packet (which you
can't decrypt), you should be able to forge a new specially-crafted,
fragmented packet and stuff the ESP payload (still encrypted, as you
intercepted it) into it. The decryption succeeds and the target crashes.
|
| 1.9.6.2 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.9.6.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.9.4.1 | 05-Apr-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1594):
sys/kern/uipc_mbuf.c: revision 1.182
sys/netinet6/frag6.c: revision 1.67
sys/netinet/ip_reass.c: revision 1.14
sys/sys/mbuf.h: revision 1.179
Remove M_PKTHDR from secondary mbufs when reassembling packets.
This is a real problem, because I found at least one component that relies
on the fact that only the first mbuf has M_PKTHDR: far from here, in
m_splithdr, we don't update m->m_pkthdr.len if M_PKTHDR is found in a
secondary mbuf. (The initial intention there was to avoid updating
m_pkthdr.len twice, the assumption was that if M_PKTHDR is set then we're
dealing with the first mbuf.) Therefore, when handling fragmented IPsec
packets (in particular IPv6, IPv4 is a bit more complicated), we may end
up with an incorrect m_pkthdr.len after authentication or decryption. In
the case of ESP, this can lead to a remote crash on this instruction:
m_copydata(m, m->m_pkthdr.len - 3, 3, lastthree);
m_pkthdr.len is bigger than the actual mbuf chain.
It seems possible to me to trigger this bug even if you don't have the ESP
key, because the fragmentation part is outside of the encrypted ESP
payload.
So if you MITM the target, and intercept an incoming ESP packet (which you
can't decrypt), you should be able to forge a new specially-crafted,
fragmented packet and stuff the ESP payload (still encrypted, as you
intercepted it) into it. The decryption succeeds and the target crashes.
|
| 1.10.2.1 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.11.8.7 | 17-Oct-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1045):
sys/netinet/ip_reass.c: revision 1.19-1.21
Hold ip_off and ip_len in the fragment entry, instead of always reading
the associated mbuf (and converting to host order). This reduces the
cache/TLB misses when processing long lists.
-
Kick fragments that would introduce several !MFFs in a reassembly chain.
The problem arises if we receive three fragments of the kind
3. A -> has MFF
1. B -> doesn't have MFF
2. C -> doesn't have MFF
Because of the received order B->C->A, we don't see that B is !MFF, and
therefore that there is a problem in this chain.
Now we do two checks, and drop us if:
* there is a fragment preceding us, and this fragment is !MFF, or
* there is a fragment following us, and we are !MFF
Spotted a long time ago.
-
Force ip_off to zero when the reassembly is complete. This was lost in my
rev1.19 - before that the IP struct was clobbered for the reassembly, but
it actually implicitly guaranteed that the first fragment of the packet
would end up with ip_off = 0, and this was a desired behavior.
|
| 1.11.8.6 | 09-Oct-2018 |
martin | Back out the following from ticket #1045 by maxv:
sys/netinet/ip_reass.c 1.19
Faster IPv4 packet reassembly - causes fallout, needs further investigation
(see PR kern/53664)
|
| 1.11.8.5 | 03-Oct-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1045):
sys/netinet/ip_reass.c: revision 1.19
Hold ip_off and ip_len in the fragment entry, instead of always reading
the associated mbuf (and converting to host order). This reduces the
cache/TLB misses when processing long lists.
|
| 1.11.8.4 | 27-Sep-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1041):
sys/netinet/ip_reass.c: revision 1.17 (patch)
sys/netinet6/frag6.c: revision 1.74 (patch)
When reassembling IPv4/IPv6 packets, ensure each fragment has been subject
to the same IPsec processing. That is to say, that all fragments are ESP,
or AH, or AH+ESP, or none.
The reassembly mechanism can be used both on the wire and inside an IPsec
tunnel, so we need to make sure all fragments of a packet were received
on only one side.
Even though I haven't tried, I believe there are configurations where it
would be possible for an attacker to inject an unencrypted fragment into a
legitimate stream of already-decrypted-and-authenticated fragments.
Typically on IPsec gateways with ESP tunnels, where we can encapsulate
fragments (as opposed to the general case, where we fragment encapsulated
data).
Note, for the record: a funnier thing, under IPv4, would be to send a
zero-sized !MFF fragment at the head of the packet, and manage to trigger
an ICMP error; M_DECRYPTED gets lost by the reassembly, and ICMP will reply
with the packet in clear (not encrypted).
|
| 1.11.8.3 | 09-Apr-2018 |
martin | Additionally pull up the following revision for ticket #668,
requested by ozaki-r:
sys/netinet/ip_reass.c 1.13
Change the error stat from IP_STAT_BADFRAGS to IP_STAT_TOOLONG. The
ping_of_death ATF test expects this counter to get increased.
|
| 1.11.8.2 | 05-Apr-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #695):
sys/kern/uipc_mbuf.c: revision 1.182
sys/netinet6/frag6.c: revision 1.67
sys/netinet/ip_reass.c: revision 1.14
sys/sys/mbuf.h: revision 1.179
Remove M_PKTHDR from secondary mbufs when reassembling packets.
This is a real problem, because I found at least one component that relies
on the fact that only the first mbuf has M_PKTHDR: far from here, in
m_splithdr, we don't update m->m_pkthdr.len if M_PKTHDR is found in a
secondary mbuf. (The initial intention there was to avoid updating
m_pkthdr.len twice, the assumption was that if M_PKTHDR is set then we're
dealing with the first mbuf.) Therefore, when handling fragmented IPsec
packets (in particular IPv6, IPv4 is a bit more complicated), we may end
up with an incorrect m_pkthdr.len after authentication or decryption. In
the case of ESP, this can lead to a remote crash on this instruction:
m_copydata(m, m->m_pkthdr.len - 3, 3, lastthree);
m_pkthdr.len is bigger than the actual mbuf chain.
It seems possible to me to trigger this bug even if you don't have the ESP
key, because the fragmentation part is outside of the encrypted ESP
payload.
So if you MITM the target, and intercept an incoming ESP packet (which you
can't decrypt), you should be able to forge a new specially-crafted,
fragmented packet and stuff the ESP payload (still encrypted, as you
intercepted it) into it. The decryption succeeds and the target crashes.
|
| 1.11.8.1 | 30-Mar-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #668):
sys/netinet/ip_reass.c: revision 1.12
Add one more check in ip_reass_packet(): make sure that the end of each
fragment does not exceed IP_MAXPACKET.
In ip_reass(), we only check the final length of the reassembled packet
against IP_MAXPACKET.
But there is an integer overflow that can happen a little earlier. We
are doing:
i = ntohs(p->ipqe_ip->ip_off) + ntohs(p->ipqe_ip->ip_len) -
ntohs(ip->ip_off);
[...]
ip->ip_off = htons(ntohs(ip->ip_off) + i);
It is possible that
ntohs(p->ipqe_ip->ip_off) + ntohs(p->ipqe_ip->ip_len) > 65535
so the computation of ip_off wraps to zero. This breaks an assumption in
the reassembler - it expects the list of fragments to be ordered by
offset, and here it's not ordered anymore. (Un)Fortunately I couldn't
turn this into anything exploitable.
With the new check, it is guaranteed that ip_off+ip_len<=65535.
|
| 1.13.2.6 | 20-Oct-2018 |
pgoyette | Sync with head
|
| 1.13.2.5 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.13.2.4 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.13.2.3 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.13.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.13.2.1 | 15-Mar-2018 |
pgoyette | Synch with HEAD
|
| 1.17.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 23-Jul-2004 |
martti | branches: 1.1.1.2.2;
Import IPFilter 4.1.3
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.2;
Import IPFilter 4.1.1
|
| 1.1.1.2.2.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.2.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.2.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.2.2.3 | 05-Aug-2004 |
skrll | Fix merge mistakes.
|
| 1.1.1.2.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.2.2.1 | 23-Jul-2004 |
skrll | file ip_rpcb_pxy.c was added on branch ktrace-lwp on 2004-08-03 10:54:42 +0000
|
| 1.1.1.1.2.1 | 13-Aug-2004 |
jmc | branches: 1.1.1.1.2.1.2;
Pullup rev 1.1.1.2 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.1.1.1.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.3 | 23-Jul-2004 |
martti | branches: 1.3.2;
Upgraded IPFilter to 4.1.3
|
| 1.2 | 24-Apr-2004 |
matt | Always include <sys/param.h> first!
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.2;
Import IPFilter 4.1.1
|
| 1.1.1.1.2.2 | 13-Aug-2004 |
jmc | Delete file (for pullup 759 requested by christos)
Bring up to ipf 4.1.3
|
| 1.1.1.1.2.1 | 27-Apr-2004 |
jdc | Pull up revision 1.2 (requested by matt in ticket #187)
Always include <sys/param.h> first!
|
| 1.3.2.2 | 25-Aug-2004 |
skrll | These are dead.
|
| 1.3.2.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.2 | 23-Jul-2004 |
martti | branches: 1.2.2;
Upgraded IPFilter to 4.1.3
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.2;
Import IPFilter 4.1.1
|
| 1.1.1.1.2.1 | 13-Aug-2004 |
jmc | Delete file (for pullup 759 requested by christos)
Bring up to ipf 4.1.3
|
| 1.2.2.2 | 25-Aug-2004 |
skrll | These are dead.
|
| 1.2.2.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_scan.c was added on branch ktrace-lwp on 2004-08-03 10:54:42 +0000
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_scan.h was added on branch ktrace-lwp on 2004-08-03 10:54:42 +0000
|
| 1.47 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.46 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.45 | 10-May-2004 |
christos | PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.44 | 28-Mar-2004 |
martti | branches: 1.44.2;
Upgraded IPFilter to 4.1.1
|
| 1.43 | 24-Feb-2004 |
wiz | parameter with two es. From Peter Postma.
|
| 1.42 | 19-Sep-2002 |
martti | branches: 1.42.6;
Resync with official IPF
|
| 1.41 | 19-Sep-2002 |
martti | Upgraded IPFilter to 3.4.29
|
| 1.40 | 09-Jun-2002 |
itojun | whitespace
|
| 1.39 | 01-Jun-2002 |
yamt | make "keep state" work for SYN without win scale option.
|
| 1.38 | 02-May-2002 |
martti | branches: 1.38.2; 1.38.4;
Fix compilation problems
|
| 1.37 | 02-May-2002 |
martti | Upgraded IPFilter to 3.4.27
|
| 1.36 | 09-Apr-2002 |
thorpej | Add missing #else
|
| 1.35 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.34 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.33 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.32 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.31 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.30 | 06-Apr-2001 |
darrenr | branches: 1.30.2;
fix fragment cache security hole
|
| 1.29 | 26-Mar-2001 |
mike | Resolve conflicts.
|
| 1.28 | 05-Feb-2001 |
chs | branches: 1.28.2;
expose the definitions of MIN() and MAX() in sys/param.h to the kernel
and use those in favor of a dozen copies scattered around the source tree.
|
| 1.27 | 09-Aug-2000 |
veego | Resolve conflicts.
|
| 1.26 | 23-May-2000 |
veego | branches: 1.26.4;
Resolve conflicts.
|
| 1.25 | 21-May-2000 |
veego | Resolve conflicts.
|
| 1.24 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.23 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.22 | 07-Feb-2000 |
veego | Fix from Darren Reed for the test failure of f11.
|
| 1.21 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.20 | 29-Dec-1999 |
veego | Fix a panic which was mentioned on the ipfilter mailing list.
Patch from Darren send to the mailing list after he released 3.3.6 and
did a bad job with using the wrong way to update the NetBSD version
of ipfilter.
|
| 1.19 | 28-Dec-1999 |
darrenr | update ipfilter code to 3.3.6
|
| 1.18 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.17 | 02-Feb-1999 |
cjs | branches: 1.17.2; 1.17.8; 1.17.14;
Remove SCCS markers and make these compile in $NetBSD$ IDs.
|
| 1.16 | 22-Nov-1998 |
mrg | merge ipf 3.2.10
|
| 1.15 | 12-Jul-1998 |
veego | Resolve conflicts from the import.
|
| 1.14 | 29-May-1998 |
veego | Fix compiler warnings: Add missing ()'s.
|
| 1.13 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.12 | 17-May-1998 |
veego | Resolve conflicts
|
| 1.11 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.10 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.9 | 21-Sep-1997 |
veego | branches: 1.9.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.8 | 06-Jul-1997 |
thorpej | branches: 1.8.2;
Restore original RCS IDs.
|
| 1.7 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.6 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.5 | 27-May-1997 |
thorpej | Make this compile on 32-bit architectures again:
- Pull in includes to get appropriate prototypes.
|
| 1.4 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.3 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.26 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.25 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.24 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.23 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.22 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.21 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.20 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.19 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.18 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.17 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.16 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.15 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.14 | 28-Dec-1999 |
darrenr | update DARRENR branch of netinet to 3.3.6
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.8.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.16 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.9.2.3 | 22-Jul-1998 |
mellon | Pull up 1.15 (veego)
|
| 1.9.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.9.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.17.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.17.8.4 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.17.8.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.17.8.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.17.8.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.17.2.4 | 14-Apr-2001 |
he | Pull up revision 1.30 (via patch, requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.17.2.3 | 25-May-2000 |
he | Apply patch (requested by darrenr):
Do not let RST TCP segments create state.
|
| 1.17.2.2 | 08-Jan-2000 |
he | Pull up revision 1.19 (requested by darrenr):
Update IPF to version 3.3.6.
|
| 1.17.2.1 | 20-Dec-1999 |
he | Pull up revision 1.18 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.26.4.4 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.26.4.3 | 09-Feb-2002 |
he | Pull up revisions 1.28-1.29,1.31-1.24 (via patch, requested by martti):
Updated IPFilter to 3.4.23.
|
| 1.26.4.2 | 14-Apr-2001 |
he | Pull up revision 1.30 (requested by darrenr):
Fix bug related to fragment cache handling.
|
| 1.26.4.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.28.2.9 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.28.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.28.2.7 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.28.2.6 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.28.2.5 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.28.2.4 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.28.2.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.28.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.28.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.30.2.5 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.30.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.30.2.3 | 15-Mar-2002 |
jdolecek | fix merge botch, it's now identical to rev 1.34
|
| 1.30.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.30.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.38.4.2 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.38.4.1 | 02-Jun-2002 |
tv | Pull up revision 1.39 (requested by yamt in ticket #116):
make "keep state" work for SYN without win scale option.
|
| 1.38.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.42.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.42.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.44.2.4 | 16-Mar-2005 |
tron | Apply patch (requested by martti in ticket #1110):
Disable the oow test because it is broken. It is killing valid packets.
|
| 1.44.2.3 | 08-Oct-2004 |
jmc | branches: 1.44.2.3.2;
Pullup patch (requested by darrenr in ticket #902)
* Prevent hang when attempting to flush state entries for ipv4 when ipv6
are present or vice versa
* Fix matching of IPv6 state entries when the initial packet is a
sent to a multicast address. This includes not updating the address as
being fixed when a second (or further) such packet is seen before a reply.
* Disable code, for now, that limited how many ICMP packets could match a
state entry based on the number of real packets seen.
|
| 1.44.2.2 | 13-Aug-2004 |
jmc | Pullup rev 1.46 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.44.2.1 | 30-May-2004 |
tron | Pull up revision 1.45 (requested by christos in ticket #416):
PR/24969: Arto Selonen: /usr/sbin/ipfs from ipfilter 4.1.1 does not work
patch applied.
|
| 1.44.2.3.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.47 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.26 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.25 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.24 | 28-Mar-2004 |
martti | branches: 1.24.2;
Upgraded IPFilter to 4.1.1
|
| 1.23 | 19-Sep-2002 |
martti | branches: 1.23.6;
Upgraded IPFilter to 3.4.29
|
| 1.22 | 02-May-2002 |
martti | branches: 1.22.4;
Upgraded IPFilter to 3.4.27
|
| 1.21 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.20 | 24-Jan-2002 |
martti | Re-sync with IPFilter
|
| 1.19 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.18 | 26-Mar-2001 |
mike | branches: 1.18.2;
Resolve conflicts.
|
| 1.17 | 09-Aug-2000 |
veego | branches: 1.17.2;
Resolve conflicts.
|
| 1.16 | 03-May-2000 |
veego | branches: 1.16.4;
Resolve conflicts.
|
| 1.15 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.14 | 12-Dec-1999 |
veego | Resolve conflicts and small fixes.
|
| 1.13 | 22-Nov-1998 |
mrg | branches: 1.13.4; 1.13.10; 1.13.16;
merge ipf 3.2.10
|
| 1.12 | 29-May-1998 |
veego | Resolve conflicts from the import of IPFilter 3.2.7.
|
| 1.11 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.10 | 14-Nov-1997 |
mrg | merge ip-filter 3.2.1
|
| 1.9 | 30-Oct-1997 |
mrg | sigh. merge ipfilter 3.2 onto the trunk. merge to the branch was a mistake.
|
| 1.8 | 21-Sep-1997 |
veego | branches: 1.8.2;
Resolve conflicts from the merge of ipf 3.2beta5.
|
| 1.7 | 06-Jul-1997 |
thorpej | branches: 1.7.2;
Restore original RCS IDs.
|
| 1.6 | 05-Jul-1997 |
darrenr | fix conflicts from import
|
| 1.5 | 28-May-1997 |
thorpej | Resolve conflicts from merge of 3.2a7, take 2. Also, eliminate some
silly differences between the NetBSD copy of the code and the
vendor branch, keeping only those which are necessary. Of those
differences that currently exist, several "portability to NetBSD"
issues, which will be fed back to the ipfilter author.
|
| 1.4 | 25-May-1997 |
darrenr | fix conflicts
|
| 1.3 | 29-Mar-1997 |
thorpej | Resolve conflicts from merge.
XXX !!! XXX !!!
I noticed a few semi-serious bugs while doing this merge, one of which
has existed for a fairly long time. Some of them are addressed in this
commit (because they caused the kernel to not compile), and are annoted
by "XXX" and "--thorpej". The other one will be addressed shortly in
a future commit, and, as far as I can tell, affects all operating systems
which IP Filter supports.
|
| 1.2 | 05-Jan-1997 |
veego | Add $NetBSD$ id's and restore the orginal Id's.
|
| 1.1 | 05-Jan-1997 |
mrg | branches: 1.1.1;
initial import of darren reed's ip-filter, version 3.1.2.
|
| 1.1.1.23 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.22 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.21 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.20 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.19 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.18 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.17 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.16 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.15 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.14 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.13 | 12-Dec-1999 |
veego | Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.12 | 22-Nov-1998 |
mrg | ip filter version 3.2.10
|
| 1.1.1.11 | 12-Jul-1998 |
veego | Import IP Filter 3.2.9
|
| 1.1.1.10 | 29-May-1998 |
veego | Import IP Filter 3.2.7
|
| 1.1.1.9 | 17-May-1998 |
veego | Import IP Filter 3.2.5
|
| 1.1.1.8 | 14-Nov-1997 |
mrg | import ip-filter 3.2.1
|
| 1.1.1.7 | 30-Oct-1997 |
mrg | import ip-filter 3.2
|
| 1.1.1.6 | 21-Sep-1997 |
veego | Import ip-filter 3.2beta5
|
| 1.1.1.5 | 05-Jul-1997 |
darrenr | import 3.2beta1 IP Filter sources
|
| 1.1.1.4 | 27-May-1997 |
thorpej | ipfilter2netbsd did not produce correct output for last import of
3.2a7. Re-import it now that ipfilter2netbsd is fixed.
|
| 1.1.1.3 | 25-May-1997 |
darrenr | Import version 3.2alpha7
|
| 1.1.1.2 | 27-Mar-1997 |
darrenr | Bring in entire 3.2alpha2 source tree
|
| 1.1.1.1 | 27-Mar-1997 |
darrenr | Update to version 3.2alpha2
|
| 1.7.2.1 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.8.2.4 | 24-Nov-1998 |
cgd | pull up rev(s) 1.13 from trunk (ipfilter 3.2.10). (mrg)
|
| 1.8.2.3 | 22-Jul-1998 |
mellon | Pull up 1.12 (veego)
|
| 1.8.2.2 | 17-Nov-1997 |
mrg | pull up from trunk: ipfilter 3.2.1 (plus bug fix for fil.c from marc boucher).
|
| 1.8.2.1 | 30-Oct-1997 |
mrg | merge ipfilter 3.2
|
| 1.13.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.13.10.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.13.10.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.13.4.1 | 20-Dec-1999 |
he | Pull up revision 1.14 (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.16.4.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.16.4.2 | 09-Feb-2002 |
he | Pull up revisions 1.18-1.20 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.16.4.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.17.2.5 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.17.2.4 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.17.2.3 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.17.2.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.17.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.18.2.3 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.18.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.18.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.22.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.23.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.23.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.24.2.1 | 13-Aug-2004 |
jmc | branches: 1.24.2.1.2;
Pullup rev 1.25 (requested by christos in ticket #759)
Bring up to ipf 4.1.3
|
| 1.24.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.26 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_sync.c was added on branch ktrace-lwp on 2004-08-03 10:54:43 +0000
|
| 1.2 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.1 | 28-Mar-2004 |
martti | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 | 28-Mar-2004 |
martti | branches: 1.1.1.1.4; 1.1.1.1.6;
Import IPFilter 4.1.1
|
| 1.1.1.1.6.1 | 06-Feb-2005 |
jmc | Pull up revision 1.2 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.1.1.1.4.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.1.1.4.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.1.1.4.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.1.1.1.4.1 | 28-Mar-2004 |
skrll | file ip_sync.h was added on branch ktrace-lwp on 2004-08-03 10:54:43 +0000
|
| 1.135 | 27-Jun-2025 |
andvar | Grammar and spelling fixes, mainly in comments. A few in documentation,
logging, test description, and SCSI ASC/ASCQ assignment descriptions.
|
| 1.134 | 10-Apr-2022 |
andvar | branches: 1.134.10;
fix various typos in comments and output/log messages.
|
| 1.133 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.132 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.131 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.130 | 28-Aug-2020 |
ozaki-r | branches: 1.130.2;
inet: reduce silent packet discards
|
| 1.129 | 28-Aug-2020 |
ozaki-r | inet, inet6: count packets dropped by IPsec
The counters count packets dropped due to security policy checks.
|
| 1.128 | 13-May-2019 |
ozaki-r | Count packets dropped by pfil
|
| 1.127 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.126 | 10-Jul-2018 |
maxv | Remove the second argument from ip_reass_packet(). We want the IP header
on the mbuf, not elsewhere. Simplifies the NPF reassembly code a little.
No real functional change.
|
| 1.125 | 08-Apr-2018 |
maxv | branches: 1.125.2;
Remove the ipre_mlast field and the TRAVERSE macro.
The goal was to store in ipre_mlast the last mbuf of the chain, so that
m_cat could be called on it. But it's not needed, since m_cat already
does the equivalent of TRAVERSE itself.
If it were needed, there would be a bug, since we don't call TRAVERSE on
ipre_mlast when creating a new reassembly entry.
|
| 1.124 | 08-Apr-2018 |
maxv | Remove unused field, and sync comment with reality.
|
| 1.123 | 03-Apr-2018 |
maxv | Remove unused fields and outdated comment.
|
| 1.122 | 10-Jan-2018 |
knakahara | branches: 1.122.2;
add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
|
| 1.121 | 11-Dec-2017 |
ryo | As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.120 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.119 | 31-Mar-2017 |
ozaki-r | branches: 1.119.6;
Don't use a single global variable to store source route information for multiple incoming packets
It's not MP-safe. So use a m_tag to store the information instead.
Pointed out by knakahara@
The fix is from OpenBSD (originally fixed in FreeBSD)
|
| 1.118 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.117 | 16-Feb-2017 |
knakahara | add l2tp(4) L2TPv3 interface.
originally implemented by IIJ SEIL team.
|
| 1.116 | 08-Dec-2016 |
ozaki-r | branches: 1.116.2;
Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.115 | 01-Aug-2016 |
knakahara | improve fast-forward performance when the number of flows exceeds IPFLOW_MAX.
In the fast-forward case, when the number of flows exceeds IPFLOW_MAX, the
performmance degraded to about 50% compared to the case less than IPFLOW_MAX
flows. This modification suppresses the degradation to 65%. Furthermore,
the modified kernel is about the same performance as the original kernel
when the number of flows is less than IPFLOW_MAX.
The original patch is implemented by ryo@n.o. Thanks.
|
| 1.114 | 21-Jun-2016 |
ozaki-r | branches: 1.114.2;
Replace ifp of ip_moptions and ip6_moptions with if_index
The motivation is the same as the mbuf's rcvif case; avoid having a pointer
of an ifnet object in ip_moptions and ip6_moptions, which is not MP-safe.
ip_moptions and ip6_moptions can be stored in a PCB for inet or inet6
that's life time is different from ifnet one and so an ifnet object can be
disappeared anytime we get it via them. Thus we need to look up an ifnet
object by if_index every time for safe.
|
| 1.113 | 13-Jun-2016 |
knakahara | make ipflow_reap() static function.
|
| 1.112 | 28-Apr-2016 |
ozaki-r | Constify rtentry of if_output
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
|
| 1.111 | 26-Apr-2016 |
ozaki-r | Stop using rt_gwroute on packet sending paths
rt_gwroute of rtentry is a reference to a rtentry of the gateway
for a rtentry with RTF_GATEWAY. That was used by L2 (arp and ndp)
to look up L2 addresses. By separating L2 nexthop caches, we don't
need a route for the purpose and we can stop using rt_gwroute.
By doing so, we can reduce referencing and modifying rtentries,
which makes it easy to apply a lock (and/or psref) to the
routing table and rtentries.
One issue to do this is to keep RTF_REJECT behavior. It seems it
was broken when we moved rtalloc1 things from L2 output routines
(e.g., ether_output) to ip_hresolv_output, but (fortunately?)
it works unexpectedly. What we mistook are:
- RTF_REJECT was checked for any routes in L2 output routines,
but in ip_hresolv_output it is checked only when the route
is RTF_GATEWAY
- The RTF_REJECT check wasn't copied to IPv6 (nd6_output)
It seems that rt_gwroute checks hid the mistakes and it looked
work (unexpectedly) and removing rt_gwroute checks unveil the
issue. So we need to fix RTF_REJECT checks in ip_hresolv_output
and also add them to nd6_output.
One more point we have to care is returning an errno; we need
to mimic looutput behavior. Originally RTF_REJECT check was
done either in L2 output routines or in looutput. The latter is
applied when a reject route directs to a loopback interface.
However, now RTF_REJECT check is done before looutput so to keep
the original behavior we need to return an errno which looutput
chooses. Added rt_check_reject_route does such tweaks.
|
| 1.110 | 20-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.109 | 20-Jan-2016 |
riastradh | Give proper prototype to rip_output.
|
| 1.108 | 04-Jun-2015 |
ozaki-r | Pull out route lookups from L2 output routines
Route lookups for routes of RTF_GATEWAY were done in L2 output
routines such as ether_output, but they should be done in L3
i.e., before L2 output routines. This change places the lookups
between L3 output routines (say ip_output) and the L2 output
routines.
The change is based on dyoung's patch submitted in the thread:
https://mail-index.netbsd.org/tech-net/2013/02/01/msg003847.html
You can find out detailed investigations by dyoung about the
issue in there.
Note that the change introduces a workaround for MPLS. ether_output
knew that it needs to fill the ethertype of a frame as MPLS,
based on a tag of an original route (rtentry), but now we don't
pass it to ehter_output. So we have to tell that in another way.
We use mtag to do so for now, which introduces some overhead.
We should fix it somehow in the future.
Discussed on tech-kern and tech-net.
|
| 1.107 | 11-Oct-2014 |
christos | branches: 1.107.2;
exposet multicast option functions which are used by the v6 code now.
|
| 1.106 | 05-Jun-2014 |
rmind | - Implement pktqueue interface for lockless IP input queue.
- Replace ipintrq and ip6intrq with the pktqueue mechanism.
- Eliminate kernel-lock from ipintr() and ip6intr().
- Some preparation work to push softnet_lock out of ipintr().
Discussed on tech-net.
|
| 1.105 | 30-May-2014 |
rmind | Use __CTASSERT() in the header.
|
| 1.104 | 29-May-2014 |
rmind | Make IGMP and multicast group management code MP-safe. Use a read-write
lock to protect the hash table of multicast address records; also, make it
private and eliminate some macros. In the long term, the lookup path ought
to be optimised.
|
| 1.103 | 23-May-2014 |
rmind | Make ip_forward() static, there is no need to expose it.
|
| 1.102 | 22-May-2014 |
rmind | - Make ip_setmoptions(), ip_getmoptions() and ip_pcbopts() static.
- ip_output: eliminate 7th variadic argument; IP_RETURNMTU is flag
always used to store MTU size into struct inpcb::inp_errormtu.
- Clean up these routines: reduce #ifdefs, variable scopes, etc.
|
| 1.101 | 22-May-2014 |
rmind | - Add in_init() and move some functions, variables and sysctls into in.c
where they belong to. Make some functions and variables static.
- ip_input.c: reduce some #ifdefs, cleanup a little.
- Move some sysctls into ip_flow.c as they belong there.
No functional change.
|
| 1.100 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.99 | 19-Mar-2014 |
liamjfoy | branches: 1.99.2;
Move ipflow into ip_var.h and fix confliction
|
| 1.98 | 19-Mar-2014 |
liamjfoy | Remove ipflow_prune and replace with ipflow_reap. ok rmind@
|
| 1.97 | 03-May-2011 |
dyoung | branches: 1.97.4; 1.97.14; 1.97.18;
*_drain() routines may be called with locks held, so instead of doing
any work in *_drain(), set a drain-needed flag. Do the work in the
fasttimo handler.
Contributed by Coyote Point Systems, Inc.
|
| 1.96 | 05-Nov-2010 |
rmind | branches: 1.96.2;
ip_reass_packet: finish abstraction; some clean-up.
Discussed some time ago with matt@.
|
| 1.95 | 25-Aug-2010 |
rmind | Use own IPv4 reassembly queue entry structure and leave struct ipqent only
for TCP. Now both struct ipfr_qent, struct ipfr_queue and hashed fragment
queue are abstracted and no longer public.
|
| 1.94 | 19-Jul-2010 |
rmind | Revert previous change of making struct ipqent invisible to userland.
|
| 1.93 | 19-Jul-2010 |
rmind | Abstract IP reassembly into single generic routine - ip_reass_packet().
Make struct ipq private and struct ipqent not visible to userland.
Push ip_len adjustment into reassembly layer.
OK matt@
|
| 1.92 | 13-Jul-2010 |
rmind | Split-off IPv4 re-assembly mechanism into a separate module. Abstract
into ip_reass_init(), ip_reass_lookup(), etc (note: abstraction is not
yet complete). No functional changes to the actual mechanism.
OK matt@
|
| 1.91 | 01-Feb-2009 |
pooka | branches: 1.91.4; 1.91.6;
Init ipflow pool dynamically instead of using a linkset.
|
| 1.90 | 12-Oct-2008 |
plunky | branches: 1.90.2;
update ip_pcbopts() to use sockopt(9) API.
cleans up function and one small fix is that we now stop copying user
options to the mbuf when the _EOL is given, previously this function
would continue to copy options.
|
| 1.89 | 16-Aug-2008 |
plunky | constify sockopt in the PRCO_SETOPT path
|
| 1.88 | 06-Aug-2008 |
plunky | Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.87 | 12-Apr-2008 |
thorpej | branches: 1.87.4; 1.87.6; 1.87.10;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.86 | 09-Apr-2008 |
thorpej | - ipflow is not used outside ip_flow.c; move its definition there.
- Make ipflow_reap() private to ip_flow.c, and introduce ipflow_prune()
for external callers to use (avoids returning an ipflow * that is never
actually used anyway).
|
| 1.85 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.84 | 06-Feb-2008 |
matt | branches: 1.84.6;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.83 | 25-Dec-2007 |
perry | Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.82 | 22-Dec-2007 |
matt | Make sure ip_newid etal doesn't return an ip_id of 0.
|
| 1.81 | 22-Dec-2007 |
matt | Add ipq_tos to struct ipqe. (Doesn't increase size since the last member
was a u_int16_t).
|
| 1.80 | 02-Oct-2007 |
dyoung | branches: 1.80.4; 1.80.6; 1.80.10;
Delete the unused second argument to ip_stripoptions(), move it
closer to its single caller in if_eon.c, try to move fewer bytes
by moving the IP header forward instead of moving the tail of the
mbuf backward, and use m_adj(9) instead of fiddling directly with
mbuf data members.
|
| 1.79 | 25-Mar-2007 |
liamjfoy | branches: 1.79.8; 1.79.10; 1.79.12;
Add net.inet.ip.hashsize to control the IPv4 fast forward hash table size.
|
| 1.78 | 17-Feb-2007 |
dyoung | branches: 1.78.4; 1.78.6; 1.78.8;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.77 | 16-Feb-2006 |
perry | branches: 1.77.20;
Change "inline" back to "__inline" in .h files -- C99 is still too
new, and some apps compile things in C89 mode. C89 keywords stay.
As per core@.
|
| 1.76 | 24-Dec-2005 |
perry | branches: 1.76.2; 1.76.4; 1.76.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.75 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.74 | 10-Dec-2005 |
elad | Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.73 | 22-Nov-2005 |
yamt | revert rev.1.72 as it isn't necessary.
|
| 1.72 | 06-May-2005 |
matt | branches: 1.72.2; 1.72.8;
Add #include <sys/protosw.h> when _KERNEL
|
| 1.71 | 29-Apr-2005 |
yamt | move decl of inetsw to its own header to avoid array of incomplete type.
found by gcc4. reported by Adam Ciarcinski.
|
| 1.70 | 07-Apr-2005 |
yamt | when doing TSO, avoid to use duplicated ip_id heavily.
XXX ip_randomid
|
| 1.69 | 15-Dec-2004 |
thorpej | branches: 1.69.2; 1.69.8;
Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.68 | 22-Apr-2004 |
matt | Constify protosw arrays. This can reduce the kernel .data section by
over 4K (if all the network protocols) are loaded.
|
| 1.67 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.66 | 18-Apr-2004 |
matt | De __P()
|
| 1.65 | 12-Dec-2003 |
scw | Make fast-ipsec and ipflow (Fast Forwarding) interoperate.
The idea is that we only clear M_CANFASTFWD if an SPD exists
for the packet. Otherwise, it's safe to add a fast-forward
cache entry for the route.
To make this work properly, we invalidate the entire ipflow
cache if a fast-ipsec key is added or changed.
|
| 1.64 | 08-Dec-2003 |
jonathan | Add new field ipq_nfrags to struct ipq. Maintain count of fragments
(fragments, not fragmented packets) in each queue entry.
Use ipq_nfrags to maintain a count of total fragments in reassembly queue.
|
| 1.63 | 06-Dec-2003 |
jonathan | Replace the single global IP reassembly list/listhead, with a
hashtable of list-heads. Independently re-invented, then reworked to
match similar code in FreeBSD.
|
| 1.62 | 26-Nov-2003 |
itojun | define RANDOM_IP_ID by default (unifdef -DRANDOM_IP_ID).
one use remains in sys/netipsec, which is kept for freebsd source code compat.
|
| 1.61 | 25-Nov-2003 |
itojun | knf
|
| 1.60 | 17-Nov-2003 |
jonathan | Revert the (default) ip_id algorithm to the pre-randomid algorithm,
due to demonstrated low-period repeated IDs from the randomized IP_id
code. Consensus is that the low-period repetition (much less than
2^15) is not suitable for general-purpose use.
Allocators of new IPv4 IDs should now call the function ip_newid().
Randomized IP_ids is now a config-time option, "options RANDOM_IP_ID".
ip_newid() can use ip_random-id()_IP_ID if and only if configured
with RANDOM_IP_ID. A sysctl knob should be provided.
This API may be reworked in the near future to support linear ip_id
counters per (src,dst) IP-address pair.
|
| 1.59 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.58 | 19-Aug-2003 |
itojun | make ip_fragment public (it is for coming PF integration)
|
| 1.57 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.56 | 29-Jun-2003 |
fvdl | branches: 1.56.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.55 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.54 | 23-Jun-2003 |
martin | Protect opt_*.h includes by _KERNEL_OPT
|
| 1.53 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.52 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.51 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.50 | 28-Jan-2003 |
wiz | success, not sucess. Noted by mjl.
|
| 1.49 | 11-Sep-2002 |
itojun | correct signedness mixup in pointer passing. sync w/kame
|
| 1.48 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.47 | 07-May-2002 |
matt | branches: 1.47.2;
Change struct ipqe to use TAILQ's instead of LIST's (primarily for TCP's
benefit currently). Rework tcp_reass code to optimize the 4 most likely causes
of out-of-order packets: first OoO pkt, next OoO pkt in seq, OoO pkt is part
of new chuck of OoO packets, and the OoO pkt fills the first hole. Add evcnts
to instrument tcp_reass (enabled by the options TCP_REASS_COUNTERS). This is
part 1/2 of tcp_reass changes.
|
| 1.46 | 21-Dec-2001 |
itojun | have rip_ctlinput to notify routing changes to raw sockets
(protosw change to be done). sync with kame
|
| 1.45 | 02-Mar-2001 |
itojun | branches: 1.45.2; 1.45.4;
increase ipstat.ips_badaddr if the packet fails to pass address checks.
|
| 1.44 | 13-Jan-2001 |
itojun | allow IP_MULTICAST_IF and IP_ADD/DROP_MEMBERSHIP to specify interface
by interface index. if the interface address specified is in 0.0.0.0/8
it will be considered as interface index in network byteorder.
getsockopt(IP_MULTICAST_IF) preserves old behavior if
setsockopt(IP_MULTICAST_IF) was done with interface address, and
returns interface index if setsockopt(IP_MULTICAST_IF) was done with
interface index (again using the form in 0.0.0.0/8).
Suggested by Dave Thaler, based on RIPv2 MIB spec (RFC1724 section 3.3).
http://mail-index.netbsd.org/tech-net/2001/01/13/0003.html
|
| 1.43 | 17-Oct-2000 |
thorpej | Add an IP_MTUDISC flag to the flags that can be passed to
ip_output(). This flag, if set, causes ip_output() to set
DF in the IP header if the MTU in the route is not locked.
This allows a bunch of redundant code, which I was never
really all that happy about adding in the first place, to
be eliminated.
Inspired by a similar change made by provos@openbsd.org when
he integrated NetBSD's Path MTU Discovery code into OpenBSD.
|
| 1.42 | 25-Aug-2000 |
tron | Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
|
| 1.41 | 30-Mar-2000 |
simonb | branches: 1.41.4;
Delete redundant decl of ip_gif_ttl - it's in <netinet/in_gif.h>.
Delete redundant decl of ip_mforward() - it's in <netinet/ip_mroute.h>.
|
| 1.40 | 20-Nov-1999 |
thorpej | Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.39 | 19-Nov-1999 |
bouyer | Update protocoles and interfaces stats counters to 64bit.
RTM_IFINFO is now 0xf, 0xe is RTM_OIFINFO which returns the old (if_msghdr14)
struct with 32bit counters (binary compat, conditioned on COMPAT_14).
Same for sysctl: node 3 is renamed NET_RT_OIFLIST, NET_RT_IFLIST is now node 4.
Change rt_msg1() to add an mbuf to the mbuf chain instead of just panic()
when the message is larger than MHLEN.
|
| 1.38 | 06-Jul-1999 |
itojun | branches: 1.38.2; 1.38.8;
sync with KAME/NetBSD 1.4, SNAP kit 19990705.
key changes are:
- icmp6 redirect fix (dst check)
- revised ip6 multicast check for loopback i/f
- several RCS ID cleanups
|
| 1.37 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.36 | 08-Oct-1998 |
thorpej | branches: 1.36.8; 1.36.10;
Use the pool allocator for ipflow entries.
|
| 1.35 | 08-Oct-1998 |
thorpej | Use the pool allocator for ipqent structures.
|
| 1.34 | 02-Jun-1998 |
thorpej | In addition to the IP flow hash table, put the flows on a list. The table
is used for fast lookup, the list for traversal of all flows. Also, use
PRT timers.
|
| 1.33 | 11-May-1998 |
thorpej | Back out previous. This problem was already fixed in a different way.
|
| 1.32 | 11-May-1998 |
matt | Let usr.sbin/tcpdump build again.
|
| 1.31 | 04-May-1998 |
matt | Default IP flow to being enabled. Add a sysctl to control the maximum
number of flows (net.inet.ip.maxflows). If set to 0, will disable fast
path forwarding.
|
| 1.30 | 30-Apr-1998 |
thorpej | Need <net/route.h>
|
| 1.29 | 29-Apr-1998 |
matt | Add support for "fast" forwarding. Add hooks in if_ethersubr.c and
if_fddisubr.c to fastpath IP forwarding. If ip_forward successfully
forwards a packet, it will create a cache (ipflow) entry. ether_input
and fddi_input will first call ipflow_fastforward with the received
packet and if the packet passes enough tests, it will be forwarded (the
ttl is decremented and the cksum is adjusted incrementally).
|
| 1.28 | 29-Apr-1998 |
matt | New TCP reassembly code. The new code reduces the memory needed by
out-of-order packets and builds the infrastructure needed for sending
SACK blocks (to be added shortly).
|
| 1.27 | 29-Apr-1998 |
kml | Add support for deletion of routes added by path MTU discovery;
uses new generic route timeout code. Add sysctl for timeout period.
|
| 1.26 | 24-Mar-1998 |
kml | Ensure that we take the IP option length into account when we calculate
the effective maximum send size for TCP. ip_optlen() and tcp_optlen()
should probably be inlined for efficiency.
|
| 1.25 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.24 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.23 | 05-Jan-1998 |
lukem | enhance ephemeral port allocation code:
* support sysctl net.inet.ip.anonportmin (lowest ephemeral port)
and net.inet.ip.anonportmax (highest ephemeral port).
these can't be set to >65535, < IPPORT_RESERVED (unless IPNOPRIVPORTS
is defined), and anonportmin has to be < anonportmax.
* use a cleaner way of only cycling through the available set once;
this will be useful for when a random allocation scheme is used
* define IPPORT_ANON{MIN,MAX} instead of IPPORT_USER{LOW,HIGH}
|
| 1.22 | 18-Oct-1997 |
kml | branches: 1.22.2;
change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.21 | 14-Oct-1997 |
thorpej | Define IP_RETURNMTU. (Matt missed this part of his diff, I guess :-)
|
| 1.20 | 24-Jun-1997 |
thorpej | branches: 1.20.4;
Eliminate use of dtom() from the network code, allowing more flexible
use of mbuf external storage and increasing performance (by eliminating
an m_pullup() for clusters in the IP reassembly code).
Changes from Koji Imada <koji@math.human.nagoya-u.ac.jp>, in PR #3628
and #3480, with ever-so-slight integration changes by me.
|
| 1.19 | 11-Jan-1997 |
thorpej | Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.18 | 25-Oct-1996 |
thorpej | Make length and offset fields unsigned. From Kevin M. Lahey <kml@nas.nasa.gov>
Add a counter to IP stats, to count packets which are discarded on the
grounds that they are too large.
|
| 1.17 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.16 | 13-Feb-1996 |
christos | branches: 1.16.4;
netinet prototypes
|
| 1.15 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.14 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.13 | 14-May-1995 |
cgd | drop (and record) malformed IP fragments. Fixes pr 1030 (differently).
|
| 1.12 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.11 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.7 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.6 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.16.4.2 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.16.4.1 | 10-Nov-1996 |
thorpej | Update from trunk:
- Make ip_len and ip_off unsigned.
- Make sure we don't accept or transmit packets larger than the
maximim IP packet size.
This fixes the so-called `death ping' bug.
Sum of work from Bill Fenner <fenner@parc.xerox.com>,
Kevin Lahey <kml@nas.nasa.gov>, and myself.
Thanks to Curt Sampson, Jukka Marin, and Kevin Lahey for testing
this under NetBSD 1.2
|
| 1.20.4.1 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.22.2.1 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.36.10.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.36.10.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.36.10.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.36.8.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.36.8.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.38.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.38.2.3 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.38.2.2 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.38.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.41.4.1 | 26-Aug-2000 |
tron | Pull up from current (approved by thorpej):
Add new sysctl variables "net.inet.ip.lowportmin" and
"net.inet.ip.lowportmax" which can be used to the set minimum
and maximum port number assigned to sockets using
IP_PORTRANGE_LOW.
syssrc/sys/netinet/in.h 1.49 -> 1.50
syssrc/sys/netinet/in_pcb.c 1.66 -> 1.67
syssrc/sys/netinet/ip_input.c 1.116 -> 1.117
syssrc/sys/netinet/ip_var.h 1.41 -> 1.42
|
| 1.45.4.4 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.45.4.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.45.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.45.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.45.2.4 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.45.2.3 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.45.2.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.45.2.1 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.47.2.1 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.56.2.7 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.56.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.56.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.56.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.56.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.56.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.56.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.69.8.1 | 13-Apr-2005 |
tron | Pull up revision 1.70 (requested by yamt in ticket #145):
when doing TSO, avoid to use duplicated ip_id heavily.
XXX ip_randomid
|
| 1.69.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.72.8.1 | 29-Nov-2005 |
yamt | sync with head.
|
| 1.72.2.6 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.72.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.72.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.72.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.72.2.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.72.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.76.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.76.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.76.2.1 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.77.20.2 | 15-Apr-2007 |
yamt | sync with head.
|
| 1.77.20.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.78.8.1 | 29-Mar-2007 |
reinoud | Pullup to -current
|
| 1.78.6.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.78.4.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.78.4.1 | 10-Apr-2007 |
ad | Sync with head.
|
| 1.79.12.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.79.10.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.79.10.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.79.10.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.79.8.1 | 04-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.80.10.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.80.6.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.80.4.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.84.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.84.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.84.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.87.10.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.87.6.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.87.4.3 | 09-Oct-2010 |
yamt | sync with head
|
| 1.87.4.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.87.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.90.2.1 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.91.6.2 | 31-May-2011 |
rmind | sync with head
|
| 1.91.6.1 | 05-Mar-2011 |
rmind | sync with head
|
| 1.91.4.3 | 06-Nov-2010 |
uebayasi | Sync with HEAD.
|
| 1.91.4.2 | 22-Oct-2010 |
uebayasi | Sync with HEAD (-D20101022).
|
| 1.91.4.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.96.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.97.18.3 | 18-May-2014 |
rmind | sync with head
|
| 1.97.18.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.97.18.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.97.14.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.97.14.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.97.4.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.99.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.107.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.107.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.107.2.5 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.107.2.4 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.107.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.107.2.2 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.107.2.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.114.2.4 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.114.2.3 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.114.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.114.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.116.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.119.6.2 | 11-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #536):
distrib/sets/lists/base/shl.mi: 1.825
distrib/sets/lists/comp/mi: 1.2168-1.2169
distrib/sets/lists/comp/shl.mi: 1.310
distrib/sets/lists/debug/mi: 1.234
distrib/sets/lists/debug/shl.mi: 1.188
distrib/sets/lists/man/mi: 1.1570
distrib/sets/lists/tests/mi: 1.772
etc/mtree/NetBSD.dist.tests: 1.150
share/man/man4/Makefile: 1.650
share/man/man4/ipsec.4: 1.42-1.43
share/man/man4/ipsecif.4: 1.1-1.5
sys/arch/amd64/conf/ALL: 1.77
sys/arch/amd64/conf/GENERIC: 1.480
sys/conf/files: 1.1191
sys/net/Makefile: 1.34
sys/net/files.net: 1.14
sys/net/if.c: 1.404
sys/net/if.h: 1.248
sys/net/if_gif.c: 1.135
sys/net/if_ipsec.c: 1.1-1.3
sys/net/if_ipsec.h: 1.1
sys/net/if_l2tp.c: 1.16
sys/net/if_types.h: 1.28
sys/netinet/in.c: 1.214
sys/netinet/in.h: 1.103
sys/netinet/in_gif.c: 1.92
sys/netinet/ip_var.h: 1.122
sys/netinet6/in6.c: 1.257
sys/netinet6/in6.h: 1.88
sys/netinet6/in6_gif.c: 1.90
sys/netinet6/ip6_var.h: 1.75
sys/netipsec/Makefile: 1.6
sys/netipsec/files.netipsec: 1.13
sys/netipsec/ipsec.h: 1.62
sys/netipsec/ipsecif.c: 1.1
sys/netipsec/ipsecif.h: 1.1
sys/netipsec/key.c: 1.246-1.247
sys/netipsec/key.h: 1.34
sys/rump/net/Makefile.rumpnetcomp: 1.20
sys/rump/net/lib/libipsec/IPSEC.ioconf: 1.1
sys/rump/net/lib/libipsec/Makefile: 1.1
sys/rump/net/lib/libipsec/ipsec_component.c: 1.1
tests/net/Makefile: 1.34
tests/net/if_ipsec/Makefile: 1.1
tests/net/if_ipsec/t_ipsec.sh: 1.1-1.2
Don't touch an SP without a reference to it
unify processing to check nesting count for some tunnel protocols.
add ipsec(4) interface, which is used for route-based VPN.
man and ATF are added later, please see man for details.
reviewed by christos@n.o, joerg@n.o and ozaki-r@n.o, thanks.
https://mail-index.netbsd.org/tech-net/2017/12/18/msg006557.html
ipsec(4) interface supports rump now.
add ipsec(4) interface ATF.
add ipsec(4) interface man as ipsecif.4.
add ipsec(4) interface to amd64/GENERIC and amd64/ALL configs.
apply in{,6}_tunnel_validate() to gif(4).
Spell IPsec that way. Simplify macro usage. Sort SEE ALSO. Bump
date for previous.
Improve wording and macro use.
Some parts are not clear to me, so someone with knowledge of ipsecif(4)
should improve this some more.
Improve ipsecif.4. Default port ipsec(4) NAT-T is tested now.
pointed out by wiz@n.o and suggested by ozaki-r@n.o, thanks.
Change the prefix of test names to ipsecif_ to distinguish from tests for ipsec(4)
New sentence, new line. Remove empty macro.
Fix PR kern/52920. Pointed out by David Binderman, thanks.
Improve wording, and put a new drawing, from me and Kengo Nakahara.
apply a little more #ifdef INET/INET6. fixes !INET6 builds.
|
| 1.119.6.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.122.2.4 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.122.2.3 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.122.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.122.2.1 | 07-Apr-2018 |
pgoyette | Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
|
| 1.125.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.130.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.134.10.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.17 | 01-Oct-2004 |
christos | Move ipf to sys/dist/ipf; Note that I followed the pattern used for pf.
I think though that the files.ipfilter and Makefile glue should go to
the dist directory, not like it is done now.
|
| 1.16 | 23-Jul-2004 |
martti | Upgraded IPFilter to 4.1.3
|
| 1.15 | 28-Mar-2004 |
martti | branches: 1.15.2;
Upgraded IPFilter to 4.1.1
|
| 1.14 | 19-Sep-2002 |
martti | branches: 1.14.6;
Upgraded IPFilter to 3.4.29
|
| 1.13 | 02-May-2002 |
martti | branches: 1.13.4;
Upgraded IPFilter to 3.4.27
|
| 1.12 | 14-Mar-2002 |
martti | Upgraded IPFilter to 3.4.25
|
| 1.11 | 24-Jan-2002 |
martti | Upgraded IPFilter to 3.4.23
|
| 1.10 | 26-Mar-2001 |
mike | branches: 1.10.2;
Resolve conflicts.
|
| 1.9 | 09-Aug-2000 |
veego | branches: 1.9.2; 1.9.4;
Resolve conflicts.
|
| 1.8 | 12-Jun-2000 |
veego | branches: 1.8.2;
Resolve conflicts.
|
| 1.7 | 23-May-2000 |
veego | branches: 1.7.2;
Resolve conflicts.
|
| 1.6 | 21-May-2000 |
veego | Resolve conflicts.
|
| 1.5 | 11-May-2000 |
veego | Resolve conflicts and fix a compile error in ip_ftp_pxy.c.
|
| 1.4 | 03-May-2000 |
veego | Resolve conflicts.
|
| 1.3 | 01-Feb-2000 |
veego | Resolve conflicts.
|
| 1.2 | 28-Dec-1999 |
darrenr | update ipfilter code to 3.3.6
|
| 1.1 | 12-Dec-1999 |
veego | branches: 1.1.1;
Initial revision
|
| 1.1.1.16 | 23-Jul-2004 |
martti | Import IPFilter 4.1.3
|
| 1.1.1.15 | 28-Mar-2004 |
martti | Import IPFilter 4.1.1
|
| 1.1.1.14 | 19-Sep-2002 |
martti | Import IPFilter 3.4.29
|
| 1.1.1.13 | 02-May-2002 |
martti | Import IPFilter 3.4.27
|
| 1.1.1.12 | 14-Mar-2002 |
martti | Import IPFilter 3.4.25
|
| 1.1.1.11 | 24-Jan-2002 |
martti | Import IPFilter 3.4.23
|
| 1.1.1.10 | 26-Mar-2001 |
mike | Import IP Filter 3.4.16
|
| 1.1.1.9 | 09-Aug-2000 |
veego | Import IP Filter 3.4.9
|
| 1.1.1.8 | 12-Jun-2000 |
veego | Import IP Filter 3.4.6
|
| 1.1.1.7 | 23-May-2000 |
veego | Import IP Filter 3.4.4
|
| 1.1.1.6 | 21-May-2000 |
veego | Import IP Filter 3.4.3
|
| 1.1.1.5 | 11-May-2000 |
veego | Import IP Filter 3.4.2
|
| 1.1.1.4 | 03-May-2000 |
veego | Import IP Filter 3.4.1
|
| 1.1.1.3 | 01-Feb-2000 |
veego | Import IP Filter 3.3.8
|
| 1.1.1.2 | 28-Dec-1999 |
darrenr | update DARRENR branch of netinet to 3.3.6
|
| 1.1.1.1 | 12-Dec-1999 |
veego | branches: 1.1.1.1.2; 1.1.1.1.4;
Import a few IP Filter 3.3.5 files under sys/netinet.
|
| 1.1.1.1.4.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.1.1.1.2.3 | 08-Jan-2000 |
he | Pull up revision 1.2 (requested by darrenr):
Update IPF to version 3.3.6.
|
| 1.1.1.1.2.2 | 20-Dec-1999 |
he | Pull up revision 1.1.1.1 (new) (requested by darrenr):
Update IPF to version 3.3.5.
|
| 1.1.1.1.2.1 | 12-Dec-1999 |
he | file ipl.h was added on branch netbsd-1-4 on 1999-12-20 21:02:07 +0000
|
| 1.7.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.8.2.3 | 18-Oct-2002 |
itojun | dist/ipf/BNF 1.5
dist/ipf/HISTORY 1.12-1.14
dist/ipf/Makefile 1.6
dist/ipf/QNX_OCL.txt 1.1 (new)
dist/ipf/common.c 1.1.1.5,1.2
dist/ipf/facpri.c 1.4
dist/ipf/fils.c 1.14-1.21
dist/ipf/ip_lfil.c deleted
dist/ipf/ip_sfil.c deleted
dist/ipf/ipf.c 1.8-1.13
dist/ipf/ipf2netbsd 1.6-1.9
dist/ipf/ipfs.c 1.6-1.10
dist/ipf/ipft_ef.c 1.4-1.7
dist/ipf/ipft_hx.c 1.4-1.5
dist/ipf/ipft_pc.c 1.4-1.5
dist/ipf/ipft_sn.c 1.4-1.5
dist/ipf/ipft_td.c 1.4-1.7
dist/ipf/ipft_tx.c 1.5-1.8
dist/ipf/iplang/iplang_y.y 1.4
dist/ipf/ipmon.c 1.8-1.17
dist/ipf/ipnat.c 1.9-1.12
dist/ipf/ipsend/44arp.c 1.3
dist/ipf/ipsend/arp.c 1.3
dist/ipf/ipsend/in_var.h 1.2
dist/ipf/ipsend/ip.c 1.4-1.5
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipresend.c 1.3-1.4
dist/ipf/ipsend/ipsend.c 1.5-1.8
dist/ipf/ipsend/ipsopt.c 1.3-1.4
dist/ipf/ipsend/iptest.c 1.4-1.5
dist/ipf/ipsend/iptests.c 1.3-1.5
dist/ipf/ipsend/lsock.c 1.3
dist/ipf/ipsend/resend.c 1.4-1.5
dist/ipf/ipsend/sbpf.c 1.3
dist/ipf/ipsend/sirix.c 1.3
dist/ipf/ipsend/sock.c 1.4-1.5
dist/ipf/ipt.c 1.5-1.10
dist/ipf/kmem.c 1.5-1.10
dist/ipf/l4check/l4check.c 1.1.1.2
dist/ipf/man/ipf.4 1.8-1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipl.4 1.4
dist/ipf/man/ipmon.8 1.8-1.10
dist/ipf/man/ipnat.4 1.3
dist/ipf/man/ipnat.5 1.5-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.4-1.7
dist/ipf/natparse.c 1.6-1.10
dist/ipf/opt.c 1.4-1.5
dist/ipf/parse.c 1.11-1.13
dist/ipf/printnat.c 1.3-1.10
dist/ipf/printstate.c 1.2-1.3
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
sys/netinet/fil.c 1.53-1.58
sys/netinet/ip_auth.c 1.25-1.30
sys/netinet/ip_compat.h 1.27-1.31
sys/netinet/ip_fil.c 1.76-1.79,1.81-1.86
sys/netinet/ip_fil.h 1.43-1.49
sys/netinet/ip_frag.c 1.27-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.22-1.26
sys/netinet/ip_h323_pxy.c 1.7 (new)
sys/netinet/ip_ipsec_pxy.c 1.2
sys/netinet/ip_log.c 1.18-1.23
sys/netinet/ip_nat.c 1.45-1.54
sys/netinet/ip_nat.h 1.25-1.27
sys/netinet/ip_netbios_pxy.c 1.2-1.4
sys/netinet/ip_proxy.c 1.28-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.35-1.42
sys/netinet/ip_state.h 1.21-1.23
sys/netinet/ipl.h 1.12-1.14
usr.sbin/ipf/ipftest/Makefile 1.19
Upgrade IPFilter to 3.4.29.
regression test is omitted. (martti)
|
| 1.8.2.2 | 09-Feb-2002 |
he | Pull up revisions 1.10-1.11 (requested by martti):
Updated IPFilter to 3.4.23
|
| 1.8.2.1 | 31-Aug-2000 |
veego | Pull up ipf 3.4.9 (requested by veego). approved by releng-1-5.
basesrc/dist/ipf/HISTORY 1.8 -> 1.9
basesrc/dist/ipf/fils.c 1.9 -> 1.10
basesrc/dist/ipf/ip_sfil.c 1.5 -> 1.6
basesrc/dist/ipf/ipf.c 1.4 -> 1.5
basesrc/dist/ipf/ipmon.c 1.4 -> 1.5
basesrc/dist/ipf/ipnat.c 1.5 -> 1.6
basesrc/dist/ipf/natparse.c 1.3 -> 1.4
basesrc/dist/ipf/parse.c 1.4 -> 1.5
basesrc/dist/ipf/iplang/iplang_y.y 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.1 1.2 -> 1.3
basesrc/dist/ipf/ipsend/ipsend.5 1.1 -> 1.2
syssrc/sys/netinet/fil.c 1.36 -> 1.37
syssrc/sys/netinet/ip_auth.c 1.17 -> 1.18
syssrc/sys/netinet/ip_fil.c 1.57 -> 1.58
syssrc/sys/netinet/ip_ftp_pxy.c 1.16 -> 1.17
syssrc/sys/netinet/ip_log.c 1.10 -> 1.11
syssrc/sys/netinet/ip_nat.c 1.34 -> 1.35
syssrc/sys/netinet/ip_nat.h 1.20 -> 1.21
syssrc/sys/netinet/ip_rcmd_pxy.c 1.4 -> 1.5
syssrc/sys/netinet/ip_state.c 1.26 -> 1.27
syssrc/sys/netinet/ip_state.h 1.16 -> 1.17
syssrc/sys/netinet/ipl.h 1.8 -> 1.9
Changes:
>3.4.9 08/08/2000 - Released
>
>implement new aging mechanism in fr_tcp_age()
>
>fix icmp state checking bug
>
>revamp buildsunos script and build both sparcv7/sparcv9 for Solaris
>if on an Ultra with a 64bit system & compiler (Caseper Dik)
>
>open ipfilter device read only if we know we can
>
>print out better information for ICMP packets in ipmon
>
>move checking for source spoofed packets to a point where we can generate
>logs of them
>
>return EFAULT from ircopyptr/iwcopyptr
>
>don't do ioctl(SIOCGETFS) for auth stats
>
>fix up freeing mbufs for post-4.3BSD
>
>fix returning of inc from ftp proxy
>
>fix bugs with ipfs -R/-W (Caseper Dik)
>
>3.4.8 19/07/2000 - Released
>
>create fake opt_inet6.h for FreeBSD-4 compile as LKM
>
>add #ifdef's for KLD_MODULE sanity
>
>NAT fastroute'd packets which come out of return-*
>
>fix upper/lower case crap in ftp proxy and get seq# checking fixed up.
>
>3.4.7 08/07/2000 - Released
>
>make "ipf -y" lookup NAT if's which are unknown
>
>prepend line numbers to ioctl error messages in ipf/ipnat
>
>don't apply patches to FreeBSD twice
>
>allow for ip_len to be on an unaligned boundary early on in fr_precheck
>
>fix printing of icmp code when it is 0
>
>correct printing of port numbers in map rules with from/to
>
>don't allow fr_func to be called at securelevel > 0 or rules to be added
>if securelevel > 0 if they have a non-zero fr_func.
|
| 1.9.4.5 | 20-Sep-2002 |
thorpej | Sync with HEAD.
|
| 1.9.4.4 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.9.4.3 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.9.4.2 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.9.4.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.9.2.3 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.9.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.9.2.1 | 09-Aug-2000 |
bouyer | file ipl.h was added on branch thorpej_scsipi on 2000-11-20 18:10:34 +0000
|
| 1.10.2.3 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.10.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.10.2.1 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.13.4.1 | 24-Oct-2002 |
lukem | Pull up upgrade to IPfilter 3.4.29 (requested by martti in ticket #905).
Affected files & revisions:
dist/ipf/HISTORY 1.14
dist/ipf/fils.c 1.17-1.21
dist/ipf/ipf.c 1.11-1.13
dist/ipf/ipfs.c 1.8-1.10
dist/ipf/ipft_ef.c 1.6-1.7
dist/ipf/ipft_td.c 1.6-1.7
dist/ipf/ipft_tx.c 1.7-1.8
dist/ipf/ipmon.c 1.12-1.17
dist/ipf/ipnat.c 1.11-1.12
dist/ipf/ipsend/ip_var.h 1.2
dist/ipf/ipsend/ipsend.c 1.8
dist/ipf/ipsend/iptests.c 1.5
dist/ipf/ipt.c 1.8-1.10
dist/ipf/kmem.c 1.8-1.10
dist/ipf/man/ipf.4 1.10
dist/ipf/man/ipf.5 1.8
dist/ipf/man/ipftest.1 1.3
dist/ipf/man/ipmon.8 1.10
dist/ipf/man/ipnat.5 1.9-1.10
dist/ipf/man/ipnat.8 1.4
dist/ipf/misc.c 1.7
dist/ipf/natparse.c 1.10
dist/ipf/parse.c 1.13
dist/ipf/printnat.c 1.8-1.10
dist/ipf/relay.c 1.5-1.6
dist/ipf/rules/example.9 1.2
etc/rc.d/ipnat 1.8
regress/sys/kern/ipf/Makefile 1.3-1.4
regress/sys/kern/ipf/dotest6 1.2
regress/sys/kern/ipf/expected/f13 1.1.1.2
regress/sys/kern/ipf/expected/i12 1.1.1.1
regress/sys/kern/ipf/expected/ni3 1.1.1.1
regress/sys/kern/ipf/expected/ni5 1.2
regress/sys/kern/ipf/input/f13 1.1.1.2
regress/sys/kern/ipf/input/ipv6.1 1.1.1.1
regress/sys/kern/ipf/input/ni3 1.1.1.1
regress/sys/kern/ipf/regress/i12 1.1.1.1
regress/sys/kern/ipf/regress/ipv6.1 1.1.1.1
regress/sys/kern/ipf/regress/ni3.ipf 1.1.1.1
regress/sys/kern/ipf/regress/ni3.nat 1.1.1.1
sys/arch/alpha/conf/ALPHA 1.169,1.171
sys/arch/amiga/conf/GENERIC 1.185-1.186
sys/arch/arc/conf/GENERIC 1.71-1.72
sys/arch/atari/conf/GENERIC.in 1.24-1.25
sys/arch/cats/conf/GENERIC 1.31-1.32
sys/arch/cobalt/conf/GENERIC 1.34-1.35
sys/arch/hp300/conf/GENERIC 1.83-1.84
sys/arch/i386/conf/CARDBUS 1.66-1.67
sys/arch/i386/conf/GENERIC 1.510,1.512
sys/arch/i386/conf/GENERIC_LAPTOP 1.58-1.59
sys/arch/i386/conf/GENERIC_PS2TINY 1.19-1.20
sys/arch/i386/conf/GENERIC_TINY 1.47-1.48
sys/arch/luna68k/conf/GENERIC 1.33-1.33
sys/arch/mac68k/conf/GENERIC 1.130-1.131
sys/arch/mac68k/conf/GENERICSBC 1.21-1.22
sys/arch/mac68k/conf/SMALLRAM 1.4-1.5
sys/arch/macppc/conf/GENERIC 1.142-1.143
sys/arch/mipsco/conf/GENERIC 1.21-1.22
sys/arch/mmeye/conf/GENERIC 1.44-1.45
sys/arch/news68k/conf/GENERIC 1.36-1.37
sys/arch/news68k/conf/GENERIC_TINY 1.18-1.19
sys/arch/newsmips/conf/GENERIC 1.50-1.51
sys/arch/ofppc/conf/GENERIC 1.56-1.57
sys/arch/pmax/conf/GENERIC 1.103-1.104
sys/arch/prep/conf/GENERIC 1.55-1.56
sys/arch/sbmips/conf/GENERIC 1.11-1.12
sys/arch/sgimips/conf/GENERIC 1.7-1.8
sys/arch/sparc/conf/GENERIC 1.138-1.139
sys/arch/sparc64/conf/GENERIC32 1.46-1.47
sys/arch/vax/conf/GENERIC 1.102-1.103
sys/arch/x68k/conf/ALL 1.55-1.56
sys/arch/x68k/conf/GENERIC 1.80-1.81
sys/lkm/netinet/if_ipl/mln_ipl.c 1.29
sys/netinet/fil.c 1.57-1.58
sys/netinet/ip_auth.c 1.29-1.30
sys/netinet/ip_compat.h 1.30-1.31
sys/netinet/ip_fil.c 1.81-1.86
sys/netinet/ip_fil.h 1.46-1.49
sys/netinet/ip_frag.c 1.33-1.34
sys/netinet/ip_frag.h 1.18
sys/netinet/ip_ftp_pxy.c 1.25-1.26
sys/netinet/ip_h323_pxy.c 1.5-1.6
sys/netinet/ip_log.c 1.22-1.23
sys/netinet/ip_nat.c 1.51-1.53
sys/netinet/ip_nat.h 1.27
sys/netinet/ip_netbios_pxy.c 1.4
sys/netinet/ip_proxy.c 1.35-1.36
sys/netinet/ip_proxy.h 1.18
sys/netinet/ip_state.c 1.41-1.42
sys/netinet/ip_state.h 1.23
sys/netinet/ipl.h 1.14
|
| 1.14.6.2 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.14.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.15.2.1 | 13-Aug-2004 |
jmc | branches: 1.15.2.1.2;
Pullup rev 1.16 (requested by christos in ticket #1727)
Sync up w. ipf 4.1.3
|
| 1.15.2.1.2.1 | 06-Feb-2005 |
jmc | Pull up revision 1.17 (requested by martti in ticket #1086)
Move ipf to sys/dist/ipf and sync w. trunk
|
| 1.2 | 10-Dec-2005 |
elad | Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.1 | 04-Sep-2004 |
manu | branches: 1.1.2; 1.1.14;
IPv4 PIM support, based on submission from Pavlin Radoslavov on tech-net@ :
two new files I forgot to add on the first cvs commit.
|
| 1.1.14.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.1.2.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.1.2.4 | 21-Sep-2004 |
skrll | SYNC WITH HEAD.
|
| 1.1.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.2.1 | 04-Sep-2004 |
skrll | file pim.h was added on branch ktrace-lwp on 2004-09-18 14:54:54 +0000
|
| 1.4 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.3 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.2 | 10-Dec-2005 |
elad | branches: 1.2.162; 1.2.164;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.1 | 04-Sep-2004 |
manu | branches: 1.1.2; 1.1.14;
IPv4 PIM support, based on submission from Pavlin Radoslavov on tech-net@ :
two new files I forgot to add on the first cvs commit.
|
| 1.1.14.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.1.2.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.1.2.4 | 21-Sep-2004 |
skrll | SYNC WITH HEAD.
|
| 1.1.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.1.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.1.2.1 | 04-Sep-2004 |
skrll | file pim_var.h was added on branch ktrace-lwp on 2004-09-18 14:54:54 +0000
|
| 1.2.164.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.2.162.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.2.162.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.15 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to in6pcb_*
|
| 1.14 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.13 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.12 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.11 | 11-Jan-2017 |
ozaki-r | Get rid of unnecessary header inclusions
|
| 1.10 | 26-Apr-2016 |
ozaki-r | branches: 1.10.2;
Sweep unnecessary route.h inclusions
|
| 1.9 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.8 | 10-Feb-2015 |
rjs | Add DCCP protocol support from KAME.
|
| 1.7 | 02-Dec-2014 |
christos | use the new printing code.
|
| 1.6 | 08-Sep-2014 |
joerg | branches: 1.6.2;
Always use cprng_fast32, even during initialisation. No point in using
random(9).
|
| 1.5 | 01-Jun-2013 |
pooka | branches: 1.5.2;
Give portalgo a compile-time override; for cases where the default default
doesn't make enough sense to even consider it (a lot of outgoing connections
from rump kernels with local port 65535).
|
| 1.4 | 07-Dec-2012 |
christos | use __BITMAP_TYPE
|
| 1.3 | 01-Dec-2012 |
christos | switch from fd_set to using bitmap macros
|
| 1.2 | 29-Nov-2012 |
christos | Add a new sysctl to mark ports as reserved, so that they are not used in
the anonymous or reserved port allocation.
|
| 1.1 | 25-Jun-2012 |
christos | branches: 1.1.2; 1.1.4;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.1.4.4 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.1.4.3 | 16-Jan-2013 |
yamt | sync with (a bit old) head
|
| 1.1.4.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.1.4.1 | 25-Jun-2012 |
yamt | file portalgo.c was added on branch yamt-pagecache on 2012-10-30 17:22:46 +0000
|
| 1.1.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.1.2.1 | 25-Feb-2013 |
tls | resync with head
|
| 1.5.2.2 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.5.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.6.2.4 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.6.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.6.2.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.6.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.10.2.1 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.3 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.2 | 29-Nov-2012 |
christos | Add a new sysctl to mark ports as reserved, so that they are not used in
the anonymous or reserved port allocation.
|
| 1.1 | 25-Jun-2012 |
christos | branches: 1.1.2; 1.1.4;
rename rfc6056 -> portalgo, requested by yamt
|
| 1.1.4.3 | 16-Jan-2013 |
yamt | sync with (a bit old) head
|
| 1.1.4.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.1.4.1 | 25-Jun-2012 |
yamt | file portalgo.h was added on branch yamt-pagecache on 2012-10-30 17:22:46 +0000
|
| 1.1.2.1 | 25-Feb-2013 |
tls | resync with head
|
| 1.187 | 20-Jun-2025 |
roy | inet: respect IP_TOS and IP_TTL for raw sockets
|
| 1.186 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.185 | 29-Jun-2024 |
riastradh | branches: 1.185.2;
netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.184 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.183 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.182 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.181 | 13-Jun-2022 |
knakahara | Refactor like in_pcb.c:r1.187 and in6_pcb.c:r1.168.
Use TAILQ_FOREACH instead of TAILQ_FOREACH_SAFE about inpt_queue.
rip_pcbnotify() doesn't use "ninph" pointer and doesn't remove elements.
|
| 1.180 | 08-Sep-2020 |
christos | Add IP_BINDANY, IPV6_BINDANY which can be used to bind to any address in
order to implement transparent proxies.
|
| 1.179 | 24-Feb-2019 |
maxv | RIP, RIP6, DDP, SCTP and SCTP6 lack a length check in their _connect()
functions. Fix the first three, and add a big XXX in the SCTP ones.
Found by KASAN, triggered by SyzKaller.
Reported-by: syzbot+9eaf98dad6ca738c250d@syzkaller.appspotmail.com
|
| 1.178 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.177 | 11-May-2018 |
maxv | branches: 1.177.2;
Make sure we have at least an IP header, and remove pointless XXXs (there
is no issue).
|
| 1.176 | 28-Apr-2018 |
maxv | Remove unused ipsec_var.h includes.
|
| 1.175 | 12-Apr-2018 |
maxv | Make 'opts' local to rip_sbappendaddr().
|
| 1.174 | 12-Apr-2018 |
maxv | Synchronize the code between raw_ip6.c<->icmp6.c<->raw_ip.c, so that it is
the same everywhere.
|
| 1.173 | 12-Apr-2018 |
maxv | Remove misleading comment; we're just checking the SP, not verifying the
AH/ESP payload. While here style a bit.
|
| 1.172 | 21-Mar-2018 |
roy | Sprinkle more soroverflow().
|
| 1.171 | 28-Feb-2018 |
maxv | branches: 1.171.2;
Remove unused ipsec_private.h includes.
|
| 1.170 | 28-Feb-2018 |
maxv | (just forgot to commit this file, the message was)
Remove duplicate IPSEC_STATINC(IPSEC_STAT_IN_POLVIO), ipsec_in_reject
already increases it. IPSEC6_STATINC is now unused, so remove it too.
|
| 1.169 | 26-Feb-2018 |
maxv | Dedup: merge ipsec4_in_reject and ipsec6_in_reject into ipsec_in_reject.
While here fix misleading comment.
ok ozaki-r@
|
| 1.168 | 14-Feb-2018 |
christos | join lines where they fit.
|
| 1.167 | 11-Dec-2017 |
ryo | As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.166 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.165 | 06-Jul-2017 |
christos | Merge the two copies SO_TIMESTAMP/SO_OTIMESTAMP processing to a single
function, and add a SOOPT_TIMESTAMP define reducing compat pollution from
5 places to 1.
|
| 1.164 | 20-Apr-2017 |
ozaki-r | branches: 1.164.4;
Remove unnecessary NULL checks for inp_socket and in6p_socket
They cannot be NULL except for programming errors.
|
| 1.163 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.162 | 24-Jan-2017 |
ozaki-r | Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.161 | 29-Sep-2016 |
roy | branches: 1.161.2;
Now that we disallow sending or receiving from invalid addresses,
allow binding to tentative addresses.
|
| 1.160 | 26-Aug-2016 |
roy | Allow bind to detached INET addresses.
|
| 1.159 | 01-Aug-2016 |
ozaki-r | Apply pserialize and psref to struct ifaddr and its variants
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
|
| 1.158 | 12-May-2016 |
ozaki-r | branches: 1.158.2;
Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.157 | 26-Apr-2016 |
ozaki-r | Sweep unnecessary route.h inclusions
|
| 1.156 | 21-Jan-2016 |
riastradh | Revert previous: ran cvs commit when I meant cvs diff. Sorry!
Hit up-arrow one too few times.
|
| 1.155 | 21-Jan-2016 |
riastradh | Give proper prototype to ip_output.
|
| 1.154 | 20-Jan-2016 |
riastradh | Give proper prototype to rip_output.
|
| 1.153 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.152 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.151 | 02-May-2015 |
roy | Add IPv4 address flags IN_IFF_TENTATIVE, IN_IFF_DUPLICATED and
IN_IFF_DETATCHED to mimic the IPv6 address behaviour.
Add SIOCGIFAFLAG_IN ioctl to retrieve the address flag via the
ifreq structure.
Add IPv4 DAD detection via the ARP methods described in RFC 5227.
Add sysctls net.inet.ip.dad_count and net.inet.arp.debug.
Discussed on tech-net@
|
| 1.150 | 26-Apr-2015 |
rtr | remove pr_generic from struct pr_usrreqs and all implementations of
pr_generic in protocols.
bump to 7.99.13
approved by rmind@
|
| 1.149 | 25-Apr-2015 |
rtr | make rip_connect_pcb take sockaddr_in * instead of mbuf *
make rip_connect_pcb static since it appears to be used only in raw_ip.c
moves m_len check to callers which is a small duplication of code
that will go away when the callers are converted to receive sockaddr *.
|
| 1.148 | 24-Apr-2015 |
rtr | make accept, getsockname and getpeername syscalls use sockaddr_big and modify
pr_{accept,sockname,peername} nam parameter type from mbuf * to sockaddr *.
* retained use of mbuftypes[MT_SONAME] for now.
* bump to netbsd version 7.99.12 for parameter type change.
patch posted to tech-net@ 2015/04/19
|
| 1.147 | 03-Apr-2015 |
rtr | * change pr_bind to accept struct sockaddr * instead of struct mbuf *
* update protocol bind implementations to use/expect sockaddr *
instead of mbuf *
* introduce sockaddr_big struct for storage of addr data passed via
sys_bind; sockaddr_big is of sufficient size and alignment to
accommodate all addr data sizes received.
* modify sys_bind to allocate sockaddr_big instead of using an mbuf.
* bump kernel version to 7.99.9 for change to pr_bind() parameter type.
Patch posted to tech-net@
http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html
The choice to use a new structure sockaddr_big has been retained since
changing sockaddr_storage size would lead to unnecessary ABI change. The
use of the new structure does not preclude future work that increases
the size of sockaddr_storage and at that time sockaddr_big may be
trivially replaced.
Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
|
| 1.146 | 10-Nov-2014 |
maxv | branches: 1.146.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.145 | 09-Aug-2014 |
rtr | branches: 1.145.2;
split PRU_CONNECT2 & PRU_PURGEIF function out of pr_generic() usrreq
switches and put into separate functions
- always KASSERT(solocked(so)) even if not implemented
(for PRU_CONNECT2 only)
- replace calls to pr_generic() with req = PRU_CONNECT2 with calls to
pr_connect2()
- replace calls to pr_generic() with req = PRU_PURGEIF with calls to
pr_purgeif()
put common code from unp_connect2() (used by unp_connect() into
unp_connect1() and call out to it when needed
patch only briefly reviewed by rmind@
|
| 1.144 | 08-Aug-2014 |
rtr | split PRU_RCVD function out of pr_generic() usrreq switches and put into
separate functions
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_RCVD with calls to
pr_rcvd()
|
| 1.143 | 05-Aug-2014 |
rtr | split PRU_SEND function out of pr_generic() usrreq switches and put into
separate functions
xxx_send(struct socket *, struct mbuf *, struct mbuf *,
struct mbuf *, struct lwp *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_SEND with calls to
pr_send()
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_send() PRUs
- l2cap_send() -> l2cap_send_pcb()
- sco_send() -> sco_send_pcb()
- rfcomm_send() -> rfcomm_send_pcb()
patch reviewed by rmind
|
| 1.142 | 05-Aug-2014 |
rtr | revert the removal of struct lwp * parameter from bind, listen and connect
user requests.
this should resolve the issue relating to nfs client hangs presented
recently by wiz on current-users@
|
| 1.141 | 03-Aug-2014 |
rtr | req cannot be PRU_SENDOOB here as per KASSERT() earlier in the
rip_usrreq() function.
- KASSERT(!control || (req == PRU_SEND || req == PRU_SENDOOB));
+ KASSERT(!control || (req == PRU_SEND));
|
| 1.140 | 02-Aug-2014 |
rtr | restore splsoftnet() in various usrreqs that were removed during the PRU
splits. we will properly review removal after the PRU split work is
complete.
|
| 1.139 | 31-Jul-2014 |
rtr | split PRU_DISCONNECT, PRU_SHUTDOWN and PRU_ABORT function out of
pr_generic() usrreq switches and put into separate functions
xxx_disconnect(struct socket *)
xxx_shutdown(struct socket *)
xxx_abort(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req =
PRU_{DISCONNECT,SHUTDOWN,ABORT}
with calls to pr_{disconnect,shutdown,abort}() respectively
rename existing internal functions used to implement above functionality
to permit use of the names for xxx_{disconnect,shutdown,abort}().
- {l2cap,sco,rfcomm}_disconnect() ->
{l2cap,sco,rfcomm}_disconnect_pcb()
- {unp,rip,tcp}_disconnect() -> {unp,rip,tcp}_disconnect1()
- unp_shutdown() -> unp_shutdown1()
patch reviewed by rmind
|
| 1.138 | 31-Jul-2014 |
ozaki-r | Define IFNET_EMPTY() and replace !IFNET_FIRST() with it
No functional change.
|
| 1.137 | 30-Jul-2014 |
rtr | split PRU_CONNECT function out of pr_generic() usrreq switches and put
into seaparate functions
xxx_listen(struct socket *, struct mbuf *)
- always KASSERT(solocked(so)) and KASSERT(nam != NULL)
- replace calls to pr_generic() with req = PRU_CONNECT with
pr_connect()
- rename existin {l2cap,sco,rfcomm}_connect() to
{l2cap,sco,rfcomm}_connect_pcb() respectively to permit
naming consistency with other protocols functions.
- drop struct lwp * parameter from unp_connect() and at_pcbconnect()
and use curlwp instead where appropriate.
patch reviewed by rmind
|
| 1.136 | 24-Jul-2014 |
rtr | split PRU_BIND and PRU_LISTEN function out of pr_generic() usrreq
switches and put into separate functions
xxx_bind(struct socket *, struct mbuf *)
xxx_listen(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_BIND with call to
pr_bind()
- replace calls to pr_generic() with req = PRU_LISTEN with call to
pr_listen()
- drop struct lwp * parameter from at_pcbsetaddr(), in_pcbbind() and
unp_bind() and always use curlwp.
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_{bind,listen}() PRUs
- l2cap_{bind,listen}() -> l2cap_{bind,listen}_pcb()
- sco_{bind,listen}() -> sco_{bind,listen}_pcb()
- rfcomm_{bind,listen}() -> rfcomm_{bind,listen}_pcb()
patch reviewed by rmind
welcome to netbsd 6.99.48
|
| 1.135 | 23-Jul-2014 |
rtr | split PRU_SENDOOB and PRU_RCVOOB function out of pr_generic() usrreq
switches and put into separate functions
xxx_sendoob(struct socket *, struct mbuf *, struct mbuf *)
xxx_recvoob(struct socket *, struct mbuf *, int)
- always KASSERT(solocked(so)) even if request is not implemented
- replace calls to pr_generic() with req = PRU_{SEND,RCV}OOB with
calls to pr_{send,recv}oob() respectively.
there is still some tweaking of m_freem(m) and m_freem(control) to come
for consistency. not performed with this commit for clarity.
reviewed by rmind
|
| 1.134 | 14-Jul-2014 |
rtr | fix fat fingered KASSERT(solocked(0)) -> KASSERT(solocked(so)) mistake.
spotted by Takahiro HAYASHI
|
| 1.133 | 09-Jul-2014 |
rtr | * split PRU_ACCEPT function out of pr_generic() usrreq switches and put
into a separate function xxx_accept(struct socket *, struct mbuf *)
note: future cleanup will take place to remove struct mbuf parameter
type and replace it with a more appropriate type.
patch reviewed by rmind
|
| 1.132 | 09-Jul-2014 |
rtr | * split PRU_PEERADDR and PRU_SOCKADDR function out of pr_generic()
usrreq switches and put into separate functions
xxx_{peer,sock}addr(struct socket *, struct mbuf *).
- KASSERT(solocked(so)) always in new functions even if request
is not implemented
- KASSERT(pcb != NULL) and KASSERT(nam) if the request is
implemented and not for tcp.
* for tcp roll #ifdef KPROF and #ifdef DEBUG code from tcp_usrreq() into
easier to cut & paste functions tcp_debug_capture() and
tcp_debug_trace()
- functions provided by rmind
- remaining use of PRU_{PEER,SOCK}ADDR #define to be removed in a
future commit.
* rename netbt functions to permit consistency of pru function names
(as has been done with other requests already split out).
- l2cap_{peer,sock}addr() -> l2cap_{peer,sock}_addr_pcb()
- rfcomm_{peer,sock}addr() -> rfcomm_{peer,sock}_addr_pcb()
- sco_{peer,sock}addr() -> sco_{peer,sock}_addr_pcb()
* split/refactor do_sys_getsockname(lwp, fd, which, nam) into
two functions do_sys_get{peer,sock}name(fd, nam).
- move PRU_PEERADDR handling into do_sys_getpeername() from
do_sys_getsockname()
- have svr4_stream directly call do_sys_get{sock,peer}name()
respectively instead of providing `which' & fix a DPRINTF string
that incorrectly wrote "getpeername" when it meant "getsockname"
- fix sys_getpeername() and sys_getsockname() to call
do_sys_get{sock,peer}name() without `which' and `lwp' & adjust
comments
- bump kernel version for removal of lwp & which parameters from
do_sys_getsockname()
note: future cleanup to remove struct mbuf * abuse in
xxx_{peer,sock}name()
still to come, not done in this commit since it is easier to do post
split.
patch reviewed by rmind
welcome to 6.99.47
|
| 1.131 | 07-Jul-2014 |
rtr | * sprinkle KASSERT(solocked(so)); in all pr_stat() functions.
* fix remaining inconsistent struct socket parameter names.
|
| 1.130 | 07-Jul-2014 |
rtr | backout change that made pr_stat return EOPNOTSUPP for protocols that
were not filling in struct stat.
decision made after further discussion with rmind and investigation of
how other operating systems behave. soo_stat() is doing just enough to
be able to call what gets returned valid and thus justifys a return of
success.
additional review will be done to determine of the pr_stat functions
that were already returning EOPNOTSUPP can be considered successful with
what soo_stat() is doing.
|
| 1.129 | 07-Jul-2014 |
rtr | * have pr_stat return EOPNOTSUPP consistently for all protocols that do
not fill in struct stat instead of returning success.
* in pr_stat remove all checks for non-NULL so->so_pcb except where the
pcb is actually used (i.e. cases where we don't return EOPNOTSUPP).
proposed on tech-net@
|
| 1.128 | 06-Jul-2014 |
rtr | * split PRU_SENSE functionality out of xxx_usrreq() switches and place into
separate xxx_stat(struct socket *, struct stat *) functions.
* replace calls using pr_generic with req == PRU_SENSE with pr_stat().
further change will follow that cleans up the pattern used to extract the
pcb and test for its presence.
reviewed by rmind
|
| 1.127 | 01-Jul-2014 |
rtr | fix parameter types in pr_ioctl, called xx_control() functions and remove
abuse of pointer to struct mbuf type.
param2 changed to u_long type and uses parameter name 'cmd' (ioctl command)
param3 changed to void * type and uses parameter name 'data'
param4 changed to struct ifnet * and uses parameter name 'ifp'
param5 has been removed (formerly struct lwp *) and uses of 'l' have been
replaced with curlwp from curproc(9).
callers have had (now unnecessary) casts to struct mbuf * removed, called
code has had (now unnecessary) casts to u_long, void * and struct ifnet *
respectively removed.
reviewed by rmind@
|
| 1.126 | 23-Jun-2014 |
rtr | where appropriate rename xxx_ioctl() struct mbuf * parameters from
`control' to `ifp' after split from xxx_usrreq().
sys_socket.c
fix wrapping of arguments to be consistent with other function calls
in the file after replacing pr_usrreq() call with pr_ioctl() which
required one less argument.
link_proto.c
fix indentation of parameters in link_ioctl() prototype to be
consistent with the rest of the file.
discussed with rmind@
|
| 1.125 | 22-Jun-2014 |
rtr | * split PRU_CONTROL functionality out of xxx_userreq() switches and place
into separate xxx_ioctl() functions.
* place KASSERT(req != PRU_CONTROL) inside xxx_userreq() as it is now
inappropriate for req = PRU_CONTROL in xxx_userreq().
* replace calls to pr_generic() with req = PRU_CONTROL with pr_ioctl().
* remove & fixup references to PRU_CONTROL xxx_userreq() function comments.
* fix various comments references for xxx_userreq() that mentioned
PRU_CONTROL as xxx_userreq() no longer handles the request.
a further change will follow to fix parameter and naming inconsistencies
retained from original code.
Reviewed by rmind@
|
| 1.124 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.123 | 22-May-2014 |
rmind | - Make ip_setmoptions(), ip_getmoptions() and ip_pcbopts() static.
- ip_output: eliminate 7th variadic argument; IP_RETURNMTU is flag
always used to store MTU size into struct inpcb::inp_errormtu.
- Clean up these routines: reduce #ifdefs, variable scopes, etc.
|
| 1.122 | 20-May-2014 |
rmind | Adjust PR_WRAP_USRREQS() to include the attach/detach functions.
We still need the kernel-lock for some corner cases.
|
| 1.121 | 19-May-2014 |
rmind | - Split off PRU_ATTACH and PRU_DETACH logic into separate functions.
- Replace malloc with kmem and eliminate M_PCB while here.
- Sprinkle more asserts.
|
| 1.120 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.119 | 18-May-2014 |
rmind | Use IFNET_FIRST() rather than open coding ifnet access.
|
| 1.118 | 25-Feb-2014 |
pooka | branches: 1.118.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.117 | 23-Nov-2013 |
christos | convert from CIRCLEQ to TAILQ.
|
| 1.116 | 05-Jun-2013 |
christos | branches: 1.116.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.115 | 05-Feb-2013 |
joerg | Remove remnants of AF_IMPLINK.
|
| 1.114 | 22-Mar-2012 |
drochner | branches: 1.114.2;
remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.113 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.112 | 17-Jul-2011 |
joerg | branches: 1.112.2; 1.112.6;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.111 | 09-Dec-2009 |
dyoung | Remove superfluous cast of a pointer to void *.
Compare a pointer with NULL, not 0.
No functional change intended.
|
| 1.110 | 16-Sep-2009 |
pooka | Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.109 | 19-Jan-2009 |
christos | Provide compatibility to the old timeval SCM_TIMESTAMP messages.
|
| 1.108 | 06-Aug-2008 |
plunky | branches: 1.108.2;
Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.107 | 24-Apr-2008 |
ad | branches: 1.107.2; 1.107.4; 1.107.8;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.106 | 23-Apr-2008 |
thorpej | Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.105 | 12-Apr-2008 |
thorpej | branches: 1.105.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.104 | 07-Apr-2008 |
thorpej | Change IP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ipstat structure; old netstat
binaries will continue to work properly.
|
| 1.103 | 06-Feb-2008 |
matt | branches: 1.103.6;
Add a new ip_id generation scheme based on a Fisher-Yates shuffle over a
sliding window. XXX replace use of arc4random RSN.
|
| 1.102 | 21-Dec-2007 |
matt | Add fix for ip_id information leakage. Since the leakage information is
primarily used with TCP SYN and RST packets and such packets are less than
the smallest sized packet that an IP stack is allowed to fragment, we simply
set ip_id to 0 for all packets 68 bytes or less.
|
| 1.101 | 27-Nov-2007 |
christos | branches: 1.101.2; 1.101.6;
require that the options argument is the right size, not that it is greater
or equal to the requested size. Suggested by Matt Thomas.
|
| 1.100 | 19-Sep-2007 |
dyoung | branches: 1.100.6;
1) Introduce a new socket option, (SOL_SOCKET, SO_NOHEADER), that
tells a socket that it should both add a protocol header to tx'd
datagrams and remove the header from rx'd datagrams:
int onoff = 1, s = socket(...);
setsockopt(s, SOL_SOCKET, SO_NOHEADER, &onoff);
2) Add an implementation of (SOL_SOCKET, SO_NOHEADER) for raw IPv4
sockets.
3) Reorganize the protocols' pr_ctloutput implementations a bit.
Consistently return ENOPROTOOPT when an option is unsupported,
and EINVAL if a supported option's arguments are incorrect.
Reorganize the flow of code so that it's more clear how/when
options are passed down the stack until they are handled.
Shorten some pr_ctloutput staircases for readability.
4) Extract common mbuf code into subroutines, add new sockaddr
methods, and introduce a new subroutine, fsocreate(), for reuse
later; use it first in sys_socket():
struct mbuf *m_getsombuf(struct socket *so)
Create an mbuf and make its owner the socket `so'.
struct mbuf *m_intopt(struct socket *so, int val)
Create an mbuf, make its owner the socket `so', put the
int `val' into it, and set its length to sizeof(int).
int fsocreate(..., int *fd)
Create a socket, a la socreate(9), put the socket into the
given LWP's descriptor table, return the descriptor at `fd'
on success.
void *sockaddr_addr(struct sockaddr *sa, socklen_t *slenp)
const void *sockaddr_const_addr(const struct sockaddr *sa, socklen_t *slenp)
Extract a pointer to the address part of a sockaddr. Write
the length of the address part at `slenp', if `slenp' is
not NULL.
socklen_t sockaddr_getlen(const struct sockaddr *sa)
Return the length of a sockaddr. This just evaluates to
sa->sa_len. I only add this for consistency with code that
appears in a portable userland library that I am going to
import.
const struct sockaddr *sockaddr_any(const struct sockaddr *sa)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
const void *sockaddr_anyaddr(const struct sockaddr *sa, socklen_t *slenp)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
|
| 1.99 | 02-Sep-2007 |
dyoung | m_copym(..., 0, M_COPYALL, ...) -> m_copypacket(..., ...).
|
| 1.98 | 02-Sep-2007 |
dyoung | m_copy() was deprecated, apparently, long ago. m_copy(...) ->
m_copym(..., M_DONTWAIT).
|
| 1.97 | 12-May-2007 |
dyoung | branches: 1.97.2; 1.97.6; 1.97.8;
KNF. Use sockaddr_in_init(). Shorten staircases. No functional
changes intended.
|
| 1.96 | 04-Mar-2007 |
christos | branches: 1.96.2; 1.96.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.95 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.94 | 25-Oct-2006 |
elad | branches: 1.94.4;
Introduce KAUTH_REQ_NETWORK_SOCKET_OPEN, to check if opening a socket is
allowed. It takes three int * arguments indicating domain, type, and
protocol. Replace previous KAUTH_REQ_NETWORK_SOCKET_RAWSOCK with it (but
keep it still).
Places that used to explicitly check for privileged context now don't
need it anymore, so I replaced these with XXX comment indiacting it for
future reference.
Documented and updated examples as well.
|
| 1.93 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.92 | 19-Sep-2006 |
elad | Remove ugly (void *) casts from network scope authorization wrapper and
calls to it.
While here, adapt code for system scope listeners to avoid some more
casts (forgotten in previous run).
Update documentation.
|
| 1.91 | 08-Sep-2006 |
elad | branches: 1.91.2;
First take at security model abstraction.
- Add a few scopes to the kernel: system, network, and machdep.
- Add a few more actions/sub-actions (requests), and start using them as
opposed to the KAUTH_GENERIC_ISSUSER place-holders.
- Introduce a basic set of listeners that implement our "traditional"
security model, called "bsd44". This is the default (and only) model we
have at the moment.
- Update all relevant documentation.
- Add some code and docs to help folks who want to actually use this stuff:
* There's a sample overlay model, sitting on-top of "bsd44", for
fast experimenting with tweaking just a subset of an existing model.
This is pretty cool because it's *really* straightforward to do stuff
you had to use ugly hacks for until now...
* And of course, documentation describing how to do the above for quick
reference, including code samples.
All of these changes were tested for regressions using a Python-based
testsuite that will be (I hope) available soon via pkgsrc. Information
about the tests, and how to write new ones, can be found on:
http://kauth.linbsd.org/kauthwiki
NOTE FOR DEVELOPERS: *PLEASE* don't add any code that does any of the
following:
- Uses a KAUTH_GENERIC_ISSUSER kauth(9) request,
- Checks 'securelevel' directly,
- Checks a uid/gid directly.
(or if you feel you have to, contact me first)
This is still work in progress; It's far from being done, but now it'll
be a lot easier.
Relevant mailing list threads:
http://mail-index.netbsd.org/tech-security/2006/01/25/0011.html
http://mail-index.netbsd.org/tech-security/2006/03/24/0001.html
http://mail-index.netbsd.org/tech-security/2006/04/18/0000.html
http://mail-index.netbsd.org/tech-security/2006/05/15/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/01/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/25/0000.html
Many thanks to YAMAMOTO Takashi, Matt Thomas, and Christos Zoulas for help
stablizing kauth(9).
Full credit for the regression tests, making sure these changes didn't break
anything, goes to Matt Fleming and Jaime Fournier.
Happy birthday Randi! :)
|
| 1.90 | 23-Jul-2006 |
ad | branches: 1.90.4;
Use the LWP cached credentials where sane.
|
| 1.89 | 14-May-2006 |
elad | integrate kauth.
|
| 1.88 | 11-Dec-2005 |
christos | branches: 1.88.4; 1.88.6; 1.88.8; 1.88.10; 1.88.12;
merge ktrace-lwp.
|
| 1.87 | 29-Apr-2005 |
yamt | branches: 1.87.2;
move decl of inetsw to its own header to avoid array of incomplete type.
found by gcc4. reported by Adam Ciarcinski.
|
| 1.86 | 11-Mar-2005 |
atatat | Revert the change that made kern.file2 and net.*.*.pcblist into nodes
instead of structs. It had other deleterious side-effects that are
rather nasty. Another solution must be found.
|
| 1.85 | 10-Mar-2005 |
atatat | Change types of kern.file2 and net.*.*.pcblist to NODE
|
| 1.84 | 09-Mar-2005 |
atatat | Add the following nodes to the sysctl tree:
net.local.stream.pcblist
net.local.dgram.pcblist
net.inet.tcp.pcblist
net.inet.udp.pcblist
net.inet.raw.pcblist
net.inet6.tcp6.pcblist
net.inet6.udp6.pcblist
net.inet6.raw6.pcblist
which allow retrieval of the pcbs in use for those protocols. The
struct involved is 32/64 bit clean and incorporates parts of struct
inpcb, struct unpcb, a bit of struct tcpcb, and two socket addresses.
|
| 1.83 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.82 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.81 | 04-Sep-2004 |
manu | branches: 1.81.4; 1.81.6;
IPv4 PIM support, based on a submission from Pavlin Radoslavov posted on
tech-net@
|
| 1.80 | 07-May-2004 |
jonathan | Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.79 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.78 | 19-Nov-2003 |
jonathan | branches: 1.78.2;
Patch back support for (badly) randomized IP ids, by request:
* Include "opt_inet.h" everywhere IP-ids are generated with ip_newid(),
so the RANDOM_IP_ID option is visible. Also in ip_id(), to ensure
the prototype for ip_randomid() is made visible.
* Add new sysctl to enable randomized IP-ids, provided the kernel was
configured with RANDOM_IP_ID. (The sysctl defaults to zero, and is
a read-only zero if RANDOM_IP_ID is not configured).
Note that the implementation of randomized IP ids is still defective,
and should not be enabled at all (even if configured) without
very careful deliberation. Caveat emptor.
|
| 1.77 | 17-Nov-2003 |
jonathan | Revert the (default) ip_id algorithm to the pre-randomid algorithm,
due to demonstrated low-period repeated IDs from the randomized IP_id
code. Consensus is that the low-period repetition (much less than
2^15) is not suitable for general-purpose use.
Allocators of new IPv4 IDs should now call the function ip_newid().
Randomized IP_ids is now a config-time option, "options RANDOM_IP_ID".
ip_newid() can use ip_random-id()_IP_ID if and only if configured
with RANDOM_IP_ID. A sysctl knob should be provided.
This API may be reworked in the near future to support linear ip_id
counters per (src,dst) IP-address pair.
|
| 1.76 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.75 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.74 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.73 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.72 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.71 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.70 | 29-Jun-2003 |
fvdl | branches: 1.70.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.69 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.68 | 27-May-2003 |
itojun | can't use M_WAIT here, i believe.
|
| 1.67 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.66 | 30-Jan-2003 |
thorpej | M_SOOPTS -> MT_SOOPTS
|
| 1.65 | 07-Nov-2002 |
thorpej | In the IP_HDRINCL case of rip_output(), if the mbuf is read-only
then copy the header into a new mbuf before modifying it.
Fixes PR 18809. Thanks to Chuq Silvers for diagnosing it.
|
| 1.64 | 22-Oct-2002 |
simonb | Oops, still need the call to va_arg() to advance the args pointer.
|
| 1.63 | 22-Oct-2002 |
simonb | "off" in rip_input() is set but not used, remove it.
static global "ripsrc" is never used, remove it.
|
| 1.62 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.61 | 09-Jun-2002 |
itojun | whitespace
|
| 1.60 | 21-Dec-2001 |
itojun | branches: 1.60.8;
have rip_ctlinput to notify routing changes to raw sockets
(protosw change to be done). sync with kame
|
| 1.59 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.58 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.57 | 25-Jul-2001 |
itojun | branches: 1.57.4;
allocate ipsec policy buffer attached to pcb in in*_pcballoc, before
giving anyone accesses to pcb (do not reveal an inconsistent ones).
sync with kame
|
| 1.56 | 03-Jul-2001 |
itojun | branches: 1.56.2;
call in{,6}_pcbpurgeif0() before in{,6}_purgeif().
|
| 1.55 | 26-Feb-2001 |
itojun | branches: 1.55.2;
make sure to validate packet against ipsec policy.
|
| 1.54 | 24-Jan-2001 |
itojun | - record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.53 | 30-Mar-2000 |
augustss | branches: 1.53.4;
Remove register declarations.
|
| 1.52 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.51 | 17-Feb-2000 |
darrenr | Change the use of pfil hooks. There is no longer a single list of all
pfil information, instead, struct protosw now contains a structure
which caontains list heads, etc. The per-protosw pfil struct is passed
to pfil_hook_get(), along with an in/out flag to get the head of the
relevant filter list. This has been done for only IPv4 and IPv6, at
present, with these patches only enabling filtering for IPPROTO_IP and
IPPROTO_IPV6, although it is possible to have tcp/udp, etc, dedicated
filters now also. The ipfilter code has been updated to only filter
IPv4 packets - next major release of ipfilter is required for ipv6.
|
| 1.50 | 02-Feb-2000 |
thorpej | PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.49 | 01-Feb-2000 |
thorpej | First-draft if_detach() implementation, originally from Bill Studnemund,
although this version has been changed somewhat:
- reference counting on ifaddrs isn't as complete as Bill's original
work was. This is hard to get right, and we should attack one
protocol at a time.
- This doesn't do reference counting or dynamic allocation of ifnets yet.
- This version introduces a new PRU -- PRU_PURGEADDR, which is used to
purge an ifaddr from a protocol. The old method Bill used didn't work
on all protocols, and it only worked on some because it was Very Lucky.
This mostly works ... i.e. works for my USB Ethernet, except for a dangling
ifaddr reference left by the IPv6 code; have not yet tracked this down.
|
| 1.48 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.47 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.46 | 13-Sep-1999 |
itojun | branches: 1.46.2; 1.46.8;
- Call in{,6}_pcbdetach if ipsec initialization is failed during PRU_ATTACH.
This situation happens on severe memory shortage. We may need more
improvements here and there.
- Grab IEEE802 address from IFT_ETHER card, even if the card is
inserted after bootup time. Is there any other card that can be
inserted afterwards? pcmcia fddi card? :-P
- RFC2373 u bit handling suggests that we SHOULD NOT copy interface id from
ethernet card to pseudo interface, when ethernet card has IEEE802/EUI64
with u bit != 0 (this means that IEEE802/EUI64 is not universally unique).
Do not use such address as, for example, interface id for gif interface.
(I have such an ethernet card myself)
This may change interface id for your gif interface. be careful upgrading
rc files.
(sync with recent KAME)
|
| 1.45 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.44 | 05-Jul-1999 |
darrenr | Call icmp_error() at the bottom of rip_input IFF rip_input is the handler
for the protocol in the specified packet.
Fix statistic gathering to not make bogus increments of ips_delivered and
ips_noproto for cases where rip_input() is called by a protocol handler
(such as icmp_input or igmp_input) which has already processed the packet.
|
| 1.43 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.42 | 30-Jan-1999 |
thorpej | branches: 1.42.4; 1.42.6;
Make programs that use raw IP work again; trim the header length from ip_len
before handing the packet off to the socket.
|
| 1.41 | 03-Apr-1998 |
thorpej | branches: 1.41.6;
Fix a bug which would cause a panic in soreceive() if multiple raw
receivers ask for ancillary data.
Noted by Francis Dupont <Francis.Dupont@inria.fr> on tech-net.
|
| 1.40 | 12-Jan-1998 |
scottr | Use option header file for MROUTING
|
| 1.39 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.38 | 16-Nov-1997 |
mycroft | On output, if the packet length doesn't match the length in the IP header,
drop the packet with EINVAL.
|
| 1.37 | 14-Oct-1997 |
matt | branches: 1.37.2;
Add support for returning maximum supported MTU when ip_output fails with
EMSGSIZE.
|
| 1.36 | 11-Jan-1997 |
thorpej | branches: 1.36.10;
Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.35 | 25-Oct-1996 |
thorpej | In rip_output(), sanity check the length of the packet to be transmitted.
If it's larger than IP_MAXPACKET, return an error condition.
Based on a patch from Bill Fenner <fenner@parc.xerox.com>
|
| 1.34 | 16-Sep-1996 |
mycroft | Make sure the sin_zero fields are filled.
|
| 1.33 | 15-Sep-1996 |
mycroft | Hash unconnected PCBs.
|
| 1.32 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.31 | 23-Jun-1996 |
mycroft | Return ENOPROTOOPT rather than picking pseudo-random error values.
Don't allow SIOCGET{VIF,SG}CNT from sockets other than the multicast router.
Restructure rip_ctloutput() like ip_ctloutput(), and fix memory leaks.
|
| 1.30 | 28-May-1996 |
pk | Prototype new rip_*() functions.
|
| 1.29 | 24-May-1996 |
mycroft | Move some code into a separate rip_bind() function.
|
| 1.28 | 23-May-1996 |
mycroft | Make sure the control mbufs are freed in all cases.
|
| 1.27 | 23-May-1996 |
mycroft | Minor changes to make this more like other protocols. Also, fix some return
values.
|
| 1.26 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.25 | 18-Feb-1996 |
christos | branches: 1.25.4;
Fix PR/2095 options MROUTING did not compile.
|
| 1.24 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.23 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.22 | 30-Nov-1995 |
pk | Handle PRU_CONTROL (David Maltz; PR#1664).
|
| 1.21 | 18-Jun-1995 |
cgd | branches: 1.21.2;
convert pcb lists to CIRCLEQs, so that the end can be looked at more
easily, and so that the original (insque/remque) logic can be effectively
mimiced. (This fixes a bug in the previous set of list changes.)
also (since terminator is no longer null) reinstate uninitted list checks,
but mark them XXX.
|
| 1.20 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.19 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.18 | 31-May-1995 |
mycroft | Integrate multicast 3.5 distribution, with several bugs fixed and general
cleanup. This is a (working) snapshot of work in progress.
|
| 1.17 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.16 | 02-Mar-1995 |
glass | Fix for two bad tests in the raw IP socket input code. Only affected
raw sockets that were bound to a local address and/or connected to a
foreign address. Fix from Dan McDonald <danmcd@itd.nrl.navy.mil>
|
| 1.15 | 12-Jan-1995 |
mycroft | Fix mbuf leak in rip_ctloutput().
|
| 1.14 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.13 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.12 | 10-Feb-1994 |
mycroft | Format police.
|
| 1.11 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.10 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.9 | 09-Jan-1994 |
mycroft | Prototype the rest.
|
| 1.8 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.7 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.6 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.5 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.4 | 21-May-1993 |
cgd | add packet size check for raw IP provided by Paul Antonov <apg@apg.kiae.su>,
to fix the "traceroute foohost 2000 == panic" problem.
|
| 1.3 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.2 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.21.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.25.4.2 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.25.4.1 | 10-Nov-1996 |
thorpej | Update from trunk:
- Make ip_len and ip_off unsigned.
- Make sure we don't accept or transmit packets larger than the
maximim IP packet size.
This fixes the so-called `death ping' bug.
Sum of work from Bill Fenner <fenner@parc.xerox.com>,
Kevin Lahey <kml@nas.nasa.gov>, and myself.
Thanks to Curt Sampson, Jukka Marin, and Kevin Lahey for testing
this under NetBSD 1.2
|
| 1.36.10.1 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.37.2.1 | 18-Nov-1997 |
mellon | Pull rev 1.38 up from trunk (mycroft)
|
| 1.41.6.1 | 11-Dec-1998 |
kenh | The beginnings of interface detach support. Still some bugs, but mostly
works for me.
This work was originally by Bill Studenmund, and cleaned up by me.
|
| 1.42.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.42.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.42.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.42.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.42.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.46.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.46.2.3 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.46.2.2 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.46.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.53.4.2 | 06-Apr-2001 |
he | Pull up revision 1.54 (requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.53.4.1 | 26-Feb-2001 |
he | Pull up revision 1.55 (requested by itojun):
Make sure to validate packet against ipsec policy.
|
| 1.55.2.6 | 07-Nov-2002 |
thorpej | Sync with HEAD.
|
| 1.55.2.5 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.55.2.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.55.2.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.55.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.55.2.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.56.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.56.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.56.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.56.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.57.4.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.60.8.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.60.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.70.2.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.70.2.6 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.70.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.70.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.70.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.70.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.70.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.78.2.1 | 10-May-2004 |
tron | Pull up revision 1.80 (requested by jonathan in ticket #280):
Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.81.6.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.81.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.81.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.87.2.8 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.87.2.7 | 21-Jan-2008 |
yamt | sync with head
|
| 1.87.2.6 | 07-Dec-2007 |
yamt | sync with head
|
| 1.87.2.5 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.87.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.87.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.87.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.87.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.88.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.88.10.3 | 06-May-2006 |
christos | - Move kauth_cred_t declaration to <sys/types.h>
- Cleanup struct ucred; forward declarations that are unused.
- Don't include <sys/kauth.h> in any header, but include it in the c files
that need it.
Approved by core.
|
| 1.88.10.2 | 10-Mar-2006 |
elad | generic_authorize() -> kauth_authorize_generic().
|
| 1.88.10.1 | 08-Mar-2006 |
elad | Adapt to kernel authorization KPI.
|
| 1.88.8.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.88.8.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.88.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.88.6.1 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.88.4.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.88.4.1 | 05-Feb-2006 |
rpaulo | inpcb_hdr is gone.
|
| 1.90.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.91.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.91.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.94.4.3 | 17-May-2007 |
yamt | sync with head.
|
| 1.94.4.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.94.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.96.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.96.2.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.96.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.97.8.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.97.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.97.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.97.6.3 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.97.6.2 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.97.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.97.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.100.6.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.100.6.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.100.6.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.101.6.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.101.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.103.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.103.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.105.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.107.8.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.107.4.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.107.2.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.107.2.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.108.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.112.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.112.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.112.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.112.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.114.2.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.114.2.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.114.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.114.2.1 | 25-Feb-2013 |
tls | resync with head
|
| 1.116.2.4 | 18-May-2014 |
rmind | sync with head
|
| 1.116.2.3 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.116.2.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.116.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.118.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.145.2.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.146.2.8 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.146.2.7 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.146.2.6 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.146.2.5 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.146.2.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.146.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.146.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.146.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.158.2.4 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.158.2.3 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.158.2.2 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.158.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.161.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.164.4.2 | 09-Apr-2018 |
bouyer | Pull up following revision(s) (requested by roy in ticket #724):
tests/net/icmp/t_ping.c: revision 1.19
sys/netinet6/raw_ip6.c: revision 1.166
sys/netinet6/ip6_input.c: revision 1.195
sys/net/raw_usrreq.c: revision 1.59
sys/sys/socketvar.h: revision 1.151
sys/kern/uipc_socket2.c: revision 1.128
tests/lib/libc/sys/t_recvmmsg.c: revision 1.2
lib/libc/sys/recv.2: revision 1.38
sys/net/rtsock.c: revision 1.239
sys/netinet/udp_usrreq.c: revision 1.246
sys/netinet6/icmp6.c: revision 1.224
tests/net/icmp/t_ping.c: revision 1.20
sys/netipsec/keysock.c: revision 1.63
sys/netinet/raw_ip.c: revision 1.172
sys/kern/uipc_socket.c: revision 1.260
tests/net/icmp/t_ping.c: revision 1.22
sys/kern/uipc_socket.c: revision 1.261
tests/net/icmp/t_ping.c: revision 1.23
sys/netinet/ip_mroute.c: revision 1.155
sbin/route/route.c: revision 1.159
sys/netinet6/ip6_mroute.c: revision 1.123
sys/netatalk/ddp_input.c: revision 1.31
sys/netcan/can.c: revision 1.3
sys/kern/uipc_usrreq.c: revision 1.184
sys/netinet6/udp6_usrreq.c: revision 1.138
tests/net/icmp/t_ping.c: revision 1.18
socket: report receive buffer overflows
Add soroverflow() which increments the overflow counter, sets so_error
to ENOBUFS and wakes the receive socket up.
Replace all code that manually increments this counter with soroverflow().
Add soroverflow() to raw_input().
This allows userland to detect route(4) overflows so it can re-sync
with the current state.
socket: clear error even when peeking
The error has already been reported and it's pointless requiring another
recv(2) call just to clear it.
socket: remove now incorrect comment that so_error is only udp
As it can be affected by route(4) sockets which are raw.
rtsock: log dropped messages that we cannot report to userland
Handle ENOBUFS when receiving messages.
Don't send messages if the receiver has died.
Sprinkle more soroverflow().
Handle ENOBUFS in recv
Handle ENOBUFS in sendto
Note value received. Harden another sendto for ENOBUFS.
Handle the routing socket overflowing gracefully.
Allow a valid sendto .... duh
Handle errors better.
Fix test for checking we sent all the data we asked to.
|
| 1.164.4.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.171.2.5 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.171.2.4 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.171.2.3 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.171.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.171.2.1 | 22-Mar-2018 |
pgoyette | Synch with HEAD, resolve conflicts
|
| 1.177.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.185.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.9 | 25-Jun-2012 |
christos | rename rfc6056 -> portalgo, requested by yamt
|
| 1.8 | 21-Jun-2012 |
yamt | for the default "bsd" algorithm, restore the pre rfc6056 changes behaviour.
fix anonportmin/max.
probably other algorithms need similar fixes.
|
| 1.7 | 21-Jun-2012 |
yamt | whitespace and cosmetics. no functional changes.
|
| 1.6 | 13-Apr-2012 |
yamt | comment
|
| 1.5 | 15-Mar-2012 |
gson | Fix random kernel memory corruption by algo_doublehash(). And by
"random" I don't mean just "arbitary" as in using an uninitialized
pointer, but random as in corrupting the contents of memory addresses
chosen using a crypto-strength random number generator.
I believe this is the likely cause of multiple reports of random
crashes over the last six months, including kern/45677 and kern/46096.
|
| 1.4 | 19-Nov-2011 |
tls | branches: 1.4.2; 1.4.4;
First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>. This change includes
the following:
An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are
fixed. Some effort is made to accumulate entropy more quickly at
boot time.
A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.
The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.
An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c). This
is based on Greg Rose's implementation from Qualcomm.
A new random stream generator, nist_ctr_drbg, is provided. It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim. This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.
An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness. The arc4random/arc4randbytes API is deprecated for
in-kernel use. It is replaced by "cprng_strong". The current
cprng_fast implementation wraps the existing arc4random
implementation. The current cprng_strong implementation wraps the
new CTR_DRBG implementation. Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.
In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data(). Performance is expected to improve.
The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.
The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot. There is
approximately a 3/10000 chance of a false positive from these
tests. Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.
A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32). A new kernel will require a new rndctl.
The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.
Manual pages for the new kernel interfaces are forthcoming.
|
| 1.3 | 25-Sep-2011 |
mrg | branches: 1.3.2;
make this build without INET6.
also, fix the rfc6056algo passed to sysctl_rfc6056_helper
(it was backwards for inet4/inet6.)
|
| 1.2 | 24-Sep-2011 |
christos | disable debugging
|
| 1.1 | 24-Sep-2011 |
christos | Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.3.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.3.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.4.4.1 | 17-Mar-2012 |
bouyer | Pull up following revision(s) (requested by gson in ticket #122):
sys/netinet/rfc6056.c: revision 1.5
Fix random kernel memory corruption by algo_doublehash(). And by
"random" I don't mean just "arbitary" as in using an uninitialized
pointer, but random as in corrupting the contents of memory addresses
chosen using a crypto-strength random number generator.
I believe this is the likely cause of multiple reports of random
crashes over the last six months, including kern/45677 and kern/46096.
|
| 1.4.2.2 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.4.2.1 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.4 | 25-Jun-2012 |
christos | rename rfc6056 -> portalgo, requested by yamt
|
| 1.3 | 22-Jun-2012 |
christos | PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.2 | 24-Sep-2011 |
christos | branches: 1.2.2;
install the header.
|
| 1.1 | 24-Sep-2011 |
christos | Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.2.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.7 | 28-Feb-2025 |
andvar | Fix various typos in comments.
|
| 1.6 | 17-Mar-2024 |
andvar | branches: 1.6.2;
Add missing "e" in few words, in comments and one log message.
|
| 1.5 | 24-Oct-2021 |
andvar | fix various typos in comments, mainly copypasta.
|
| 1.4 | 06-Sep-2021 |
andvar | fix various typos in comments.
|
| 1.3 | 03-Jun-2019 |
msaitoh | Fix typo in comment (s/seperate/separate/).
|
| 1.2 | 27-Jun-2017 |
rjs | branches: 1.2.4; 1.2.8;
Pack structs.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.8.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.2.4.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.4.1 | 27-Jun-2017 |
jdolecek | file sctp.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.6.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.14 | 11-Apr-2024 |
knakahara | Fix invalid IPv6 route when ipsecif(4) is deleted tunnel. Pointed out by ohishi@IIJ.
The pointed bug is fixed by modification in nd6_need_cache().
Others are similar bugs.
XXX pullup-9, 10
|
| 1.13 | 09-Feb-2024 |
andvar | fix spelling mistakes, mainly in comments and log messages.
|
| 1.12 | 25-Jun-2019 |
rjs | branches: 1.12.28;
Split out the prototypes for add/delete address into a separate header file.
|
| 1.11 | 28-Jun-2017 |
rjs | branches: 1.11.4; 1.11.8;
Put back some commented out code.
|
| 1.10 | 17-Jan-2017 |
ozaki-r | Fix build w/ SCTP and w/o SCTP_DEBUG
|
| 1.9 | 16-Jan-2017 |
christos | ip6_sprintf -> IN6_PRINT so that we pass the size.
|
| 1.8 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.7 | 15-Dec-2016 |
ozaki-r | branches: 1.7.2;
Restore nd6.h inclusion to resolve implicit dependency
|
| 1.6 | 13-Dec-2016 |
ozaki-r | Remove unnecessary inclusions of nd6.h
|
| 1.5 | 07-Jul-2016 |
ozaki-r | branches: 1.5.2;
Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.4 | 07-Jul-2016 |
ozaki-r | Use IFADDR_FOREACH instead of IFADDR_FOREACH_SAFE
No item is removed in the loop.
|
| 1.3 | 12-May-2016 |
ozaki-r | Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.2 | 03-Apr-2016 |
mlelstv | Replace generic queue macros with IFNET/IFADDR macros.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_asconf.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.5.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.5.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.7.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.11.8.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.11.4.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.11.4.1 | 28-Jun-2017 |
jdolecek | file sctp_asconf.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.12.28.1 | 18-Apr-2024 |
martin | Pull up following revision(s) (requested by knakahara in ticket #659):
sys/netinet6/in6_ifattach.c: revision 1.122
sys/netinet/sctp_asconf.c: revision 1.14
sys/netinet6/nd6.c: revision 1.282
Fix invalid IPv6 route when ipsecif(4) is deleted tunnel. Pointed out by ohishi@IIJ.
The pointed bug is fixed by modification in nd6_need_cache().
Others are similar bugs.
|
| 1.4 | 25-Jun-2019 |
rjs | Split out the prototypes for add/delete address into a separate header file.
|
| 1.3 | 08-Jun-2019 |
rjs | Don't need 'extern' for function prototypes.
|
| 1.2 | 24-Feb-2019 |
kamil | Appease GCC7 in sctp_asconf.h
Do not declare types inside function parameter list.
Add decklarations of types before these function prototypes.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18; 1.1.22;
Add core networking support for SCTP.
|
| 1.1.22.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.1.22.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_asconf.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_asconf.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.5 | 02-Feb-2024 |
andvar | fix various typos in comments.
|
| 1.4 | 18-May-2022 |
andvar | s/yeild/yield/
|
| 1.3 | 01-Jan-2022 |
msaitoh | s/implemenation/implementation/ in comment.
|
| 1.2 | 27-Dec-2019 |
msaitoh | s/inital/initial/
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18; 1.1.22;
Add core networking support for SCTP.
|
| 1.1.22.1 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_constants.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_constants.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.3 | 13-Aug-2019 |
rjs | Remove unused checksum code.
|
| 1.2 | 12-Aug-2016 |
jdolecek | branches: 1.2.14; 1.2.18;
sprinkle const on sctp_crc_c[]
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_crc32.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.18.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.2.14.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.14.1 | 12-Aug-2016 |
jdolecek | file sctp_crc32.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.2 | 13-Aug-2019 |
rjs | Remove unused checksum code.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18; 1.1.22;
Add core networking support for SCTP.
|
| 1.1.22.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_crc32.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_crc32.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_hashdriver.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_hashdriver.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_hashdriver.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_hashdriver.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.4 | 08-Dec-2023 |
andvar | fix triple s typos in comments.
|
| 1.3 | 05-Dec-2021 |
msaitoh | s/convience/convenience/ in comment.
|
| 1.2 | 27-Jun-2017 |
rjs | branches: 1.2.4;
Pack structs.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_header.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.4.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.4.1 | 27-Jun-2017 |
jdolecek | file sctp_header.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.18 | 17-Apr-2025 |
andvar | s/valdid/valid/ and s/valdiation/validation/ in comments.
|
| 1.17 | 04-Dec-2024 |
andvar | s/transmite/transmitte/ in comments.
|
| 1.16 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.15 | 09-Feb-2024 |
andvar | branches: 1.15.2;
s/anthing/anything/ and s/be to/too/ in comments.
|
| 1.14 | 15-Jan-2024 |
andvar | Fix few typos in comments, mainly s/argment/argument/.
|
| 1.13 | 05-Apr-2023 |
andvar | remove some double ee typos in comments.
|
| 1.12 | 28-May-2022 |
andvar | fix various typos in comments.
|
| 1.11 | 24-May-2022 |
andvar | fix various typos in comment, documentation and log messages.
|
| 1.10 | 08-Apr-2022 |
andvar | s/postion/position/
|
| 1.9 | 07-Apr-2022 |
andvar | fix various typos in comments.
|
| 1.8 | 22-Dec-2018 |
maxv | Replace: M_MOVE_PKTHDR -> m_move_pkthdr. No functional change, since the
former is a macro to the latter.
|
| 1.7 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.6 | 01-May-2018 |
maxv | branches: 1.6.2;
Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.5 | 10-Dec-2017 |
rjs | branches: 1.5.2;
Improve compliance to RFC 6458.
|
| 1.4 | 25-Apr-2016 |
rjs | branches: 1.4.16;
Fix build when IPSEC enabled.
|
| 1.3 | 15-Feb-2016 |
rtr | Fix building of IPv4-Mapped IPv6 addresses.
As discussed on tech-net@ use in6_sin_2_v4mapsin6() to build mapped
addresses.
|
| 1.2 | 13-Dec-2015 |
christos | branches: 1.2.2;
PR/50528: David Binderman: remove sizeof(sizeof(x))
|
| 1.1 | 13-Oct-2015 |
rjs | Add core networking support for SCTP.
|
| 1.2.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.2.2.3 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.2.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.2.2.1 | 13-Dec-2015 |
skrll | file sctp_indata.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.4.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.4.16.1 | 25-Apr-2016 |
jdolecek | file sctp_indata.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.5.2.3 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.5.2.2 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.5.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.6.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.15.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_indata.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_indata.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.19 | 11-Jul-2025 |
andvar | Fix various typos, mainly in comments and log/error messages.
|
| 1.18 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.17 | 02-Feb-2024 |
andvar | branches: 1.17.2;
fix various typos in comments.
|
| 1.16 | 08-Apr-2022 |
andvar | s/postion/position/
|
| 1.15 | 19-Sep-2021 |
andvar | fix various typos in comments, messages and documentation.
|
| 1.14 | 28-May-2019 |
msaitoh | s/recieve/receive/
|
| 1.13 | 24-Feb-2019 |
kamil | Add missing FALLTHROUGH in sctp_input.c
Requested by GCC NetBSD/i386 kUBSan KCOC build.
|
| 1.12 | 12-Feb-2019 |
rjs | Add some fallthrough annotations.
|
| 1.11 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.10 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.9 | 01-May-2018 |
maxv | branches: 1.9.2;
Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.8 | 26-Feb-2018 |
maxv | branches: 1.8.2;
Dedup: merge ipsec4_in_reject and ipsec6_in_reject into ipsec_in_reject.
While here fix misleading comment.
ok ozaki-r@
|
| 1.7 | 27-Jun-2017 |
rjs | branches: 1.7.4;
Check outgoing cookie size before accessing any contents.
Spotted in FreeBSD by maya.
|
| 1.6 | 23-Jun-2017 |
rjs | Make arguments match debug message.
|
| 1.5 | 20-Apr-2017 |
ozaki-r | Fix build of kernel with SCTP
|
| 1.4 | 20-Apr-2017 |
ozaki-r | Remove unnecessary NULL checks for inp_socket and in6p_socket
They cannot be NULL except for programming errors.
|
| 1.3 | 10-Jun-2016 |
ozaki-r | branches: 1.3.2; 1.3.4;
Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.2 | 25-Apr-2016 |
rjs | Fix build when IPSEC enabled.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.5 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.4 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_input.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.3.4.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.3.2.1 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.7.4.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.7.4.1 | 27-Jun-2017 |
jdolecek | file sctp_input.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.8.2.3 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.8.2.2 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.8.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.9.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.17.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_input.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_input.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.39 | 12-Jun-2025 |
ozaki-r | sctp: follow the recent change of ip_newid()
|
| 1.38 | 14-Apr-2025 |
andvar | fix few typos in comments and update link to Printer Device Class spec.
|
| 1.37 | 08-Sep-2024 |
rillig | fix a/an grammar in obvious cases
|
| 1.36 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.35 | 09-Feb-2024 |
andvar | branches: 1.35.2;
fix spelling mistakes, mainly in comments and log messages.
|
| 1.34 | 13-Sep-2023 |
bouyer | handle EHOSTDOWN the same way as EHOSTUNREACH in sctp_med_chunk_output().
Compile-tested only (I don't have a sctp setup); proposed in
http://mail-index.netbsd.org/tech-net/2023/09/11/msg008611.html
LGTM from Greg Troxel and Robert Swindells
|
| 1.33 | 04-Nov-2022 |
ozaki-r | branches: 1.33.2;
inpcb: rename functions to in6pcb_*
|
| 1.32 | 28-Oct-2022 |
ozaki-r | Adjust dccp and sctp for struct inpcb separation
|
| 1.31 | 31-May-2022 |
andvar | fix various typos in comments, documentation and messages.
|
| 1.30 | 28-May-2022 |
andvar | fix various typos, mainly in comments.
|
| 1.29 | 08-Apr-2022 |
andvar | fix various typos, mainly in comments, but also log messages, docs, game text.
|
| 1.28 | 05-Dec-2021 |
msaitoh | s/futher/further/ in comment.
|
| 1.27 | 05-Dec-2021 |
msaitoh | s/measurment/measurement/ in comment.
|
| 1.26 | 21-Oct-2021 |
andvar | fix various typos, mainly in comments, but also in man pages and log messages.
|
| 1.25 | 07-Sep-2021 |
andvar | s/aquire/acquire/ in comments, also one typo fix acqure->acquire.
|
| 1.24 | 03-Sep-2021 |
andvar | fix typos in comments, mainly s/extention/extension/ and s/sufficent/sufficient/
|
| 1.23 | 24-Jul-2021 |
andvar | Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.
|
| 1.22 | 13-Jun-2020 |
roy | branches: 1.22.6;
SCTP: Use ifp->if_mtu rather than ND_IFINFO(ifp)->linkmtu
|
| 1.21 | 26-Dec-2019 |
msaitoh | Fix typo in comment.
|
| 1.20 | 03-Dec-2019 |
msaitoh | s/upate/update/ in comment.
|
| 1.19 | 13-Nov-2019 |
ozaki-r | Get rid of unnecessary NULL checks for rt_ifa and ifa_ifp
They are always non-NULL nowadays.
|
| 1.18 | 22-Dec-2018 |
maxv | Replace M_ALIGN and MH_ALIGN by m_align.
|
| 1.17 | 16-Sep-2018 |
skrll | interrupt has two 'r's
fix another typo while I'm here (flsah)
|
| 1.16 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.15 | 03-May-2018 |
maxv | branches: 1.15.2;
Remove m_copy completely.
|
| 1.14 | 01-May-2018 |
maxv | Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.13 | 30-Mar-2018 |
maya | correct typo: and and -> and (comments only)
heads up on this being a common typo from chris28.
|
| 1.12 | 10-Dec-2017 |
rjs | branches: 1.12.2;
Improve compliance to RFC 6458.
|
| 1.11 | 27-Jun-2017 |
rjs | branches: 1.11.4;
Use host byte order for a debug message.
|
| 1.10 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.9 | 23-Dec-2016 |
maya | branches: 1.9.2;
Remove extraneous parentheses. no functional change
Appeases clang
|
| 1.8 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.7 | 07-Jul-2016 |
ozaki-r | branches: 1.7.2;
Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.6 | 10-Jun-2016 |
ozaki-r | Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.5 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.4 | 12-May-2016 |
ozaki-r | Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.3 | 25-Apr-2016 |
rjs | Fix build when IPSEC enabled.
|
| 1.2 | 03-Apr-2016 |
mlelstv | Replace generic queue macros with IFNET/IFADDR macros.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_output.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.7.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.7.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.9.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.11.4.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.11.4.1 | 27-Jun-2017 |
jdolecek | file sctp_output.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.12.2.6 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.12.2.5 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.12.2.4 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.12.2.3 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.12.2.2 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.12.2.1 | 07-Apr-2018 |
pgoyette | Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
|
| 1.15.2.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.15.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.15.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.22.6.1 | 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.33.2.2 | 29-Jul-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1140):
sys/netinet/ip_output.c: revision 1.330
sys/netinet/sctp_output.c: revision 1.39
sys/netinet/ip_mroute.c: revision 1.166
sys/netipsec/ipsecif.c: revision 1.24
sys/netipsec/xform_ipip.c: revision 1.80
sys/netinet/ip_output.c: revision 1.327
sys/netinet/ip_output.c: revision 1.328
sys/netinet/ip_input.c: revision 1.406
sys/netinet/ip_output.c: revision 1.329
sys/netinet/in_var.h: revision 1.105
in: get rid of unused argument from ip_newid() and ip_newid_range()
in: take a reference of ifp on IP_ROUTETOIF
The ifp could be released after ia4_release(ia).
in: narrow the scope of ifa in ip_output (NFC)
sctp: follow the recent change of ip_newid()
in: avoid racy ifa_acquire(rt->rt_ifa) in ip_output()
If a rtentry is being destroyed asynchronously, ifa referenced by rt_ifa
can be destructed and taking ifa_acquire(rt->rt_ifa) aborts with a
KASSERT failure. Fortunately, the ifa is not actually freed because of
a reference by rt_ifa, it can be available (except some functions like
psref) so as long the rtentry is held.
PR kern/59527
in: avoid racy ia4_acquire(ifatoia(rt->rt_ifa) in ip_rtaddr()
Same as the case of ip_output(), it's racy and should be avoided.
PR kern/59527
|
| 1.33.2.1 | 13-Sep-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #863):
sys/netinet/sctp_output.c: revision 1.34
handle EHOSTDOWN the same way as EHOSTUNREACH in sctp_med_chunk_output().
Compile-tested only (I don't have a sctp setup); proposed in
http://mail-index.netbsd.org/tech-net/2023/09/11/msg008611.html
LGTM from Greg Troxel and Robert Swindells
|
| 1.35.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_output.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_output.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.27 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.26 | 15-Oct-2022 |
andvar | branches: 1.26.8;
fix various typos in documentation and comments.
mainly in words functionality, functional, function.
|
| 1.25 | 28-May-2022 |
andvar | fix various typos in comments.
|
| 1.24 | 24-May-2022 |
andvar | fix various typos in comment, documentation and log messages.
|
| 1.23 | 10-Dec-2021 |
andvar | s/occured/occurred/ in comments, log messages and man pages.
|
| 1.22 | 19-Sep-2021 |
andvar | fix various typos in comments, messages and documentation.
|
| 1.21 | 30-Apr-2020 |
riastradh | Omit needless #include <sys/rnd.h>.
|
| 1.20 | 19-Jan-2020 |
riastradh | Replace kooky sctp random number generation by cprng_strong32().
|
| 1.19 | 26-Dec-2019 |
msaitoh | branches: 1.19.2;
Fix typo in comment.
|
| 1.18 | 11-Dec-2018 |
christos | PR/53775: Havard Eidnes: bind(2) may inaccurately return EADDRNOTAVAIL,
it should return EADDRINUSE.
|
| 1.17 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.16 | 27-Feb-2018 |
maxv | branches: 1.16.2; 1.16.4;
Dedup: merge
ipsec4_get_policy and ipsec6_get_policy
ipsec4_delete_pcbpolicy and ipsec6_delete_pcbpolicy
The already-existing ipsec_get_policy() function is inlined in the new
one.
|
| 1.15 | 17-Oct-2017 |
rjs | branches: 1.15.2;
Make SCTP work when IPSEC is also defined.
|
| 1.14 | 17-Oct-2017 |
rjs | Move call to sofree() to end of sctp_inpcb_free() and re-aquire
softnet_lock.
Logic copied from in_pcb.c.
|
| 1.13 | 17-Oct-2017 |
rjs | Remove duplicate assignment, comment doesn't match it anyway.
|
| 1.12 | 17-Oct-2017 |
rjs | Remove some foreign conditional code. NFC intended.
|
| 1.11 | 17-Oct-2017 |
rjs | Wrap pcb list check with #ifdef DEBUG.
|
| 1.10 | 17-Oct-2017 |
rjs | Remove function prototype that is no longer required. NFC
|
| 1.9 | 28-Jun-2017 |
rjs | Whitespace.
|
| 1.8 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.7 | 07-Jul-2016 |
ozaki-r | branches: 1.7.2;
Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.6 | 22-May-2016 |
rjs | Remove rtcache reference to route before freeing the containing struct.
|
| 1.5 | 12-May-2016 |
ozaki-r | Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.4 | 25-Apr-2016 |
rjs | Fix build when IPSEC enabled.
|
| 1.3 | 14-Apr-2016 |
rjs | Remove stray debug printf().
|
| 1.2 | 03-Apr-2016 |
mlelstv | Replace generic queue macros with IFNET/IFADDR macros.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_pcb.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.7.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.15.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.15.2.1 | 17-Oct-2017 |
jdolecek | file sctp_pcb.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.16.4.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.16.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.16.2.2 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.16.2.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.19.2.1 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.26.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.8 | 02-Jun-2023 |
andvar | follow the steps of Andrew Doran (ad) commit and fix more s/loose/lose/ typos.
also s/beyound/beyond/ and few others along the way, mainly in comments.
|
| 1.7 | 28-Oct-2022 |
ozaki-r | Adjust dccp and sctp for struct inpcb separation
|
| 1.6 | 28-Oct-2022 |
ozaki-r | Adjust pf, wg, dccp and sctp for struct inpcb integration
|
| 1.5 | 24-May-2022 |
andvar | fix various typos in comment, documentation and log messages.
|
| 1.4 | 09-Aug-2021 |
andvar | fix various typos in compatibility, mainly in comments.
|
| 1.3 | 19-Jan-2020 |
riastradh | Replace kooky sctp random number generation by cprng_strong32().
|
| 1.2 | 08-Jun-2019 |
rjs | branches: 1.2.4;
Don't need 'extern' for function prototypes.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18; 1.1.22;
Add core networking support for SCTP.
|
| 1.1.22.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.1.22.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_pcb.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_pcb.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.4.1 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.2 | 25-Apr-2016 |
rjs | branches: 1.2.16;
Fix build when IPSEC enabled.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_peeloff.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.16.1 | 25-Apr-2016 |
jdolecek | file sctp_peeloff.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_peeloff.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_peeloff.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.1 | 25-Jun-2019 |
rjs | branches: 1.1.10;
Split out the prototypes for add/delete address into a separate header file.
|
| 1.1.10.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.1.10.1 | 25-Jun-2019 |
martin | file sctp_route.h was added on branch phil-wifi on 2020-04-13 08:05:16 +0000
|
| 1.4 | 10-Aug-2023 |
andvar | fix typos in comments s/iton/tion/ or s/ton/tion/.
|
| 1.3 | 24-Jun-2023 |
msaitoh | Fix typo in comment.
|
| 1.2 | 25-Apr-2016 |
rjs | branches: 1.2.16;
Fix build when IPSEC enabled.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_structs.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.16.1 | 25-Apr-2016 |
jdolecek | file sctp_structs.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.6 | 16-Feb-2022 |
andvar | fix various typos, mainly in comments.
|
| 1.5 | 01-May-2018 |
maxv | Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.4 | 10-Dec-2017 |
rjs | branches: 1.4.2;
Add ipsec option header.
|
| 1.3 | 08-Dec-2016 |
ozaki-r | branches: 1.3.14;
Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.2 | 25-Apr-2016 |
rjs | branches: 1.2.2;
Fix build when IPSEC enabled.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.4 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_timer.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.3.14.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.3.14.1 | 08-Dec-2016 |
jdolecek | file sctp_timer.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.4.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18;
Add core networking support for SCTP.
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_timer.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_timer.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.4 | 31-Jul-2018 |
rjs | Change implementation of sctp_connectx() to use ioctl(2).
|
| 1.3 | 10-Dec-2017 |
rjs | branches: 1.3.2; 1.3.4;
Improve compliance to RFC 6458.
|
| 1.2 | 28-Jun-2017 |
rjs | branches: 1.2.4;
Pack assoc structs.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_uio.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.4.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.4.1 | 28-Jun-2017 |
jdolecek | file sctp_uio.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.3.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.3.2.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.27 | 08-Sep-2024 |
rillig | s/effect/affect/ in a few places
|
| 1.26 | 08-Sep-2024 |
rillig | fix a/an grammar in obvious cases
|
| 1.25 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.24 | 09-Feb-2024 |
andvar | branches: 1.24.2;
fix spelling mistakes, mainly in comments and log messages.
|
| 1.23 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.22 | 06-Aug-2022 |
andvar | s/blity/bility/ in various words, mainly in comments.
|
| 1.21 | 29-May-2022 |
andvar | fix various typos in comments and log messages.
|
| 1.20 | 27-Apr-2020 |
rjs | Do sctp_connectx() handling using ioctl() for IPv6 as well.
|
| 1.19 | 25-Jun-2019 |
rjs | Split out the prototypes for add/delete address into a separate header file.
|
| 1.18 | 25-Feb-2019 |
maxv | RIP6, CAN, SCTP and SCTP6 lack a length check in their _send() functions.
Fix RIP6 and CAN, add a big XXX in the SCTP ones.
Found by KASAN, triggered by SyzKaller.
Reported-by: syzbot+0b9692ae0f49f93b7dc7@syzkaller.appspotmail.com
|
| 1.17 | 24-Feb-2019 |
maxv | RIP, RIP6, DDP, SCTP and SCTP6 lack a length check in their _connect()
functions. Fix the first three, and add a big XXX in the SCTP ones.
Found by KASAN, triggered by SyzKaller.
Reported-by: syzbot+9eaf98dad6ca738c250d@syzkaller.appspotmail.com
|
| 1.16 | 15-Feb-2019 |
rjs | This really was a missing break.
Spotted by Rin Okuyama.
|
| 1.15 | 12-Feb-2019 |
rjs | Add some fallthrough annotations.
|
| 1.14 | 28-Jan-2019 |
martin | Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.13 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.12 | 31-Jul-2018 |
rjs | Enable SCTP sysctl nodes.
Rename auto asconf one to match FreeBSD.
|
| 1.11 | 31-Jul-2018 |
rjs | Change implementation of sctp_connectx() to use ioctl(2).
|
| 1.10 | 01-May-2018 |
maxv | branches: 1.10.2;
Remove now unused net_osdep.h includes, the other BSDs did the same.
|
| 1.9 | 10-Dec-2017 |
rjs | branches: 1.9.2;
Improve compliance to RFC 6458.
|
| 1.8 | 17-Oct-2017 |
rjs | branches: 1.8.2;
Make SCTP work when IPSEC is also defined.
|
| 1.7 | 17-Oct-2017 |
rjs | Set SPL level to match usage for TCP.
|
| 1.6 | 07-Jul-2016 |
ozaki-r | branches: 1.6.10;
Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.5 | 12-May-2016 |
ozaki-r | Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.4 | 25-Apr-2016 |
rjs | Fix build when IPSEC enabled.
|
| 1.3 | 03-Apr-2016 |
mlelstv | Replace generic queue macros with IFNET/IFADDR macros.
|
| 1.2 | 13-Dec-2015 |
christos | branches: 1.2.2;
PR/50529: David Binderman: Remove double sizeof
|
| 1.1 | 13-Oct-2015 |
rjs | Add core networking support for SCTP.
|
| 1.2.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.2.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.2.2.3 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.2.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.2.2.1 | 13-Dec-2015 |
skrll | file sctp_usrreq.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.6.10.1 | 29-Jan-2019 |
msaitoh | Pull up following revision(s) (requested by martin in ticket #1175):
sys/net/link_proto.c 1.37
sys/netatalk/ddp_usrreq.c 1.72
sys/netbt/hci_socket.c 1.46
sys/netbt/l2cap_socket.c 1.36
sys/netbt/rfcomm_socket.c 1.38
sys/netbt/sco_socket.c 1.38
sys/netinet/sctp_usrreq.c 1.14
sys/netinet/tcp_usrreq.c 1.223
sys/netinet6/raw_ip6.c 1.173
sys/netinet6/sctp6_usrreq.c 1.17
sys/netinet6/udp6_usrreq.c 1.146
sys/netmpls/mpls_proto.c 1.32
sys/netnatm/natm.c patch
Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.8.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.8.2.1 | 17-Oct-2017 |
jdolecek | file sctp_usrreq.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.9.2.2 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.9.2.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.10.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.10.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.24.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.4 | 27-Apr-2020 |
rjs | Do sctp_connectx() handling using ioctl() for IPv6 as well.
|
| 1.3 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.2 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2; 1.1.18; 1.1.20; 1.1.22;
Add core networking support for SCTP.
|
| 1.1.22.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.1.20.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.1.20.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 13-Oct-2015 |
jdolecek | file sctp_var.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctp_var.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.21 | 11-Jul-2025 |
andvar | Fix various typos, mainly in comments and log/error messages.
|
| 1.20 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.19 | 08-Apr-2022 |
andvar | branches: 1.19.10;
s/postion/position/
|
| 1.18 | 05-Dec-2021 |
msaitoh | s/measurment/measurement/ in comment.
|
| 1.17 | 24-Jul-2021 |
andvar | Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.
|
| 1.16 | 19-Jan-2020 |
riastradh | branches: 1.16.10;
Replace kooky sctp random number generation by cprng_strong32().
|
| 1.15 | 13-Aug-2019 |
rjs | branches: 1.15.2;
Remove unused checksum code.
|
| 1.14 | 08-Nov-2018 |
msaitoh | "s/ are are / are /" in comment. No functional change.
|
| 1.13 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.12 | 16-Jan-2017 |
christos | branches: 1.12.12; 1.12.14; 1.12.16;
ip6_sprintf -> IN6_PRINT so that we pass the size.
|
| 1.11 | 16-Jan-2017 |
ryo | Make ip6_sprintf(), in_fmtaddr(), lla_snprintf() and icmp6_redirect_diag() mpsafe.
Reviewed by ozaki-r@
|
| 1.10 | 07-Jul-2016 |
ozaki-r | branches: 1.10.2; 1.10.4;
Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
|
| 1.9 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.8 | 22-May-2016 |
rjs | Use const for arguments to sctp_is_same_scope().
|
| 1.7 | 12-May-2016 |
ozaki-r | Protect ifnet list with psz and psref
The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).
Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.
|
| 1.6 | 26-Apr-2016 |
rjs | Fix build when IPSEC enabled.
|
| 1.5 | 11-Apr-2016 |
ozaki-r | Sweep unncessary radix.h inclusions
|
| 1.4 | 03-Apr-2016 |
mlelstv | Replace generic queue macros with IFNET/IFADDR macros.
|
| 1.3 | 06-Mar-2016 |
christos | PR/50899: David Binderman: optimize memset
|
| 1.2 | 15-Feb-2016 |
rtr | Fix building of IPv4-Mapped IPv6 addresses.
As discussed on tech-net@ use in6_sin_2_v4mapsin6() to build mapped
addresses.
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.7 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.6 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.1.2.5 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.4 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.1.2.3 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctputil.c was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.10.4.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.10.2.1 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.12.16.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.12.16.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.12.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.12.14.2 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.12.14.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.12.12.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.12.12.1 | 16-Jan-2017 |
jdolecek | file sctputil.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.15.2.1 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.16.10.1 | 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.19.10.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.5 | 05-Jul-2024 |
rin | sctp_m_freem: Safely accept NULL argument as m_freem(9)
|
| 1.4 | 14-Apr-2024 |
andvar | branches: 1.4.2;
s/force_comile_error/force_compile_error/
|
| 1.3 | 19-Jan-2020 |
riastradh | Replace kooky sctp random number generation by cprng_strong32().
|
| 1.2 | 22-May-2016 |
rjs | branches: 1.2.16; 1.2.20; 1.2.26;
Use const for arguments to sctp_is_same_scope().
|
| 1.1 | 13-Oct-2015 |
rjs | branches: 1.1.2;
Add core networking support for SCTP.
|
| 1.1.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 27-Dec-2015 |
skrll | Sync with HEAD (as of 26th Dec)
|
| 1.1.2.1 | 13-Oct-2015 |
skrll | file sctputil.h was added on branch nick-nhusb on 2015-12-27 12:10:07 +0000
|
| 1.2.26.1 | 25-Jan-2020 |
ad | Sync with head.
|
| 1.2.20.1 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.2.16.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.2.16.1 | 22-May-2016 |
jdolecek | file sctputil.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.4.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.37 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.36 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.35 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.34 | 01-Nov-2019 |
christos | branches: 1.34.8;
Add comments to the tcp flags.
|
| 1.33 | 10-Jan-2017 |
christos | branches: 1.33.16;
add a couple of lint comments.
|
| 1.32 | 02-Jan-2017 |
christos | Fix TCP signature code:
1. pack options more tightly instead of being generous with no/op
2. put TCP_SIGNATURE option before SACK
3. fix computation of options length, by deferring it
XXX: Really we should move the options setting code in one place instead
of having two copies one for input and one for output.
XXX: tcp_optlen/tcp_hdrsiz need to be fixed; they were wrong before too.
|
| 1.31 | 14-Feb-2015 |
he | branches: 1.31.2;
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
|
| 1.30 | 07-Jan-2012 |
christos | branches: 1.30.6; 1.30.22; 1.30.24;
make standalone
|
| 1.29 | 11-Dec-2011 |
christos | u_int -> uint
|
| 1.28 | 25-Dec-2007 |
perry | branches: 1.28.44; 1.28.48;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.27 | 02-Aug-2007 |
rmind | branches: 1.27.4; 1.27.10; 1.27.12; 1.27.16; 1.27.20;
TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.26 | 20-Jun-2007 |
christos | branches: 1.26.2;
- per socket keepalive settings
- settable connection establishment timeout
|
| 1.25 | 09-Oct-2006 |
rpaulo | branches: 1.25.8; 1.25.10;
Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.24 | 25-Sep-2006 |
rpaulo | Remove line that shouldn't have been committed.
|
| 1.23 | 25-Sep-2006 |
rpaulo | PR/14806: NetBSD uses the wrong default TCP MSS.
No objections in tech-net.
|
| 1.22 | 05-Sep-2006 |
rpaulo | branches: 1.22.2; 1.22.4;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.21 | 10-Dec-2005 |
elad | branches: 1.21.4; 1.21.8;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.20 | 21-Jul-2005 |
riz | Add a definition for TCPOLEN_SIGLEN from OpenBSD, so a kernel with
options TCP_SIGNATURE will compile again after the new PMTU checks
were brought in from OpenBSD. Approved by christos.
|
| 1.19 | 07-Mar-2005 |
yamt | branches: 1.19.4;
tcp_sack_option: the max number of sack blocks in a packet is 4, not 3.
|
| 1.18 | 07-Dec-2004 |
yamt | branches: 1.18.4; 1.18.6;
remove TCPOPT_MD5SIGNATURE because no one in our tree uses it
and it's duplicated with TCPOPT_SIGNATURE.
i preferred TCPOPT_SIGNATURE because it's used by FreeBSD and OpenBSD.
|
| 1.17 | 07-May-2004 |
kleink | Add definitions for the (currently unimplemented) ECN TCP flags;
from Chuck Swiger in PR standards/25058.
|
| 1.16 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.15 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.14 | 05-May-2003 |
bjh21 | branches: 1.14.2;
Header cleanup: Hide all of this file apart from the socket options
from POSIX/XNS applications.
|
| 1.13 | 26-May-2001 |
matt | Add TCP_MD5SIGNATURE option.
|
| 1.12 | 05-Jul-2000 |
christos | branches: 1.12.2;
added a linted comment about non-portable bitfields. Unfortunately it cannot
be fixed portably.
|
| 1.11 | 20-Nov-1999 |
thorpej | Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.10 | 04-Oct-1998 |
matt | branches: 1.10.12; 1.10.18;
Adapt the NEWRENO changes from the UCSB diffs of BSDI 3.0's TCP
to NetBSD. Ignore the SACK & FACK stuff for now.
|
| 1.9 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.8 | 17-Apr-1995 |
cgd | spacing cleaup. also, minor type mixup fixups.
|
| 1.7 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.6 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.5 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.4 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.10.18.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.10.12.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.12.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.14.2.7 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.14.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.14.2.5 | 08-Mar-2005 |
skrll | Sync with HEAD.
|
| 1.14.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.14.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.14.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.14.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.18.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.18.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.19.4.4 | 21-Jan-2008 |
yamt | sync with head
|
| 1.19.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.19.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.19.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.21.8.1 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.21.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.22.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.22.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.25.10.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.25.8.2 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.25.8.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.26.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.27.20.2 | 02-Aug-2007 |
rmind | TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.27.20.1 | 02-Aug-2007 |
rmind | file tcp.h was added on branch matt-mips64 on 2007-08-02 02:42:41 +0000
|
| 1.27.16.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.27.12.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.27.10.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.27.4.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.28.48.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.28.44.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.30.24.2 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.30.24.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.30.22.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by he in ticket #530):
sys/netinet/tcp_output.c: revision 1.180
sys/netinet/tcp_input.c: revision 1.336
sys/netinet/tcp_usrreq.c: revision 1.203
share/man/man4/tcp.4: revision 1.30
sys/netinet/tcp.h: revision 1.31
sys/netinet/tcp_subr.c: revision 1.258
sys/netinet/tcp_var.h: revision 1.176
sys/netinet/tcp_var.h: revision 1.177
sys/sys/param.h: bump revision
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.30.6.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.31.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.31.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.33.16.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.34.8.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.29 | 14-May-2024 |
andvar | fix recently committed typos by msaitoh in few more places, as well as few more.
mainly s/contigous/contiguous/ and s/miliseconds/milliseconds/ in comments.
|
| 1.28 | 31-Jul-2021 |
andvar | s/threshhold/threshold
|
| 1.27 | 09-Oct-2019 |
msaitoh | branches: 1.27.12;
All of snd_wnd, snd_cwnd and snd_ssthresh in stuct tcpcb are u_long,
so use u_long and ulmin() instead of u_int and uimin(). Found by lgtm bot.
XXX TCP's sequence number is uint32_t, so it might be good to change some
entries in struct tcpcb to uint32_t instead of u_long. FreeBSD did it.
|
| 1.26 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.25 | 03-May-2018 |
maxv | branches: 1.25.2;
Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.24 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.23 | 02-Jan-2017 |
skrll | branches: 1.23.14;
Restore behaviour to pre- tcp_congctl.c:1.18 for SACK. Further analysis
of the change is required.
OK kefren@
PR/51753 tcp SACK causes SSH disconnect
|
| 1.22 | 13-Dec-2016 |
ozaki-r | Remove unnecessary inclusions of nd6.h
|
| 1.21 | 26-Apr-2016 |
ozaki-r | branches: 1.21.2;
Sweep unnecessary route.h inclusions
|
| 1.20 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.19 | 18-Nov-2013 |
kefren | branches: 1.19.4; 1.19.6; 1.19.8; 1.19.10;
Cubic changes:
* correct W(t) calculation
* check wmax limits
* change W_max in slow and fast retransmit
* correct rtt approximation
Reno:
* move comment I forgot behind after fast_retransmit() split
|
| 1.18 | 12-Nov-2013 |
kefren | * implement TCP CUBIC congestion control algorithm
* move tcp_sack_newack bits inside reno and newreno_fast_retransmit_newack
* notify ECN peer about cwnd shrink in [new]reno_slow_retransmit
Based on the patch proposed on tech-net@ on Nov 7 with minor improvments:
* adapt wmax for no-fast convergence case
* correct cbrt calculation for big window sizes (>750KB)
|
| 1.17 | 25-Oct-2013 |
martin | Mark a diagnostic-only variable
|
| 1.16 | 08-Apr-2011 |
yamt | branches: 1.16.4; 1.16.14; 1.16.18;
simplify code a little. no functional changes.
|
| 1.15 | 28-Apr-2008 |
martin | branches: 1.15.22; 1.15.28;
Remove clause 3 and 4 from TNF licenses
|
| 1.14 | 29-Feb-2008 |
matt | branches: 1.14.2; 1.14.4;
Rework tcp congctl selection code so that the congctl entries can be const.
Don't access tcp_congctl stuff outside of tcp_congctl.c, use routines to
update t_congctl. This code is slightly now more complicated.
|
| 1.13 | 11-Jul-2007 |
xtraeme | branches: 1.13.8; 1.13.24; 1.13.28;
Replace a simple lock with a mutex and make it static.
|
| 1.12 | 16-Nov-2006 |
christos | branches: 1.12.2; 1.12.6; 1.12.12;
__unused removal on arguments; approved by core.
|
| 1.11 | 21-Oct-2006 |
yamt | branches: 1.11.2;
constify.
|
| 1.10 | 19-Oct-2006 |
yamt | tcp_reno_newack: remove an __unused because it's now used.
|
| 1.9 | 19-Oct-2006 |
yamt | tcp_reno_newack: regardless of sysctl setting, use L=1*SMSS when
we are doing retransmission.
|
| 1.8 | 19-Oct-2006 |
yamt | implement RFC3465 appropriate byte counting.
from Kentaro A. Kurahone, with minor adjustments by me.
the ack prediction part of the original patch was omitted because
it's a separate change. reviewed by Rui Paulo.
|
| 1.7 | 15-Oct-2006 |
rpaulo | Move comments to proper places.
|
| 1.6 | 15-Oct-2006 |
rpaulo | Add a new tcp_congctl(9) structure member for congestion experienced callback.
Needed by HSTCP.
|
| 1.5 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.4 | 10-Oct-2006 |
rpaulo | tcp_reno_newack(): bring the exact original code.
tcp_newreno_newack(): call tcp_reno_newack() if partialacks < 0.
|
| 1.3 | 10-Oct-2006 |
yamt | tcp_reno_newack/tcp_newreno_newack: remove stale comments.
|
| 1.2 | 10-Oct-2006 |
yamt | tcp_newreno_newack: actually inflate cwnd as it used to do.
|
| 1.1 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.11.2.3 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.11.2.2 | 22-Oct-2006 |
yamt | sync with head
|
| 1.11.2.1 | 21-Oct-2006 |
yamt | file tcp_congctl.c was added on branch yamt-splraiseipl on 2006-10-22 06:07:28 +0000
|
| 1.12.12.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.12.6.4 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.12.6.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.12.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.12.6.1 | 16-Nov-2006 |
yamt | file tcp_congctl.c was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.12.2.2 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.12.2.1 | 16-Nov-2006 |
ad | file tcp_congctl.c was added on branch newlock2 on 2006-11-18 21:39:36 +0000
|
| 1.13.28.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.13.28.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.13.24.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.13.8.1 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.14.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.14.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.15.28.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.15.22.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.16.18.1 | 18-May-2014 |
rmind | sync with head
|
| 1.16.14.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.16.14.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.16.4.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.19.10.1 | 18-Jan-2017 |
skrll | Sync with netbsd-5
|
| 1.19.8.1 | 05-Jan-2017 |
martin | Pull up following revision(s) (requested by skrll in ticket #1347):
sys/netinet/tcp_congctl.c: revision 1.23
Restore behaviour to pre- tcp_congctl.c:1.18 for SACK. Further analysis
of the change is required.
OK kefren@
PR/51753 tcp SACK causes SSH disconnect
|
| 1.19.6.3 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.19.6.2 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.19.6.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.19.4.1 | 05-Jan-2017 |
martin | Pull up following revision(s) (requested by skrll in ticket #1347):
sys/netinet/tcp_congctl.c: revision 1.23
Restore behaviour to pre- tcp_congctl.c:1.18 for SACK. Further analysis
of the change is required.
OK kefren@
PR/51753 tcp SACK causes SSH disconnect
|
| 1.21.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.23.14.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.23.14.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.23.14.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.25.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.25.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.27.12.1 | 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.7 | 12-Nov-2013 |
kefren | * implement TCP CUBIC congestion control algorithm
* move tcp_sack_newack bits inside reno and newreno_fast_retransmit_newack
* notify ECN peer about cwnd shrink in [new]reno_slow_retransmit
Based on the patch proposed on tech-net@ on Nov 7 with minor improvments:
* adapt wmax for no-fast convergence case
* correct cbrt calculation for big window sizes (>750KB)
|
| 1.6 | 14-Apr-2011 |
yamt | branches: 1.6.4; 1.6.14; 1.6.18;
- comments
- g/c stale extern
|
| 1.5 | 28-Apr-2008 |
martin | branches: 1.5.22; 1.5.28;
Remove clause 3 and 4 from TNF licenses
|
| 1.4 | 29-Feb-2008 |
matt | branches: 1.4.2; 1.4.4;
Rework tcp congctl selection code so that the congctl entries can be const.
Don't access tcp_congctl stuff outside of tcp_congctl.c, use routines to
update t_congctl. This code is slightly now more complicated.
|
| 1.3 | 21-Oct-2006 |
yamt | branches: 1.3.2; 1.3.4; 1.3.8; 1.3.30; 1.3.50; 1.3.54;
constify.
|
| 1.2 | 15-Oct-2006 |
rpaulo | Add a new tcp_congctl(9) structure member for congestion experienced callback.
Needed by HSTCP.
|
| 1.1 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.3.54.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.3.54.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.3.50.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.3.30.1 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.3.8.3 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.3.8.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.3.8.1 | 21-Oct-2006 |
yamt | file tcp_congctl.h was added on branch yamt-lazymbuf on 2006-12-30 20:50:33 +0000
|
| 1.3.4.2 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.3.4.1 | 21-Oct-2006 |
ad | file tcp_congctl.h was added on branch newlock2 on 2006-11-18 21:39:36 +0000
|
| 1.3.2.2 | 22-Oct-2006 |
yamt | sync with head
|
| 1.3.2.1 | 21-Oct-2006 |
yamt | file tcp_congctl.h was added on branch yamt-splraiseipl on 2006-10-22 06:07:28 +0000
|
| 1.4.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.4.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.5.28.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.5.22.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.6.18.1 | 18-May-2014 |
rmind | sync with head
|
| 1.6.14.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.6.4.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.32 | 03-May-2018 |
maxv | Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.31 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.30 | 26-Apr-2016 |
ozaki-r | branches: 1.30.16;
Sweep unnecessary route.h inclusions
|
| 1.29 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.28 | 18-Apr-2009 |
tsutsui | branches: 1.28.22; 1.28.40;
Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.27 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.26 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.25 | 04-Mar-2007 |
christos | branches: 1.25.40; 1.25.50; 1.25.56;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.24 | 11-Dec-2005 |
christos | branches: 1.24.26;
merge ktrace-lwp.
|
| 1.23 | 06-Sep-2005 |
rpaulo | Wrap two big lines.
|
| 1.22 | 02-Jun-2005 |
riz | branches: 1.22.2;
Fix some const fallout.
|
| 1.21 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.20 | 13-Jan-2005 |
drochner | branches: 1.20.2; 1.20.4;
compile tcp_debug.c only if the TCP_DEBUG option is set,
and remove the "#ifdef TCP_DEBUG" around everything
|
| 1.19 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.18 | 09-Jun-2002 |
itojun | branches: 1.18.6;
whitespace
|
| 1.17 | 13-Nov-2001 |
lukem | branches: 1.17.8;
add RCSIDs
|
| 1.16 | 08-Jul-2001 |
abs | branches: 1.16.2;
Rename TCPDEBUG to TCP_DEBUG, defopt TCP_DEBUG and TCP_NDEBUG, and
make all usage of tcp_trace dependent on TCP_DEBUG - resulting in
a 31K saving on an INET enabled i386 kernel.
|
| 1.15 | 08-Jul-2001 |
abs | Give TCPDEBUG a chance of working - fix printf() types, add missing &s,
and remove attempt to use a non existant tcphdr field.
|
| 1.14 | 01-Jul-1999 |
itojun | branches: 1.14.14;
IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.13 | 13-Oct-1996 |
christos | branches: 1.13.24; 1.13.26;
backout previous kprintf change
|
| 1.12 | 13-Oct-1996 |
christos | backout previous kprintf changes
|
| 1.11 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.10 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.9 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.8 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.7 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.6 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.3 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.13.26.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.13.26.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.13.26.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.13.24.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.14.14.3 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.14.14.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.14.14.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.16.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.16.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.17.8.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.18.6.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.18.6.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.18.6.4 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.18.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.18.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.18.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.20.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.20.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.22.2.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.22.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.24.26.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.25.56.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.25.50.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.25.40.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.28.40.2 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.28.40.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.28.22.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.30.16.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.30.16.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.21 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.20 | 03-Feb-2021 |
roy | tcp_debug: restore __packed
|
| 1.19 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.18 | 01-May-2018 |
maxv | branches: 1.18.14;
Redefine the structure, not to rely on tcpiphdr.
|
| 1.17 | 04-Mar-2007 |
christos | branches: 1.17.128;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.16 | 10-Dec-2005 |
elad | branches: 1.16.26;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.15 | 02-Jun-2005 |
riz | branches: 1.15.2;
Fix some const fallout.
|
| 1.14 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.13 | 09-Jul-2001 |
itojun | branches: 1.13.22;
do not #ifdef in headers. usr.sbin/trpt needs it.
|
| 1.12 | 08-Jul-2001 |
abs | Rename TCPDEBUG to TCP_DEBUG, defopt TCP_DEBUG and TCP_NDEBUG, and
make all usage of tcp_trace dependent on TCP_DEBUG - resulting in
a 31K saving on an INET enabled i386 kernel.
|
| 1.11 | 30-May-2001 |
mrg | use _KERNEL_OPT
|
| 1.10 | 29-Apr-2001 |
fvdl | Make it possible to override TCP_NDEBUG. The default value of 100
wastes quite a bit of space (0xfa00).
|
| 1.9 | 31-Jul-1999 |
itojun | branches: 1.9.14;
sync with recent KAME.
- loosen ipsec restriction on packet diredtion.
- revise icmp6 redirect handling on IsRouter bit.
- tcp/udp notification processing (link-local address case)
- cosmetic fixes (better code share across *BSD).
|
| 1.8 | 09-Jul-1999 |
thorpej | defopt INET6, and put it in opt_inet.h (most places already include this
file, which is why the file list is so short).
|
| 1.7 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.6 | 10-Feb-1998 |
perry | branches: 1.6.10; 1.6.12;
add/cleanup multiple inclusion protection.
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.6.12.2 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.6.12.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.6.10.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.6.10.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.9.14.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.9.14.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.13.22.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.13.22.4 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.13.22.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.13.22.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.13.22.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.15.2.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.15.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.16.26.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.17.128.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.18.14.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.16 | 07-Apr-2018 |
maxv | Remove dead code.
|
| 1.15 | 10-Dec-2005 |
elad | branches: 1.15.162;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.14 | 20-Apr-2004 |
matt | branches: 1.14.12;
export tcpstates for _KERNEL and remove tcp_usrreq.c's incorrect
declartion.
|
| 1.13 | 20-Nov-2003 |
yamt | comments on tcp_outflags.
|
| 1.12 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.11 | 20-Oct-2001 |
matt | branches: 1.11.18;
Make tcp_outflags & tcpstates const.
|
| 1.10 | 09-Jul-1998 |
mycroft | branches: 1.10.26; 1.10.28;
Back out the change from TCP/IP vol 2, in revision 1.7, which removed TH_FIN
from the output flags for CLOSING state. There is no harm in retransmitting
the FIN, and this change has unexpected side effects that break simultaneous
close behaviour.
|
| 1.9 | 03-Jul-1998 |
thorpej | Fix TCPS_HAVERCVDFIN() to actually catch all TCP states in which a FIN
has been received (CLOSE_WAIT, CLOSING, LAST_ACK, and TIME_WAIT).
From David Borman <dab@bsdi.com>.
|
| 1.8 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.7 | 18-Jun-1997 |
kleink | branches: 1.7.8;
As per RFC 793, don't retransmit the FIN during a simultaneous close.
From Thorsten Frueauf <frueauf@ira.uka.de> and W. Richard Stevens in PR/3737
and TCP/IP Illustrated, Vol. 2, respectively.
|
| 1.6 | 14-Oct-1994 |
mycroft | Don't return received data to the user until the initial handshake is complete.
Also use TCPS_HAVEESTABLISHED() in a few other places.
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.7.8.1 | 01-Oct-1998 |
cgd | pull up revs 1.9-1.10 from trunk. (mycroft)
|
| 1.10.28.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.10.26.1 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.11.18.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.11.18.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.11.18.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.11.18.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.14.12.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.15.162.1 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.441 | 08-Oct-2024 |
rin | tcp_reass: Mitigate CVE-2018-6922 (SegmentSmack)
at a level of FreeBSD, by introducing an arbitrary (100) limit to
the length of TCP reassembly queues:
https://github.com/freebsd/freebsd-src/commit/95a914f6316874f5b0c45d491f2843dc810071ef
Originally authored by ryo@.
We thank Tomoyuki Sahara <tsahara at iij>, who has analyzed the
problem again, updated the patch, and carried out experiments for
vulnerability scenarios. The confidential PR below is based on
his work.
PR security/58708
|
| 1.440 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.439 | 29-Jun-2024 |
riastradh | branches: 1.439.2;
netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.438 | 04-Nov-2022 |
ozaki-r | branches: 1.438.2;
inpcb: rename functions to in6pcb_*
|
| 1.437 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.436 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.435 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.434 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.433 | 24-May-2022 |
andvar | fix various typos in comment, documentation and log messages.
|
| 1.432 | 23-Mar-2022 |
andvar | fix few typos in comments, mainly s/paramenters/parameters/.
|
| 1.431 | 09-Aug-2021 |
andvar | fix typos in asymmetry, asymmetric(al), symmetrical.
|
| 1.430 | 06-Aug-2021 |
andvar | fix various typos in comments.
|
| 1.429 | 31-Jul-2021 |
andvar | s/threshhold/threshold
|
| 1.428 | 08-Mar-2021 |
christos | branches: 1.428.4;
Remove the unused "addin" argument (it was always 0) and go back using
a random iss by default (instead of rfc1948)
|
| 1.427 | 19-Feb-2021 |
jakllsch | it's spelled struct tcphdr, not struct tcp_hdr
|
| 1.426 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.425 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.424 | 29-Sep-2020 |
msaitoh | branches: 1.424.2;
s/occurence/occurrence/
|
| 1.423 | 13-Sep-2020 |
roy | inet: Fix build without ARP
|
| 1.422 | 11-Sep-2020 |
roy | ARP: Use ND rather than our own.
This brings the benefit of Neighbour Unreachability Detection which is
something ARP sorely lacks.
The new timings mirror those of IPv6 and are adjustable via sysctl(8).
Unlike IPv6 ND, these are global and not per interface.
|
| 1.421 | 11-Sep-2020 |
roy | tcp_input: Adjust for ND changes
|
| 1.420 | 11-Sep-2020 |
kardel | PR/kern 55567
fix the data-only fast path. RCV.UP and SND.WL1 could be left behind
on long sequences of data only packets. pull them along to avoid relative
sequence wraps.
consistent with FreeBSD
addresses second failure mode of PR/kern 55567.
pullup to netbsd-8
pullup to netbsd-9
|
| 1.419 | 02-Sep-2020 |
kardel | Fix fast path for uni directional transfers
pure ACK case:
drag snd_wl2 along so only newer
ACKs can update the window size.
also avoids the state where snd_wl2
is eventually larger than th_ack and thus
blocking the window update mechanism and
the connection gets stuck for a loooong
time in the zero sized send window state.
see PR/kern 55567
ok thorpej@, also found in FreeBSD
|
| 1.418 | 06-Jul-2020 |
christos | - always set both ip and ip6, otherwise a kernel assertion can be triggered
- move alignment early so that we do less work
|
| 1.417 | 16-Nov-2019 |
maxv | Call rtcache_unref() only when the checks succeed, instead of relying on
another NULL check in rtcache_unref().
Because, in order to resolve the address of the second argument, we do a
dereference on 'tp', which is theoretically allowed to be NULL. The five
callers of nd6_hint() never pass a NULL argument however, so by luck the
actual NULL deref never happens.
Maybe the NULL check on 'tp' in should be replaced to a KASSERT ensuring
it isn't NULL, for clarity.
Reported by kUBSan.
|
| 1.416 | 25-Sep-2019 |
jnemeth | PR/54572 - Edgar Fu� -- error in comment
|
| 1.415 | 06-Aug-2019 |
riastradh | Clamp tcp timer quantities to reasonable ranges.
Reported-by: syzbot+259675123340bf46a6de@syzkaller.appspotmail.com
|
| 1.414 | 01-Jun-2019 |
kamil | branches: 1.414.2;
Replace potentially misaligned pointer dereference + htonl() with be32dec()
Reported by kUBSan.
|
| 1.413 | 08-Nov-2018 |
msaitoh | "s/ an an / an /" in comment. No functional change.
|
| 1.412 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.411 | 14-Sep-2018 |
maxv | rename toff -> off
|
| 1.410 | 14-Sep-2018 |
maxv | rename off -> thlen
|
| 1.409 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.408 | 18-May-2018 |
maxv | branches: 1.408.2;
IP6_EXTHDR_GET -> M_REGION_GET, no functional change.
|
| 1.407 | 03-May-2018 |
maxv | Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.406 | 28-Apr-2018 |
maxv | Remove unused ipsec_var.h includes.
|
| 1.405 | 08-Apr-2018 |
maxv | Remove the ipre_mlast field and the TRAVERSE macro.
The goal was to store in ipre_mlast the last mbuf of the chain, so that
m_cat could be called on it. But it's not needed, since m_cat already
does the equivalent of TRAVERSE itself.
If it were needed, there would be a bug, since we don't call TRAVERSE on
ipre_mlast when creating a new reassembly entry.
|
| 1.404 | 03-Apr-2018 |
maxv | Remove ipsec_copy_policy and ipsec_copy_pcbpolicy. No functional change,
since we used only ipsec_copy_pcbpolicy, and it was a no-op.
Originally we were using ipsec_copy_policy to optimize the IPsec-PCB
cache: when an ACK was received in response to a SYN, we used to copy the
SP cached in the SYN's PCB into the ACK's PCB, so that
ipsec_getpolicybysock could use the cached SP instead of requerying it.
Then we switched to ipsec_copy_pcbpolicy which has always been a no-op. As
a result the SP cached in the SYN was/is not copied in the ACK, and the
first call to ipsec_getpolicybysock had to query the SP and cache it
itself. It's not totally clear to me why this change was made.
But it has been this way for years, and after a conversation with Ryota
Ozaki it turns out the optimization is not valid anymore due to
MP-ification, so it won't be re-enabled.
ok ozaki-r@
|
| 1.403 | 30-Mar-2018 |
maxv | Fix the log. mtod never returns NULL, so 'ip' is always non-NULL, and the
'ip6' branch is never taken. As a result we log garbage on IPv6 packets.
Use ip_v instead.
|
| 1.402 | 30-Mar-2018 |
maxv | Use consttime_memequal instead of memcmp, to prevent side channels. This
functions returns 1 when the buffers are equal, contrary to memcmp, hence
the !.
|
| 1.401 | 29-Mar-2018 |
rmind | tcp_urp_drop: fix a bug introduced in 1.390 rev (hi maxv@).
|
| 1.400 | 29-Mar-2018 |
maxv | Remove TCPREASS_DEBUG. It was introduced 20 years ago when the reassembler
was being developed, but it's irrelevant today. Makes the code clearer.
|
| 1.399 | 29-Mar-2018 |
maxv | Reorder/Fix comments to clarify.
|
| 1.398 | 29-Mar-2018 |
maxv | Remove two more 'else' branches.
|
| 1.397 | 29-Mar-2018 |
maxv | Fix memory leak, we may reallocate 'tcp_saveti' after 'findpcb'. It's not
a tragic bug, because it happens only on sockets with debug enabled.
|
| 1.396 | 29-Mar-2018 |
maxv | Remove 'else', makes it clearer that we leave.
|
| 1.395 | 29-Mar-2018 |
maxv | Clarify with KASSERT.
|
| 1.394 | 29-Mar-2018 |
maxv | Simplify the computation:
m->m_pkthdr.len - sizeof(struct tcphdr) - optlen - hlen
= m->m_pkthdr.len - (sizeof(struct tcphdr) + optlen + hlen)
= m->m_pkthdr.len - [tcp_len]
= toff
|
| 1.393 | 28-Mar-2018 |
maxv | Several changes in syn_cache_respond:
* Replace idiotic diagnostic check by KASSERT. max_linkhdr+tlen<=MCLBYTES
is a widespread assumption.
* Improve initialization of 'tp'.
* Put panics in dead branches.
* Merge two switches.
|
| 1.392 | 28-Mar-2018 |
maxv | Remove unused variable.
|
| 1.391 | 28-Mar-2018 |
maxv | Remove two unused args from syn_cache_get().
|
| 1.390 | 28-Mar-2018 |
maxv | Dedup: introduce tcp_urp_drop() and use it.
|
| 1.389 | 28-Mar-2018 |
maxv | Minor changes: style, improve comments (and put them at the correct place),
use NULL for pointers, and add {}s to prevent confusion.
|
| 1.388 | 23-Mar-2018 |
maxv | Remove #ifdef INET. Nobody is doing that in the kernel, and there are
even IPv4 places that are not covered here.
|
| 1.387 | 23-Mar-2018 |
maxv | Improve a bit here and there. Replace bcopy by memcpy/memmove.
|
| 1.386 | 22-Mar-2018 |
maxv | Don't pass a pointer to tcp_reass, otherwise it looks like it can modify
tlen while it doesn't.
|
| 1.385 | 22-Mar-2018 |
maxv | Rearrange a bit. No real functional change.
|
| 1.384 | 22-Mar-2018 |
maxv | Don't call tcp_input_checksum again, it was already called earlier, no
need to checksum twice.
Then call tcp_fields_to_host a bit earlier, so that we don't need to call
it in each branch.
|
| 1.383 | 01-Mar-2018 |
maxv | branches: 1.383.2;
Revert rev1.183 (2003).
It was intended as an optimization, but it increases the attack surface:
the IPsec policy is not enforced on RST packets when the socket is in the
LISTEN state, and an (unauthenticated) attacker could jam the connection
between two IPsec hosts by sending RST packets between the client's SYN
and ACK packets.
Discussed with ozaki-r@.
|
| 1.382 | 28-Feb-2018 |
maxv | Remove unused ipsec_private.h includes.
|
| 1.381 | 28-Feb-2018 |
maxv | Remove duplicate IPSEC_STATINC(IPSEC_STAT_IN_POLVIO), ipsec_in_reject
already increases it. IPSEC6_STATINC is now unused, so remove it too.
|
| 1.380 | 26-Feb-2018 |
maxv | Dedup: merge ipsec4_in_reject and ipsec6_in_reject into ipsec_in_reject.
While here fix misleading comment.
ok ozaki-r@
|
| 1.379 | 12-Feb-2018 |
maxv | Remove unused argument from tcp_signature_getsav.
|
| 1.378 | 12-Feb-2018 |
maxv | Add a KASSERT.
|
| 1.377 | 12-Feb-2018 |
maxv | Remove the 'm' argument from syn_cache_respond(); all it does with it is
freeing it, so free in the caller instead.
|
| 1.376 | 12-Feb-2018 |
maxv | Remove this multicast check. Multicast packets are already dropped at
the beginning of the function.
|
| 1.375 | 09-Feb-2018 |
maxv | Style, and move the 'ip_srcroute' call after 'tcp_dooptions', otherwise
we're leaking 'ipopts'. (Harmless, since TCP_SIGNATURE is disabled.)
|
| 1.374 | 08-Feb-2018 |
maxv | Style, rename a variable, and remove an unreachable case.
|
| 1.373 | 08-Feb-2018 |
maxv | Move the IPv4 multicast check earlier; we want to kick multicast packets
all the time, and not just when they are SYNs.
The IPv6 multicast check is already done earlier, so this block of code
can be removed.
|
| 1.372 | 08-Feb-2018 |
maxv | Remove the unused 'multicast' argument from tcp_vtw_input, and remove
the now-unused multicast detection code. It couldn't have been correct on
IPv6, since multicast packets are kicked at the beginning of the function.
|
| 1.371 | 08-Feb-2018 |
maxv | Remove the default case, the beginning of the function already ensures
af == AF_INET || af == AF_INET6.
|
| 1.370 | 08-Feb-2018 |
maxv | Dedup code.
|
| 1.369 | 08-Feb-2018 |
maxv | Remove the IN6_IS_ADDR_V4MAPPED checks in the protocol functions. They
are useless, because the IPv6 entry point (ip6_input) already performs
them.
The checks were first added in the protocol functions:
Wed Dec 22 04:03:02 1999 UTC (18 years, 1 month ago) by itojun
"drop IPv6 packets with v4 mapped address on src/dst. they are illegal
and may be used to fool IPv6 implementations (by using ::ffff:127.0.0.1 as
source you may be able to pretend the packet is from local node)"
Shortly afterwards they were also added in the IPv6 entry point, but
where not removed from the protocol functions:
Mon Jan 31 10:33:22 2000 UTC (18 years ago) by itojun
"be proactive about malicious packet on the wire. we fear that v4 mapped
address to be used as a tool to hose security filters (like bypassing
"local host only" filter by using ::ffff:127.0.0.1)."
OpenBSD did the same a few months ago. FreeBSD has never had these checks.
|
| 1.368 | 08-Feb-2018 |
maxv | Style, and remove outdated comments.
|
| 1.367 | 08-Feb-2018 |
maxv | Remove this check, it is already done at the beginning of the function.
|
| 1.366 | 08-Feb-2018 |
maxv | Reduce the indentation level of this huge block (without realigning yet,
for proofreadability). No functional change.
|
| 1.365 | 08-Feb-2018 |
maxv | Move the SO_DEBUG block earlier, to reduce the indentation level.
|
| 1.364 | 08-Feb-2018 |
dholland | Typos.
|
| 1.363 | 15-Nov-2017 |
ozaki-r | Convert SYN_CACHE_TIMER_ARM macro to static inline function (NFC)
|
| 1.362 | 15-Nov-2017 |
ozaki-r | Make syn_cache_timer static
|
| 1.361 | 15-Nov-2017 |
ozaki-r | Reduce return points (NFC)
|
| 1.360 | 03-Aug-2017 |
ozaki-r | Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
|
| 1.359 | 07-Jul-2017 |
ozaki-r | Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
|
| 1.358 | 19-Jun-2017 |
ozaki-r | Fix KASSERT in tcp_input
inp can be NULL when receiving an IPv4 packet on an IPv4-mapped IPv6
address. In that case KASSERT(sotoinpcb(so) == inp) always fails.
Should fix PR kern/52304 (at least it fixes the same panic as the
report)
|
| 1.357 | 20-Apr-2017 |
ozaki-r | branches: 1.357.4;
Remove unnecessary NULL checks for inp_socket and in6p_socket
They cannot be NULL except for programming errors.
|
| 1.356 | 31-Mar-2017 |
ozaki-r | Don't use a single global variable to store source route information for multiple incoming packets
It's not MP-safe. So use a m_tag to store the information instead.
Pointed out by knakahara@
The fix is from OpenBSD (originally fixed in FreeBSD)
|
| 1.355 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.354 | 07-Feb-2017 |
ozaki-r | Add missing NULL checks for m_get_rcvif
|
| 1.353 | 04-Jan-2017 |
kre | branches: 1.353.2;
Remove redundant tests: if optlen === 0, then optlen % 4 != 2 (it is 0)
so there is no need to test both.
|
| 1.352 | 02-Jan-2017 |
christos | Fix TCP signature code:
1. pack options more tightly instead of being generous with no/op
2. put TCP_SIGNATURE option before SACK
3. fix computation of options length, by deferring it
XXX: Really we should move the options setting code in one place instead
of having two copies one for input and one for output.
XXX: tcp_optlen/tcp_hdrsiz need to be fixed; they were wrong before too.
|
| 1.351 | 31-Dec-2016 |
christos | remove ancient ipsec code, and don't conditionalize tcp signatures on ipsec_used
|
| 1.350 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.349 | 15-Nov-2016 |
mrg | apply a #ifdef INET6 so the previous compiles without INET6.
|
| 1.348 | 15-Nov-2016 |
mlelstv | Enforce alignment requirements that are violated in some cases.
For machines that don't need strict alignment (i386,amd64,vax,m68k) this
is a no-op.
Fixes PR kern/50766 but should be improved.
|
| 1.347 | 10-Jun-2016 |
ozaki-r | branches: 1.347.2;
Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.346 | 10-Jun-2016 |
ozaki-r | Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.345 | 15-Feb-2016 |
rtr | Reduce code duplication.
Split creation of IPv4-Mapped IPv6 addresses into its own function
and use it.
No functional change intended. As posted to tech-net@
|
| 1.344 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.343 | 24-Jul-2015 |
matt | Make sure that snd_win doesn't go negative.
|
| 1.342 | 15-Jul-2015 |
ozaki-r | Remove unused arguments and the associated code from nd6_nud_hint()
from OpenBSD
|
| 1.341 | 24-May-2015 |
rtr | remove transitional functions in{,6}_pcbconnect_m() that were used in
converting protocol user requests to accept sockaddr instead of mbufs.
remove tcp_input copy in to mbuf from sockaddr and just copy to sockaddr
to make it possible for the transitional functions to go away.
no version bump since these functions only existed for a short time and
were commented as adapters (they appeared in 7.99.15).
|
| 1.340 | 15-May-2015 |
kefren | Don't try to do PCB lookup for bad checksummed segments
Fixes PR/43510 and PR/48452
|
| 1.339 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.338 | 27-Apr-2015 |
ozaki-r | Introduce in6_selecthlim_rt to consolidate an idiom for rt->rt_ifp
It consolidates a scattered routine:
(rt = rtcache_validate(&in6p->in6p_route)) != NULL ? rt->rt_ifp : NULL
|
| 1.337 | 14-Mar-2015 |
rtr | Move code that is conditional on options INET6 into #ifdef INET6.
* Re-organize some variable declarations to limit #ifdef's.
* Move INET and INET6 code into respective switch cases to simplify
#ifdef INET6.
No intended functional change.
|
| 1.336 | 14-Feb-2015 |
he | Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
|
| 1.335 | 02-Dec-2014 |
christos | use the new printing code.
|
| 1.334 | 08-Aug-2014 |
rtr | branches: 1.334.2; 1.334.4;
split PRU_RCVD function out of pr_generic() usrreq switches and put into
separate functions
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_RCVD with calls to
pr_rcvd()
|
| 1.333 | 30-May-2014 |
rmind | tcp_signature_getsav: handle !ipsec_used case and fix the build (hi christos!).
|
| 1.332 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.331 | 01-Mar-2014 |
maxv | branches: 1.331.2;
';;' -> ';'
no functional change
spotted by my code scanner
ok christos@
|
| 1.330 | 12-Nov-2013 |
kefren | * implement TCP CUBIC congestion control algorithm
* move tcp_sack_newack bits inside reno and newreno_fast_retransmit_newack
* notify ECN peer about cwnd shrink in [new]reno_slow_retransmit
Based on the patch proposed on tech-net@ on Nov 7 with minor improvments:
* adapt wmax for no-fast convergence case
* correct cbrt calculation for big window sizes (>750KB)
|
| 1.329 | 15-Sep-2013 |
martin | Remove unused variable
|
| 1.328 | 29-Aug-2013 |
rmind | Remove SS_ISCONFIRMING, it is unused and TP4 will not come back.
|
| 1.327 | 06-Jun-2013 |
christos | branches: 1.327.2;
merge error paths, pass the address of sav; pointed out by Greg Troxel
|
| 1.326 | 05-Jun-2013 |
christos | IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.325 | 22-Jun-2012 |
christos | branches: 1.325.2;
remove unintended commit (this was to avoid a bug in the hme driver which
I have not been able to reproduce)
|
| 1.324 | 22-Jun-2012 |
christos | PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.323 | 13-Apr-2012 |
yamt | comment
|
| 1.322 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.321 | 11-Jan-2012 |
drochner | branches: 1.321.2; 1.321.6; 1.321.8;
fix build in the (FAST_)IPSEC & TCP_SIGNATURE case
|
| 1.320 | 31-Dec-2011 |
christos | - fix offsetof usage, and redundant defines
- kill pointer casts to 0
|
| 1.319 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.318 | 19-Nov-2011 |
tls | branches: 1.318.2;
First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>. This change includes
the following:
An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are
fixed. Some effort is made to accumulate entropy more quickly at
boot time.
A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.
The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.
An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c). This
is based on Greg Rose's implementation from Qualcomm.
A new random stream generator, nist_ctr_drbg, is provided. It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim. This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.
An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness. The arc4random/arc4randbytes API is deprecated for
in-kernel use. It is replaced by "cprng_strong". The current
cprng_fast implementation wraps the existing arc4random
implementation. The current cprng_strong implementation wraps the
new CTR_DRBG implementation. Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.
In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data(). Performance is expected to improve.
The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.
The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot. There is
approximately a 3/10000 chance of a false positive from these
tests. Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.
A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32). A new kernel will require a new rndctl.
The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.
Manual pages for the new kernel interfaces are forthcoming.
|
| 1.317 | 31-Oct-2011 |
yamt | branches: 1.317.2;
fix a double unlock bug introduced by tcp_input.c rev.1.312.
|
| 1.316 | 31-Aug-2011 |
plunky | NULL does not need a cast
|
| 1.315 | 17-Jul-2011 |
joerg | Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.314 | 25-May-2011 |
gdt | Remove erroneous additional tick in RTO estimation. The variable
ts_rtt is 1 plus the RTT, so that 0 can mean invalid measurement.
However, the code failed to subtract the 1 back out before use. With
this change, TCP from Massachusetts to France now typically has 1s RTO
values, rather than 1.5s.
This bug was found and fixed by Bev Schwartz of BBN. This material is
based upon work supported by the Defense Advanced Research Projects
Agency and Space and Naval Warfare Systems Center, Pacific, under
Contract No. N66001-09-C-2073. Approved for Public Release,
Distribution Unlimited
|
| 1.313 | 17-May-2011 |
dholland | typo in comment
|
| 1.312 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.311 | 25-Apr-2011 |
yamt | tcp_input: simplify redundant assignment. no functional changes.
|
| 1.310 | 20-Apr-2011 |
wiz | Fix typos.
|
| 1.309 | 20-Apr-2011 |
gdt | Rewrite comments about TCP RTO calculations.
Long ago, the storage representations of srtt and rttvar were changed
from the 4.4BSD scheme, and the comments are out of sync with the
code. This commit rewrites most of the comments that explain the RTO
calculations, and points out some issues in the code.
Joint work with Bev Schwartz of BBN (original analysis and comments),
but I have rewritten and extended them, so errors are mine.
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073. Approved for Public
Release, Distribution Unlimited
|
| 1.308 | 14-Apr-2011 |
yamt | comments
|
| 1.307 | 09-Mar-2011 |
yamt | fix a typo in rev.1.283, which broke tcp dupack and duppack statistics.
|
| 1.306 | 02-Dec-2010 |
plunky | branches: 1.306.2;
fix potential mbuf overflow, from Alexander Danilov on tech-net
|
| 1.305 | 26-May-2010 |
bouyer | Make sure SYN_CACHE_TIMER_ARM() has been run before calling syn_cache_put()
as it will reschedule the timer. Fixes PR kern/43318.
|
| 1.304 | 21-Apr-2010 |
bouyer | syn_cache_put(): defer all pool_put() to the callout. Reschedule
the callout if needed so frees are not delayed too much.
syn_cache_timer(): we can't call syn_cache_put() here any more,
so move code deleted from syn_cache_put() here.
Avoid KASSERT() in kern_timeout.c because pool_put() is called from
ipintr context, as reported in
http://mail-index.netbsd.org/tech-kern/2010/03/19/msg007762.html
Thanks to Andrew Doran and Mindaugas Rasiukevicius for help and review.
|
| 1.303 | 16-Apr-2010 |
rmind | tcp_input: set ECE flag even if CWR flag is active.
Submitted by Richard Scheffenegger in PR/43150.
|
| 1.302 | 01-Apr-2010 |
tls | Oops. Fix LOCKDEBUG panic -- and spurious calls to tcp_output()! -- in
previous. Be careful with that {}, Eugene.
|
| 1.301 | 01-Apr-2010 |
tls | After discussion with ad@: it appears that KERNEL_LOCK also protects
the driver output path (that is, ifp->if_output()). In the case of
entry through the socket code, we are fine, because pru_usrreq takes
KERNEL_LOCK. However, there are a few other ways to cause output
which require protection:
1) direct calls to tcp_output() in tcp_input()
2) fast-forwarding code (ip_flow) -- protected elsewise
against itself by the softnet lock.
3) *Possibly* the ARP code. I have currently persuaded
myself that it is safe because of how it's called.
4) Possibly the ICMP code.
This change addresses #1 and #2.
|
| 1.300 | 26-Jan-2010 |
pooka | branches: 1.300.2; 1.300.4;
tcp sockbuf autoscaling was initially added turned off because it
was experimental. People (including myself) have been running with
it turned on for eons now, so flip the default to enabled.
|
| 1.299 | 09-Sep-2009 |
darran | Make tcp msl (max segment life) tunable via sysctl net.inet.tcp.msl.
Okayed by tls@.
|
| 1.298 | 18-Jul-2009 |
minskim | Check the minimum ttl only when pcb is available.
|
| 1.297 | 17-Jul-2009 |
minskim | Add the IP_MINTTL socket option.
The IP_MINTTL option may be used on SOCK_STREAM sockets to discard
packets with a TTL lower than the option value. This can be used to
implement the Generalized TTL Security Mechanism (GTSM) according to
RFC 3682.
OK'ed by christos@.
|
| 1.296 | 20-Jun-2009 |
christos | Follow exactly the recommendation of draft-ietf-tcpm-tcpsecure-11.txt:
Don't check gainst the last ack received, but the expected sequence number.
This makes RST handling independent of delayed ACK. From Joanne M Mikkelson.
|
| 1.295 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.294 | 18-Mar-2009 |
cegger | bcmp -> memcmp
|
| 1.293 | 15-Mar-2009 |
cegger | ansify function definitions
|
| 1.292 | 29-Jan-2009 |
pooka | branches: 1.292.2;
stinkset purge: POOL_INIT -> pool_init
also, make the syncache pool static in scope
|
| 1.291 | 04-Aug-2008 |
tls | branches: 1.291.2; 1.291.4; 1.291.8;
Unlock reassembly queue before calling sorwakeup(), not after. In unusual
cases with in-kernel consumers which might send data on the same socket,
we can deadlock on the reassembly queue otherwise (observed while testing
accept filters).
|
| 1.290 | 28-Jul-2008 |
matt | Reacquire softnet_lock after calling soabort which returns with the socket
unlocked.
|
| 1.289 | 04-Jul-2008 |
ad | branches: 1.289.2;
tcp_input: add a couple of assertions.
|
| 1.288 | 03-Jul-2008 |
ad | syn_cache_get: remove new endpoint's socket from head's queue if aborting
the connection. Should fix KASSERT(so->so_head == NULL).
|
| 1.287 | 28-Apr-2008 |
martin | branches: 1.287.2; 1.287.4;
Remove clause 3 and 4 from TNF licenses
|
| 1.286 | 24-Apr-2008 |
ad | branches: 1.286.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.285 | 23-Apr-2008 |
thorpej | Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.284 | 12-Apr-2008 |
thorpej | branches: 1.284.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.283 | 08-Apr-2008 |
thorpej | Change TCP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old tcpstat structure; old netstat
binaries will continue to work properly.
|
| 1.282 | 01-Mar-2008 |
rmind | Welcome to 4.99.55:
- Add a lot of missing selinit() and seldestroy() calls.
- Merge selwakeup() and selnotify() calls into a single selnotify().
- Add an additional 'events' argument to selnotify() call. It will
indicate which event (POLL_IN, POLL_OUT, etc) happen. If unknown,
zero may be used.
Note: please pass appropriate value of 'events' where possible.
Proposed on: <tech-kern>
|
| 1.281 | 27-Feb-2008 |
matt | Convert stragglers to ansi definitions from old-style definitons.
Remember that func() is not ansi, func(void) is.
|
| 1.280 | 20-Feb-2008 |
yamt | branches: 1.280.2; 1.280.6;
make TCP_SETUP_ACK, ICMP_CHECK, TCP_FIELDS_TO_HOST, and TCP_FIELDS_TO_NET
static functions.
|
| 1.279 | 05-Feb-2008 |
yamt | - start tcp timestamp from 1 instead of 0.
- add a comment to explain why:
+ * We start with 1, because 0 doesn't work with linux, which
+ * considers timestamp 0 in a SYN packet as a bug and disables
+ * timestamps.
|
| 1.278 | 04-Feb-2008 |
yamt | redo tcp_input.c rev.1.230 correctly.
revision 1.230
date: 2005/06/30 02:58:28; author: christos; state: Exp; lines: +20 -4
Normalize our PAWS code with Free and Open, as mentioned in tech-security.
reviewed by christos@ and matt@.
|
| 1.277 | 29-Jan-2008 |
yamt | revert tcp_output.c 1.253 because it has an ill effect when sending
small (not full-sized) segments.
http://mail-index.NetBSD.org/tech-net/2008/01/27/0009.html
|
| 1.276 | 14-Jan-2008 |
dyoung | Use rtcache_validate() instead of rtcache_getrt(). Shorten staircase
in in_losing().
|
| 1.275 | 20-Dec-2007 |
martin | A few missing ifdefs to make non-INET6 kernels build again.
|
| 1.274 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.273 | 16-Dec-2007 |
elad | Really fix low port allocation, by always passing a valid lwp to
in_pcbbind().
Okay dyoung@.
Note that the network code is another candidate for major cleanup... also
note that this issue is likely to be present in netinet6 code, too.
|
| 1.272 | 09-Nov-2007 |
dyoung | branches: 1.272.2; 1.272.6;
Change macros SYN_CACHE_PUT() and SYN_CACHE_RM() into inline
subroutines syn_cache_put() and syn_cache_rm().
|
| 1.271 | 04-Nov-2007 |
rmind | branches: 1.271.2;
Pick the smallest possible TCP window scaling factor that will still allow
us to scale up to sb_max. This might fix the problems with some firewalls.
Taken from FreeBSD (silby).
OK by <dyoung>.
|
| 1.270 | 02-Aug-2007 |
yamt | branches: 1.270.2; 1.270.4; 1.270.8; 1.270.10;
our tcp timestamps are in PR_SLOWHZ, not HZ.
|
| 1.269 | 02-Aug-2007 |
rmind | TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.268 | 09-Jul-2007 |
ad | branches: 1.268.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.267 | 20-Jun-2007 |
christos | - per socket keepalive settings
- settable connection establishment timeout
|
| 1.266 | 18-May-2007 |
riz | Fix compilation in the TCP_SIGNATURE case:
- don't use void * for pointer arithmetic
- don't try to modify const parameters
A kernel with 'options TCP_SIGNATURE' works as well as it ever did, now.
(ie, clunky, but passable)
|
| 1.265 | 18-May-2007 |
riz | Revert a small part of revision 1.254 - remove const qualifier from
the struct tcphdr * argument of tcp_dooptions(). RFC2385 support
(options TCP_SIGNATURE) needs to modify the header during options
processing, and this revision broke it.
OK yamt@.
|
| 1.264 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.263 | 12-Mar-2007 |
ad | branches: 1.263.2;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.262 | 04-Mar-2007 |
christos | branches: 1.262.2;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.261 | 22-Feb-2007 |
thorpej | TRUE -> true, FALSE -> false
|
| 1.260 | 10-Feb-2007 |
degroote | branches: 1.260.2;
Commit my SoC work
Add ipv6 support for fast_ipsec
Note that currently, packet with extensions headers are not correctly
supported
Change the ipcomp logic
|
| 1.259 | 15-Dec-2006 |
joerg | Introduce new helper functions to abstract the route caching.
rtcache_init and rtcache_init_noclone lookup ro_dst and store
the result in ro_rt, taking care of the reference counting and
calling the domain specific route cache.
rtcache_free checks if a route was cashed and frees the reference.
rtcache_copy copies ro_dst of the given struct route, checking that
enough space is available and incrementing the reference count of the
cached rtentry if necessary.
rtcache_check validates that the cached route is still up. If it isn't,
it tries to look it up again. Afterwards ro_rt is either a valid again
or NULL.
rtcache_copy is used internally.
Adjust to callers of rtalloc/rtflush in the tree to check the sanity of
ro_dst first (if necessary). If it doesn't fit the expectations, free
the cache, otherwise check if the cached route is still valid. After
that combination, a single check for ro_rt == NULL is enough to decide
whether a new lookup needs to be done with a different ro_dst.
Make the route checking in gre stricter by repeating the loop check
after revalidation.
Remove some unused RADIX_MPATH code in in6_src.c. The logic is slightly
changed here to first validate the route and check RTF_GATEWAY
afterwards. This is sementically equivalent though.
etherip doesn't need sc_route_expire similiar to the gif changes from
dyoung@ earlier.
Based on the earlier patch from dyoung@, reviewed and discussed with
him.
|
| 1.258 | 09-Dec-2006 |
dyoung | Here are various changes designed to protect against bad IPv4
routing caused by stale route caches (struct route). Route caches
are sprinkled throughout PCBs, the IP fast-forwarding table, and
IP tunnel interfaces (gre, gif, stf).
Stale IPv6 and ISO route caches will be treated by separate patches.
Thank you to Christoph Badura for suggesting the general approach
to invalidating route caches that I take here.
Here are the details:
Add hooks to struct domain for tracking and for invalidating each
domain's route caches: dom_rtcache, dom_rtflush, and dom_rtflushall.
Introduce helper subroutines, rtflush(ro) for invalidating a route
cache, rtflushall(family) for invalidating all route caches in a
routing domain, and rtcache(ro) for notifying the domain of a new
cached route.
Chain together all IPv4 route caches where ro_rt != NULL. Provide
in_rtcache() for adding a route to the chain. Provide in_rtflush()
and in_rtflushall() for invalidating IPv4 route caches. In
in_rtflush(), set ro_rt to NULL, and remove the route from the
chain. In in_rtflushall(), walk the chain and remove every route
cache.
In rtrequest1(), call rtflushall() to invalidate route caches when
a route is added.
In gif(4), discard the workaround for stale caches that involves
expiring them every so often.
Replace the pattern 'RTFREE(ro->ro_rt); ro->ro_rt = NULL;' with a
call to rtflush(ro).
Update ipflow_fastforward() and all other users of route caches so
that they expect a cached route, ro->ro_rt, to turn to NULL.
Take care when moving a 'struct route' to rtflush() the source and
to rtcache() the destination.
In domain initializers, use .dom_xxx tags.
KNF here and there.
|
| 1.257 | 06-Dec-2006 |
yamt | add some more tcp mowners.
|
| 1.256 | 06-Dec-2006 |
yamt | - make tcp_reass static.
- constify.
|
| 1.255 | 16-Nov-2006 |
christos | branches: 1.255.2; 1.255.4;
__unused removal on arguments; approved by core.
|
| 1.254 | 21-Oct-2006 |
yamt | - constify.
- make tcp_dooptions and tcpipqent_pool static.
|
| 1.253 | 17-Oct-2006 |
yamt | tcp_input: if we have SACK, don't enter fastrecovery on three dupacks.
otherwise, we can enter fastrecovery due to DSACKs, which we treat
as dupacks here. PR/34748. reviewed by Rui Paulo.
|
| 1.252 | 15-Oct-2006 |
rpaulo | Move comments to proper places.
|
| 1.251 | 15-Oct-2006 |
rpaulo | Add a new tcp_congctl(9) structure member for congestion experienced callback.
Needed by HSTCP.
|
| 1.250 | 12-Oct-2006 |
rpaulo | PR 34776: don't accept TCP connections to broadcast addresses.
Move the multicast/broadcast check above (before creating a syn_cache entry)
By Yasuoka Yasuoka.
|
| 1.249 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.248 | 10-Oct-2006 |
rpaulo | Revert previous. The check is now done in tcp_congctl.
|
| 1.247 | 10-Oct-2006 |
yamt | tcp_input: don't call congctl->newack when doing fast retransmit.
|
| 1.246 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.245 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.244 | 05-Sep-2006 |
rpaulo | branches: 1.244.2; 1.244.4;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.243 | 07-Jun-2006 |
kardel | merge FreeBSD timecounters from branch simonb-timecounters
- struct timeval time is gone
time.tv_sec -> time_second
- struct timeval mono_time is gone
mono_time.tv_sec -> time_uptime
- access to time via
{get,}{micro,nano,bin}time()
get* versions are fast but less precise
- support NTP nanokernel implementation (NTP API 4)
- further reading:
Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf
NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
|
| 1.242 | 27-May-2006 |
bouyer | Revert rev 1.241: calling m_makewritable() in tcp_input causes problems when
it has to change the mbuf chain. I experience hard hang on a Xen2 domU after
TCP connections have been closed, and a crash has been reported which may be
caused by this too.
|
| 1.241 | 25-May-2006 |
bouyer | If we're going to byteswap fields in the TCP header, make sure the mbuf
area is writable first.
|
| 1.240 | 15-Apr-2006 |
christos | branches: 1.240.2;
Coverity CID 1152: Add KASSERT before deref.
|
| 1.239 | 18-Feb-2006 |
rpaulo | branches: 1.239.2; 1.239.4; 1.239.6;
PR 13952: Noritoshi Demizu: correct the TCP window information update check.
|
| 1.238 | 02-Feb-2006 |
riz | branches: 1.238.2;
If TCP_SIGNATURE is defined, include netinet6/scope6_var.h for the
prototype of in6_clearscope(). Kernels with options TCP_SIGNATURE now
compile again after the IPv6 scoped address changes.
|
| 1.237 | 15-Nov-2005 |
dsl | branches: 1.237.2; 1.237.4;
Pass the current process structure to in_pcbconnect() so that it can
pass it to in_pcbbind() so that can allocate a low numbered port
if setsockopt() has been used to set IP_PORTRANGE to IP_PORTRANGE_LOW.
While there, fail in_pcbconnect() if the in_pcbbind() fails - rather
than sending the request out from a port of zero.
This has been largely broken since the socket option was added in 1998.
|
| 1.236 | 12-Aug-2005 |
christos | branches: 1.236.6;
If called from syn_cache_add, we need to initialize t_state before calling
tcp_dooptions. Pointed out by yamt.
|
| 1.235 | 12-Aug-2005 |
hubertf | Clarify comment that "the protocol specification dated September, 1981"
is really RFC 793.
|
| 1.234 | 11-Aug-2005 |
christos | Don't process TCP options in SYN packets after the connection has
been established. (FreeBSD-SA-05:15.tcp)
|
| 1.233 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.232 | 10-Aug-2005 |
yamt | device independent part of ipv6 rx checksum offloading.
|
| 1.231 | 19-Jul-2005 |
christos | Implement PMTU checks from:
http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.
From OpenBSD.
|
| 1.230 | 30-Jun-2005 |
christos | branches: 1.230.2;
Normalize our PAWS code with Free and Open, as mentioned in tech-security.
|
| 1.229 | 06-Jun-2005 |
yamt | tcp_input: don't overload opti.ts_ecr.
|
| 1.228 | 29-May-2005 |
christos | - add const
- remove bogus casts
- avoid nested variables
|
| 1.227 | 26-Apr-2005 |
manu | Fix build problem after recent NAT-T changes
|
| 1.226 | 03-Apr-2005 |
yamt | tcp_input: update a comment to match with the code.
|
| 1.225 | 29-Mar-2005 |
yamt | protect tcpipqent with splvm.
|
| 1.224 | 16-Mar-2005 |
yamt | branches: 1.224.2;
simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]),
build it directly from the segment list when needed.
|
| 1.223 | 02-Mar-2005 |
mycroft | Copyright maintenance.
|
| 1.222 | 28-Feb-2005 |
jonathan | Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.221 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.220 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.219 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.218 | 28-Jan-2005 |
mycroft | Several changes based on comparison with NS:
1) dupseg_fix_=true from NS: do not count a segment with completely duplicate
data as a duplicate ack. This can occur due to duplicate packets in the
network, or due to fast retransmit from the other side.
2) dupack_reset_=false from NS: do not reset the duplicate ack counter or exit
fast recovery if we happen to get data or a window update along with a
duplicate ack.
3) In the "very old ack" case that itojun added, send an ACK before dropping
the segment, to try to update the other side's send sequence number.
4) Check the ssthresh crossover point with >= rather than >. Otherwise we
start to do "exponential" growth immediately following recovery, where we
should be doing "linear". This is what NS does.
|
| 1.217 | 27-Jan-2005 |
mycroft | There is no reason to adjust ts_recent_age for ts_timebase; it's strictly an
internal variable.
|
| 1.216 | 27-Jan-2005 |
mycroft | Do the other TCP_PAWS_IDLE check unsigned as well. It doesn't do us any harm,
and it could detect even older time stamps. (Really, to be 100% correct, there
should be a timer that clears these out -- but it probably doesn't matter in
the real world.)
|
| 1.215 | 27-Jan-2005 |
mycroft | Also check whether an echoed RTT is very large -- this *could* cause the
smoothing function to overflow. I use TCP_PAWS_IDLE (24 days) for this.
|
| 1.214 | 27-Jan-2005 |
mycroft | Introduce a new state variable, t_partialacks. It has 3 states:
* t_partialacks<0 means we are not in fast recovery.
* t_partialacks==0 means we are in fast recovery, but we have not received
any partial acks yet.
* t_partialacks>0 means we are in fast recovery, and we have received
partial acks.
This is used to implement 2 changes in RFC 3782:
* We keep the notion that we are in fast recovery separate from t_dupacks, so
it is not reset due to out-of-order acks. (This affects both the Reno and
NewReno cases.)
* We only reset the retransmit timer on the first partial ack -- preventing us
from possibly taking one RTO per segment once fast recovery is initiated.
As before, it is hard to measure any difference between Reno and NewReno in the
real-world cases that I've tested.
|
| 1.213 | 26-Jan-2005 |
mycroft | Fix two problems in our TCP stack:
1) If an echoed RFC 1323 time stamp appears to be later than the current time,
ignore it and fall back to old-style RTT calculation. This prevents ending
up with a negative RTT and panicking later.
2) Fix NewReno. This involves a few changes:
a) Implement the send_high variable in RFC 2582. Our implementation is
subtly different; it is one *past* the last sequence number transmitted
rather than being equal to it. This simplifies some logic and makes
the code smaller. Additional logic was required to prevent sequence
number wraparound problems; this is not mentioned in RFC 2582.
b) Make sure we reset t_dupacks on new acks, but *not* on a partial ack.
All of the new ack code is pushed out into tcp_newreno(). (Later this
will probably be a pluggable function.) Thus t_dupacks keeps track of
whether we're in fast recovery all the time, with Reno or NewReno, which
keeps some logic simpler.
c) We do not need to update snd_recover when we're not in fast recovery.
See tech-net for an explanation of this.
d) In the gratuitous fast retransmit prevention case, do not send a packet.
RFC 2582 specifically says that we should "do nothing".
e) Do not inflate the congestion window on a partial ack. (This is done by
testing t_dupacks to see whether we're still in fast recovery.)
This brings the performance of NewReno back up to the same as Reno in a few
random test cases (e.g. transferring peer-to-peer over my wireless network).
I have not concocted a good test case for the behavior specific to NewReno.
|
| 1.212 | 21-Dec-2004 |
yamt | branches: 1.212.2; 1.212.4;
factor out receive side tcp/udp checksum handling code so that they
can be used by eg. packet filters.
reviewed by Christos Zoulas on tech-net@.
(slightly tweaked since then to make tcp and udp similar.)
|
| 1.211 | 18-Dec-2004 |
yamt | tcp_input: add missing loopback checksum omission code for ipv6.
|
| 1.210 | 15-Dec-2004 |
thorpej | Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.209 | 15-Sep-2004 |
yamt | fix ipqent pool corruption problems. make tcp reass code use
its own pool of ipqent rather than sharing it with ip reass code.
PR/24782.
|
| 1.208 | 26-Jun-2004 |
itojun | correct TCP-MD5 support. Jeff Rizzo
|
| 1.207 | 23-May-2004 |
jonathan | Remove now-unused variable.
|
| 1.206 | 18-May-2004 |
itojun | fix MD5 signature support to actually validate inbound signature, and
drop packet if fails.
|
| 1.205 | 07-May-2004 |
jonathan | Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.204 | 27-Apr-2004 |
matt | When a packet is received that overlaps the left side of the window,
check for RST *before* trimming data and adjust its sequence number.
|
| 1.203 | 26-Apr-2004 |
itojun | make TCP MD5 signature work with KAME IPSEC (#define IPSEC).
support IPv6 if KAME IPSEC (RFC is not explicit about how we make data stream
for checksum with IPv6, but i'm pretty sure using normal pseudo-header is the
right thing).
XXX
current TCP MD5 signature code has giant flaw:
it does not validate signature on input (can't believe it! what is the point?)
|
| 1.202 | 26-Apr-2004 |
matt | Remove #else clause of __STDC__
|
| 1.201 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.200 | 25-Apr-2004 |
simonb | Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.199 | 25-Apr-2004 |
itojun | fix how we send RST against ACK. markus@openbsd
|
| 1.198 | 25-Apr-2004 |
itojun | indent for little bit better readability
|
| 1.197 | 24-Apr-2004 |
itojun | fix comment; we no longer move ip+tcp into the same mbuf
|
| 1.196 | 22-Apr-2004 |
ragge | Avoid performance problem in tcp_reass() when appending mbufs to a chain
by keeping a pointer to the last mbuf in the chain.
|
| 1.195 | 20-Apr-2004 |
itojun | follow draft-ietf-tcpm-tcpsecure-00.txt 3.2 (B):
if SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.194 | 20-Apr-2004 |
itojun | - respond to RST by ACK, as suggested in NISCC recommendation
- rate-limit ACKs against RSTs and SYNs
|
| 1.193 | 17-Apr-2004 |
matt | If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
|
| 1.192 | 14-Apr-2004 |
ragge | Add back one line which was accidentially removed (by me) a while ago.
Spotted by Markus Friedl (markus at openbsd.org).
|
| 1.191 | 29-Mar-2004 |
atatat | Make these compile without INET. tcp_input probably needs a lot more
work...
|
| 1.190 | 10-Mar-2004 |
drochner | branches: 1.190.2;
fix tcp/udp checksum test in the M_CSUM_NO_PSEUDOHDR case
(this can never have worked)
now I can use a "bge" gigabit interface with hw checksumming
ttcp-t: 2147483648 bytes in 18.31 real seconds = 114527.11 KB/sec +++
woow!
|
| 1.189 | 26-Feb-2004 |
itojun | KNF
|
| 1.188 | 02-Jan-2004 |
itojun | some corrections from markus@openbsd;
- callout_ack() was called with wrong argument
- no need for xor with timestamp as we are using arc4random()
- minor typo/cleanup
|
| 1.187 | 19-Nov-2003 |
jonathan | Footwork for fast-ipsec and IPv6: when compiling sys/netinet/tcp_input.c
for both FAST_IPSEC and INET6, include <netipsec/ipsec6.h>.
|
| 1.186 | 24-Oct-2003 |
ragge | Fix the bug in the tcp transmit prediction code.
During testing the prediction counters show a hit-rate on about 85% for
packets sent on a local LAN, and better than 99% for intercontinental
high-speed bulk traffic (!).
|
| 1.185 | 23-Oct-2003 |
mycroft | Remove all the code to maintain ia_inpcbs. This information was only used to
close sockets on address changes, which was deemed to be a bad idea and was
summarily removed, so there is no point in wasting effort on maintaining it
any more.
|
| 1.184 | 10-Sep-2003 |
itojun | cut-and-paste error. Valeriy E. Ushakov
|
| 1.183 | 10-Sep-2003 |
itojun | if IPsec inbound policy mismatches, respond to SYN with RST (instead of
just dropping it), allow client to react quickly.
|
| 1.182 | 06-Sep-2003 |
itojun | clarify flowlabel handling
|
| 1.181 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.180 | 22-Aug-2003 |
itojun | make sure so is properly initialized
|
| 1.179 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.178 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.177 | 22-Aug-2003 |
jonathan | Replace the set_socket() method of passing an extra struct socket*
argument to ip6_output() with a new explicit struct in6pcb* argument.
(The underlying socket can be obtained via in6pcb->inp6_socket.)
In preparation for fast-ipsec. Reviewed by itojun.
|
| 1.176 | 21-Aug-2003 |
jonathan | Honour the M_CSUM_NO_PSEUDOHDR, if set on inbound TCP and UDP packets.
Tested against bcm5700 with patched if_bge.c.
|
| 1.175 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.174 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.173 | 20-Jul-2003 |
he | As a temporary workaround, apply the fix from PR#20390, thereby
cooperating with the callout code in working around the race
condition caused by the TCP code's use of the callout facility.
Instead of unconditionally releasing memory in tcp_close() and
SYN_CACHE_PUT(), check whether any of the related callout handlers
are about to be invoked (but have not yet done callout_ack()), and
if so, just mark the associated data structure (tcpcb or syn cache
entry) as "dead", and test for this (and release storage) in the
callout handler functions.
|
| 1.172 | 02-Jul-2003 |
ragge | Fix previous bug. Thanks to Enami for spotting the (obvious) error, and
to other people with much help with bug reports etc.
While fixing, change some of the code I added last time to make it
cleaner and simpler.
|
| 1.171 | 29-Jun-2003 |
fvdl | branches: 1.171.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.170 | 29-Jun-2003 |
ragge | Add code to remember where in the send queue of mbufs the last packet was
sent from. This change avoid a linear search through all mbufs when using
large TCP windows, and therefore permit high-speed connections on long
distances.
Tested on a 1 Gigabit connection between Lule� and San Francisco, a distance
of about 15000km. With TCP windows of just over 20 Mbytes it could keep up
with 950Mbit/s.
After discussions with Matt Thomas and Jason Thorpe.
|
| 1.169 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.168 | 30-May-2003 |
itojun | inherit IPV6_V6ONLY from listening socket. PR 21713
|
| 1.167 | 17-May-2003 |
itojun | no need for ip_v recovery in output path too
(tcp_template includes ip_v setting)
|
| 1.166 | 17-May-2003 |
itojun | ip checksum logic no longer damage ip_v
|
| 1.165 | 16-May-2003 |
itojun | use strlcpy
|
| 1.164 | 14-May-2003 |
itojun | always use PULLDOWN_TEST codepath.
|
| 1.163 | 01-Mar-2003 |
thorpej | Allow TCP connections to hosts on a local network to use a larger
slow start initial window. Default this larger initial window to
4 packets, allowing it to be adjusted with net.inet.tcp.init_win_local.
|
| 1.162 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.161 | 25-Feb-2003 |
he | Swap neighboring lines of callout_init() and bzero() of container
struct in syn_cache_add(); the bzero() invalidates whatever
callout_init() has done (which might matter, but presently doesn't).
|
| 1.160 | 04-Jan-2003 |
wiz | Spell output with two ts.
|
| 1.159 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.158 | 22-Oct-2002 |
thorpej | Make sure TF_REQ_TSTMP and TF_REQ_SCALE get set correctly in the new
TCPCB in the passive-open case.
Fixes PR 18677.
|
| 1.157 | 22-Oct-2002 |
simonb | In tcp_input():
Remove the set-but-not-used "proto" variable.
Guard the "ostate" variable in #ifdef TCP_DEBUG.
Remove the set-but-not-used "parentinpcb" variable in syn_cache_get().
|
| 1.156 | 16-Oct-2002 |
itojun | correct log_refused check (TH_SYN, !TH_RST and !TH_ACK). PR 18669
|
| 1.155 | 11-Sep-2002 |
itojun | correct signedness mixup in pointer passing. sync w/kame
|
| 1.154 | 05-Sep-2002 |
itojun | always consult SS_CANTRCVMORE. PR 18185
|
| 1.153 | 28-Aug-2002 |
thorpej | Fix a problem introduced in rev 1.103, where we recycle a TIME_WAIT
TCPCB .. the fields need to be converted back to net-order, because
the packet is checksummed after the TCPCB lookup happens.
From YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>.
|
| 1.152 | 19-Aug-2002 |
itojun | better sync w/kame on deprecated address handling. check af == AF_INET6.
|
| 1.151 | 19-Aug-2002 |
itojun | pull in deprecated address handling from KAME sys/netinet6/tcp6_input.c.
|
| 1.150 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.149 | 18-Jul-2002 |
wrstuden | When a new connection arrives on a listening port, copy over the
value of the TCP_NODELAY socket option from the listener to the
newly connected connection. Agrees with how Linux & FreeBSD behave,
and goes more with the spirit of accept(2) creating a socket with
the same properties as the listener.
Analysis by Kevin Lahey. Closes PR 17616 by myself.
|
| 1.148 | 03-Jul-2002 |
thorpej | Rename sbappend_stream() to sbappendstream(), per suggestion from
Jonathan Stone.
|
| 1.147 | 03-Jul-2002 |
thorpej | Make insertion of data into socket buffers O(C):
* Keep pointers to the first and last mbufs of the last record in the
socket buffer.
* Use the sb_lastrecord pointer in the sbappend*() family of functions
to avoid traversing the packet chain to find the last record.
* Add a new sbappend_stream() function for stream protocols which
guarantee that there will never be more than one record in the
socket buffer. This function uses the sb_mbtail pointer to perform
the data insertion. Make TCP use sbappend_stream().
On a profiling run, this makes sbappend of a TCP transmission using
a 1M socket buffer go from 50% of the time to .02% of the time.
Thanks to Bill Sommerfeld and YAMAMOTO Takashi for their debugging
assistance!
|
| 1.146 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.145 | 29-Jun-2002 |
yamt | split logging code in order to reduce maximum stack usage.
|
| 1.144 | 11-Jun-2002 |
itojun | share policy-on-pcb for listening socket. sync w/kame
todo: share even more, avoid frequent updates of spidx
|
| 1.143 | 09-Jun-2002 |
itojun | whitespace
|
| 1.142 | 28-May-2002 |
itojun | use arc4random() where possible.
XXX is it necessary to do microtime() on tcp syn cache?
|
| 1.141 | 07-May-2002 |
matt | branches: 1.141.2; 1.141.4;
Change struct ipqe to use TAILQ's instead of LIST's (primarily for TCP's
benefit currently). Rework tcp_reass code to optimize the 4 most likely causes
of out-of-order packets: first OoO pkt, next OoO pkt in seq, OoO pkt is part
of new chuck of OoO packets, and the OoO pkt fills the first hole. Add evcnts
to instrument tcp_reass (enabled by the options TCP_REASS_COUNTERS). This is
part 1/2 of tcp_reass changes.
|
| 1.140 | 24-Mar-2002 |
christos | Change the multicast/broadcast test to happen later, and when we are
in listen mode. Fixes panic with telnet ::1 port, where the port is an
ipv4 open port.
|
| 1.139 | 22-Mar-2002 |
itojun | no need to check in_broadaddr/IN_MULTICAST in dropwithreset label.
suggested by enami
|
| 1.138 | 22-Mar-2002 |
itojun | make sure we don't touch "ip" in IPv6 path
|
| 1.137 | 19-Mar-2002 |
christos | Drop connections to the broadcast address. From BUGTRAQ. This is a security
issue because it can by-pass ipf rules unintentionally.
|
| 1.136 | 12-Mar-2002 |
itojun | support tcp_log_refused for IPv6. From: Andrew Brown <atatat@atatdot.net>
|
| 1.135 | 08-Mar-2002 |
thorpej | Pool deals fairly well with physical memory shortage, but it doesn't
deal with shortages of the VM maps where the backing pages are mapped
(usually kmem_map). Try to deal with this:
* Group all information about the backend allocator for a pool in a
separate structure. The pool references this structure, rather than
the individual fields.
* Change the pool_init() API accordingly, and adjust all callers.
* Link all pools using the same backend allocator on a list.
* The backend allocator is responsible for waiting for physical memory
to become available, but will still fail if it cannot callocate KVA
space for the pages. If this happens, carefully drain all pools using
the same backend allocator, so that some KVA space can be freed.
* Change pool_reclaim() to indicate if it actually succeeded in freeing
some pages, and use that information to make draining easier and more
efficient.
* Get rid of PR_URGENT. There was only one use of it, and it could be
dealt with by the caller.
From art@openbsd.org.
|
| 1.134 | 24-Jan-2002 |
itojun | place NRL copyright notice itself, not a reference to it.
|
| 1.133 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.132 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.131 | 17-Sep-2001 |
thorpej | branches: 1.131.2;
Split the pre-computed ifnet checksum flags into Tx and Rx directions.
Add capabilities bits that indicate an interface can only perform
in-bound TCPv4 or UDPv4 checksums. There is at least one Gig-E chip
for which this is true (Level One LXT-1001), and this is also the
case for the Intel i82559 10/100 Ethernet chips.
|
| 1.130 | 11-Sep-2001 |
thorpej | Use callouts for SYN cache timers, rather than traversing time queues
in tcp_slowtimo().
|
| 1.129 | 10-Sep-2001 |
thorpej | Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.
|
| 1.128 | 10-Sep-2001 |
thorpej | Change the way receive idle time and round trip time are measured.
Instead of incrementing t_idle and t_rtt in tcp_slowtimo(), we now
take a timstamp (via tcp_now) and use subtraction to compute the
delta when we actually need it (using unsigned arithmetic so that
tcp_now wrapping is handled correctly).
Based on similar changes in FreeBSD.
|
| 1.127 | 08-Jul-2001 |
abs | branches: 1.127.2; 1.127.4;
Rename TCPDEBUG to TCP_DEBUG, defopt TCP_DEBUG and TCP_NDEBUG, and
make all usage of tcp_trace dependent on TCP_DEBUG - resulting in
a 31K saving on an INET enabled i386 kernel.
|
| 1.126 | 19-Jun-2001 |
wiz | `existent', not `existant'
|
| 1.125 | 02-Jun-2001 |
thorpej | Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.124 | 08-May-2001 |
itojun | correct faith prefix determination. use sys/netinet/if_faith.c:faithprefix()
to determine. sync with kame.
(without this change, non-faith socket may mistakenly accept for-faith traffic)
|
| 1.123 | 20-Mar-2001 |
thorpej | Two changes, designed to make us even more resilient against TCP
ISS attacks (which we already fend off quite well).
1. First-cut implementation of RFC1948, Steve Bellovin's cryptographic
hash method of generating TCP ISS values. Note, this code is experimental
and disabled by default (experimental enough that I don't export the
variable via sysctl yet, either). There are a couple of issues I'd
like to discuss with Steve, so this code should only be used by people
who really know what they're doing.
2. Per a recent thread on Bugtraq, it's possible to determine a system's
uptime by snooping the RFC1323 TCP timestamp options sent by a host; in
4.4BSD, timestamps are created by incrementing the tcp_now variable
at 2 Hz; there's even a company out there that uses this to determine
web server uptime. According to Newsham's paper "The Problem With
Random Increments", while NetBSD's TCP ISS generation method is much
better than the "random increment" method used by FreeBSD and OpenBSD,
it is still theoretically possible to mount an attack against NetBSD's
method if the attacker knows how many times the tcp_iss_seq variable
has been incremented. By not leaking uptime information, we can make
that much harder to determine. So, we avoid the leak by giving each
TCP connection a timebase of 0.
|
| 1.122 | 24-Jan-2001 |
itojun | branches: 1.122.2;
- record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.121 | 10-Dec-2000 |
itojun | remove NRL code leftover. sync with kame
|
| 1.120 | 19-Oct-2000 |
itojun | remove #ifdef TCP6. it is not likely for us to bring in sys/netinet6/tcp6*.c
(separate TCP/IPv6 stack) into netbsd-current.
|
| 1.119 | 17-Oct-2000 |
itojun | be more friendly with INET-less build.
XXX we need to do more to do a working INET-less build
|
| 1.118 | 17-Oct-2000 |
thorpej | Add an IP_MTUDISC flag to the flags that can be passed to
ip_output(). This flag, if set, causes ip_output() to set
DF in the IP header if the MTU in the route is not locked.
This allows a bunch of redundant code, which I was never
really all that happy about adding in the first place, to
be eliminated.
Inspired by a similar change made by provos@openbsd.org when
he integrated NetBSD's Path MTU Discovery code into OpenBSD.
|
| 1.117 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.116 | 27-Jul-2000 |
itojun | implement net.inet.tcp.rstppslimit to limit TCP RSTs by packet-per-second
basis. default: 100pps
set default value for net.inet.tcp.rstratelimit to 0 (disabled),
NOTE: it does not work right for smaller-than-1/hz interval. maybe we should
nuke it, or make it impossible to set smaller-than-1/hz value.
|
| 1.115 | 27-Jul-2000 |
itojun | be proactive about unspecified IPv6 source address. pcb layer uses
unspecified address (::) to mean "unbounded" or "unconnected",
and can be confused by packets from outside.
use of :: as source is not documented well in IPv6 specification.
not sure if it presents a real threat. the worst case scenario is a DoS
against TCP listening socket:
- outsider transmit TCP SYN with :: as IPv6 source
- receiving side creates TCP control block with:
local address = my addres
remote address = :: (meaning "unconnected")
state = SYN_RCVD
note that SYN ACK will not be sent due to ip6_output() filter.
this stays until it timeouts.
- the TCP control block prevents listening TCP control block from
being contacted (DoS).
udp6/raw6 socket may have similar problem, but as they are connectionless,
it may too much to filter it out.
|
| 1.114 | 23-Jul-2000 |
itojun | add an DIAGNOSTIC case for MCLBYTES assumption
|
| 1.113 | 09-Jul-2000 |
itojun | be more cautious about tcp option length field. drop bogus ones earlier.
not sure if there is a real threat or not, but it seems that there's
possibility for overrun/underrun (like non-NOP option with optlen > cnt).
|
| 1.112 | 06-Jul-2000 |
itojun | - do not use bitfield for router renumbering header.
- add protection mechanism against ND cache corruption due to bad NUD hints.
- more stats
- icmp6 pps limitation. TOOD: should implement ppsratecheck(9).
|
| 1.111 | 05-Jul-2000 |
thorpej | Fix some zero-vs-NULL confusion.
|
| 1.110 | 02-Jul-2000 |
itojun | repair kernel faithd(8) support. there were two mistakes:
(1) tcp6_input dropped packets for translation
(2) in6_pcblookup_connect was too strict
|
| 1.109 | 30-Jun-2000 |
itojun | remove old mbuf assumption (ip header and tcp header are on the same mbuf).
this is for m_pulldown use. (sync with kame)
|
| 1.108 | 05-May-2000 |
matt | branches: 1.108.4;
remove superfluous test (snd_una is always > iss since th_ack must > iss
(first test at start of case) and th_ack is assigned to snd_una).
|
| 1.107 | 05-May-2000 |
matt | From PR #3733: Only disarm timer if SYN contained the ACK bit since if
it didn't it would be a crossing/simultaneous SYN and doesn't mean the
remote TCP received our SYN.
|
| 1.106 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.105 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.104 | 15-Feb-2000 |
thorpej | Add support for rate-limiting RSTs sent in response to no socket for
an incoming packet. Default minimum interval is 10ms. The interval
is changeable via the "net.inet.tcp.rstratelimit" sysctl variable.
|
| 1.103 | 12-Feb-2000 |
thorpej | In the tcp_input() path:
- Filter out multicast destinations explicitly for every incoming packet,
not just SYNs. Previously, non-SYN multicast destination would be
filtered out as a side effect of PCB lookup. Remove now redundant
similar checks in the dropwithreset case and in syn_cache_add().
- Defer the TCP checksum until we know that we want to process the
packet (i.e. have a non-CLOSED connection or a listen socket).
|
| 1.102 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.101 | 22-Dec-1999 |
itojun | drop IPv6 packets with v4 mapped address on src/dst. they are illegal
and may be used to fool IPv6 implementations (by using ::ffff:127.0.0.1 as
source you may be able to pretend the packet is from local node)
|
| 1.100 | 15-Dec-1999 |
itojun | do not overwrite traffic class field when we write IPv6 version field.
|
| 1.99 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.98 | 11-Dec-1999 |
itojun | implement upper-layer reachability confirmation for IPv6 ND (RFC2461 7.3.1).
fix code to reject "tcp to IPv6 anycast".
sync with recent KAME.
|
| 1.97 | 08-Dec-1999 |
itojun | do not drop from IP header to tcp option until sbappend(), to reduce
requirement to mbuf chain.
part of KAME sync, committed separately for its (possible) impact.
|
| 1.96 | 23-Sep-1999 |
itojun | branches: 1.96.2; 1.96.8;
cleanup and correct TCP MSS consideration with IPsec headers.
MSS advertisement must always be:
max(if mtu) - ip hdr siz - tcp hdr siz
We violated this in the previous code so it was fixed.
tcp_mss_to_advertise() now takes af (af on wire) as its argument,
to compute right ip hdr siz.
tcp_segsize() will take care of IPsec header size.
One thing I'm not really sure is how to handle IPsec header size in
*rxsegsizep (inbound segment size estimation).
The current code subtracts possible *outbound* IPsec size from *rxsegsizep,
hoping that the peer is using the same IPsec policy as me.
It may not be applicable, could TCP gulu please comment...
|
| 1.95 | 10-Sep-1999 |
simonb | s/acknowledgment/acknowledgement/
|
| 1.94 | 26-Aug-1999 |
thorpej | Fix a problem discovered by the snd_recover update fix. A bit of the
New Reno fast recovery code was being executed even when New Reno was
disabled, resulting in an unfortunate interaction with the traditional
fast recovery code, the end resulting being that the very condition
that would trigger the traditional fast recovery mechanism caused fast
recovery to be disabled!
Problem reported by Ted Lemon, and some analytical help from Charles Hannum.
|
| 1.93 | 25-Aug-1999 |
itojun | When listening socket goes away, remove assockated syn cache entires.
Stale syn cache entries are useless because none of them will be used
if there is no listening socket, as tcp_input looks up listening socket by
in_pcblookup*() before looking into syn cache.
This fixes race condition due to dangling socket pointer from syn cache
entries to listening socket (this was introduced when ipsec is merged in).
This should preserve currently implemented behavior (but not 4.4BSD
behavior prior to syn cache).
Tested in KAME repository before commit, but we'd better run some
regression tests.
|
| 1.92 | 23-Aug-1999 |
christos | PR/8254: Wolfgang Rupprecht: Incorrect logging of tcp connections; Fix src/dst
confusion.
|
| 1.91 | 11-Aug-1999 |
thorpej | Fix a few bugs in the TCP New Reno code:
- Make sure that snd_recover is always at least snd_una. If we don't do
this, there can be confusion when sequence numbers wrap around on a
large loss-free data transfer.
- When doing a New Reno retransmit, snd_una hasn't been updated yet,
and the socket's send buffer has not yet dropped off ACK'd data, so
don't muddle with snd_una, so that tcp_output() gets the correct data
offset.
- When doing a New Reno retransmit, make sure the congestion window is
open one segment beyond the ACK'd data, so that we can actually perform
the retransmit.
Partially derived from, although more complete than, similar changes in
OpenBSD, which in turn originated from Tom Henderson <tomh@cs.berkeley.edu>.
|
| 1.90 | 11-Aug-1999 |
thorpej | Make sure the echoed RFC 1323 timestamp is valid before using it to
compute the round trip time. From Mark Allman <mallman@lerc.nasa.gov>.
|
| 1.89 | 22-Jul-1999 |
itojun | - implement IPv6 pmtud, which is necessary for TCP6.
- fix memory leak on SO_DEBUG over TCP.
|
| 1.88 | 17-Jul-1999 |
itojun | no need to include faith.h on non-IPv6 build, so wrap by #ifdef.
(dunno if it's better to always include it or not)
|
| 1.87 | 17-Jul-1999 |
itojun | fix faith interface support. need testing.
(i understand this is a dirty hack, of course)
|
| 1.86 | 14-Jul-1999 |
itojun | Use proper ip protocol # field and tcp hdr on sending RST against SYN,
when ip header and tcp header are not adjacent to each other
(i.e. when ip6 options are attached).
To test this, try
telnet @::1@::1 port
toward a port without responding server. Prior to the fix, the kernel will
generate broken RST packet.
|
| 1.85 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.84 | 02-Jul-1999 |
itojun | avoid "variable not initialized" warnings on some of the platforms.
|
| 1.83 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.82 | 23-May-1999 |
ad | Add new sysctl (net.inet.tcp.log_refused) that when set, causes refused TCP
connections to be logged.
|
| 1.81 | 03-May-1999 |
thorpej | Fix an ininitialized variable that the MIPS compiler caught (but the
SPARC, Alpha, Arm, and i386 compilers missed).
|
| 1.80 | 29-Apr-1999 |
thorpej | Implement retransmit logic for the SYN cache engine. Fixes a rare condition
where one side can think a connection exists, where the other side thinks
the connection was never established.
The original problem was first reported by Ty Sarna in PR #5909. The
original fix I made to the code didn't cover all cases. The problem this
fix addresses was reported by Christoph Badura via private e-mail.
Many thanks to Bill Sommerfeld for helping me to test this code, and
for finding a subtle bug.
|
| 1.79 | 22-Apr-1999 |
simonb | Don't extern sb_max, <sys/socketvar.h> provides a definition.
|
| 1.78 | 09-Apr-1999 |
kml | Ensure that out of window SYNs receive an ACK in responce, rather than
being dropped. This fixes a bug reported by Jason Thorpe.
|
| 1.77 | 05-Feb-1999 |
matt | branches: 1.77.2;
According to Dave Borman, the iss should be using snd_nxt and not rcv_nxt
(from tcp_impl mailing-list).
|
| 1.76 | 04-Feb-1999 |
explorer | REALLY only update the window when we get an ACK. (the old code seemed broken)
|
| 1.75 | 24-Jan-1999 |
thorpej | * Completely rewrite syn_cache_respond().
- Don't use tcp_respond(), instead create the tcp/ip header from scratch,
and send it ourself.
- Reuse the mbuf that carried the SYN, or allocate one if that is not
available.
- Cache the route we look up to do the Path MTU Discovery check, and
transfer the reference to that route to the inpcb when the connection
completes.
* Macro'ize a small, but often repeated code fragment.
|
| 1.74 | 19-Jan-1999 |
mycroft | Don't screw with ip_len; just subtract from it where we actually use the
value.
|
| 1.73 | 19-Jan-1999 |
mycroft | Don't overwrite the checksum fields when checking them. There's no reason to
do this, and it screws up ICMP replies.
XXX The returned IP checksum and length are still wrong.
|
| 1.72 | 18-Dec-1998 |
thorpej | Add a lock around the TCPCB's sequence queue, to prevent tcp_drain()
from corrupting the queue if called from a device's interrupt context.
Similar in nature to the problem reported in PR #5684.
|
| 1.71 | 08-Oct-1998 |
thorpej | Use the pool allocator for ipqent structures.
|
| 1.70 | 06-Oct-1998 |
matt | Fix boolean dyslexic test. Duh!
|
| 1.69 | 06-Oct-1998 |
matt | Add a sysctl for newreno (default to off).
|
| 1.68 | 04-Oct-1998 |
matt | Adapt the NEWRENO changes from the UCSB diffs of BSDI 3.0's TCP
to NetBSD. Ignore the SACK & FACK stuff for now.
|
| 1.67 | 19-Sep-1998 |
mycroft | Fix a typo (not mine) in a comment.
|
| 1.66 | 19-Sep-1998 |
mycroft | If we're in LISTEN state and all of RST, SYN and ACK are clear, send a RST.
|
| 1.65 | 10-Sep-1998 |
mouse | Create tcp.keepidle, tcp.keepintvl, tcp.keepcnt, tcp.slowhz sysctls.
|
| 1.64 | 09-Sep-1998 |
thorpej | Use an algorithm similar to that in tcp_notify() to determine if
syn_cache_unreach() should remove the entry, or just continue on.
Algorithm is to only remove the entry if we've had more than one unreach
error and have retransmitted 3 or more times. This prevents the following
scenario, as noted in PR #5909 (PR from Ty Sarna, scenario from
Charles Hannum):
* Host A sends a SYN.
* Host A retransmits the SYN.
* Host B gets the first SYN and sends a SYN-ACK.
* Host B gets the second SYN and sends a SYN-ACK.
* One of the SYN-ACK bounces with an
ICMP unreachable, causing the `SYN cache' entry to be
removed with no notification.
* Host A receives the other SYN-ACK, sends an ACK, and goes to
ESTABLISHED state.
Should fix PR #5909.
|
| 1.63 | 02-Aug-1998 |
thorpej | Use the pool allocator for syn_cache entries.
|
| 1.62 | 17-Jul-1998 |
thorpej | Clarify that we are using the Loss Window if a retransmission occurred
during the three-way handshake.
|
| 1.61 | 02-Jun-1998 |
thorpej | Add a comment explaining why we do _not_ ACK data that might accompany
a SYN (avoidance of a DoS attack).
|
| 1.60 | 11-May-1998 |
thorpej | Nuke TUBA per my note to tech-net; there's no reason to keep it around.
|
| 1.59 | 07-May-1998 |
thorpej | Rework the syn cache code somewhat:
- Don't use home-grown queue manipulation. Use <sys/queue.h> instead. The
data structures are a little larger, but we are otherwise wasting the
memory chunk anyway (we're already a 64-byte malloc bucket).
- Fix a bug in the cache-is-full case: if the oldest element removed from
the first non-empty bucket was the only element in the bucket, the
bucket wouldn't be removed from the bucket cache, causing queue corruption
later.
- Optimize the syn cache timers by using PRT timers rather than home-grown
decrement-and-propagate timers.
This code is now a fair bit smaller, and significantly easier to read
and understand.
|
| 1.58 | 06-May-1998 |
thorpej | Use macros from tcp_timer.h to manipulate TCP timers, so that their
implementation can be changed easily.
|
| 1.57 | 03-May-1998 |
thorpej | Once again, move a declaration for the benefit of TUBA (grumble).
|
| 1.56 | 02-May-1998 |
thorpej | Oops, move a variable declaration so TUBA won't lose.
|
| 1.55 | 02-May-1998 |
thorpej | Reintroduce the immediate ACK-on-PUSH behavior removed in revision 1.47,
but make the decision to do this dependent on the sysctl variable
net.inet.tcp.ack_on_push, which is disabled by default.
|
| 1.54 | 29-Apr-1998 |
matt | New TCP reassembly code. The new code reduces the memory needed by
out-of-order packets and builds the infrastructure needed for sending
SACK blocks (to be added shortly).
|
| 1.53 | 29-Apr-1998 |
thorpej | Change RFC1323 timestamp update rule per Section 3.4 of RFC1323.bis. Old
rule was to update the timestamp if the sequence numbers are in range. New
rule adds a check that the timestamp is advancing, thus preventing our notion
of the most recent timestamp from incorrectly moving backwards.
|
| 1.52 | 28-Apr-1998 |
thorpej | Log the peer's IP address on received window scale factors larger than
TCP_MAX_WINSHIFT (14), as recommended in Section 2.3 of RFC1323.
|
| 1.51 | 13-Apr-1998 |
kml | Fix to ensure that the correct MSS is advertised for loopback
TCP connections by using the MTU of the interface. Also added
a knob, mss_ifmtu, to force all connections to use the MTU of
the interface to calculate the advertised MSS.
|
| 1.50 | 07-Apr-1998 |
thorpej | Remember any source routes that may have accompanied a SYN.
|
| 1.49 | 03-Apr-1998 |
thorpej | Now that we have a flags word in the syn cache entry, use a flag to indicate
"peer will do timestamps" rather than a bitfield, and give the now-unsed
bit to the hash, making it now 32 bits.
|
| 1.48 | 03-Apr-1998 |
thorpej | Clean up some comments wrt. the syn cache code.
|
| 1.47 | 31-Mar-1998 |
thorpej | Back out a change made some time ago, that would cause the NetBSD TCP
to ACK immediately any packet that arrived with PSH set. This breaks
delayed ACKs in a few specific common cases that delayed ACKs were
supposed to help, and ends up not making much (if any) difference in
the case where where this ACK-on-PSH change was supposed to help.
Per discussion with several members of the TCPIMPL and TCPSAT IETF
working groups.
|
| 1.46 | 31-Mar-1998 |
thorpej | Fix a potential-congestion case in the larger initial congestion window
code, as clarified in the TCPIMPL WG meeting at IETF #41: If the SYN
(active open) or SYN,ACK (passive open) was retransmitted, the initial
congestion window for the first slow start of that connection must be
one segment.
|
| 1.45 | 19-Mar-1998 |
kml | Fix a retransmission bug introduced by the Brakmo and Peterson
RTO estimation changes. Under some circumstances it would return a value
of 0, while the old Van Jacobson RTO code would return a minimum of 3.
This would result in 12 retransmissions, each 1 second apart.
This takes care of those instances, and ensures that t_rttmin is
used everywhere as a lower bound.
|
| 1.44 | 19-Feb-1998 |
thorpej | Update copyright (sigh, should have done this long ago).
|
| 1.43 | 24-Jan-1998 |
mellon | Always set sc->sc_timeout (it was missed in one case). This fixes a problem where SYN cache entries are sometimes timed out almost immediately.
|
| 1.42 | 24-Jan-1998 |
mycroft | Fix an old editing error from merging a bug fix into Lite,
that might cause us to erroneously drop a FIN.
Also, minor changes so the code looks more like Stevens vol 2 figure 28.30.
|
| 1.41 | 21-Jan-1998 |
mellon | Never free the mbuf that we give to tcp_respond(). The previous change corrected an inconsistency but in exactly the wrong way.
|
| 1.40 | 18-Jan-1998 |
mellon | In syn_cache_get(), don't free incoming packet before jumping to resetandabort, but do free it after sending the reset.
|
| 1.39 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.38 | 31-Dec-1997 |
thorpej | Implement a queue for delayed ACK processing. This queue is used in
tcp_fasttimo() in lieu of scanning all open TCP connections.
|
| 1.37 | 11-Dec-1997 |
thorpej | Fix the "stretch ACK violation" bug documented in internet draft
draft-ietf-tcpimpl-prob-02.txt. Also, fix another bug in the header
prediction case where an ACK would not be sent when it should be.
|
| 1.36 | 21-Nov-1997 |
thorpej | Slight change to the previous: just drop the packet in the self-connect
case. Sending an RST to ourselves is a little silly, considering that
we'll just attempt to remove a non-existent compressed state entry and
then drop the packet anyway.
|
| 1.35 | 21-Nov-1997 |
thorpej | In tcp_input(), if the PCB we lookup for an incoming packet is a listen
socket:
- If we received a SYN,ACK, send an RST.
- If we received a SYN, and the connection attempt appears to come from
itself, send an RST, since it cannot possibly be valid.
|
| 1.34 | 08-Nov-1997 |
kml | TCP MSS fixes to provide cleaner slow-start and recovery.
|
| 1.33 | 10-Oct-1997 |
explorer | branches: 1.33.2;
Add hooks to use the kernel random system to generate TCP sequence numbers.
|
| 1.32 | 22-Sep-1997 |
thorpej | Fix several annoyances related to MSS handling in BSD TCP:
- Don't overload t_maxseg. Previous behavior was to set it to the min
of the peer's advertised MSS, our advertised MSS, and tcp_mssdflt
(for non-local networks). This breaks PMTU discovery running on
either host. Instead, remember the MSS we advertise, and use it
as appropriate (in silly window avoidance).
- Per last bullet, split tcp_mss() into several functions for handling
MSS (ours and peer's), and performing various tasks when a connection
becomes ESTABLISHED.
- Introduce a new function, tcp_segsize(), which computes the max size
for every segment transmitted in tcp_output(). This will eventually
be used to hook in PMTU discovery.
|
| 1.31 | 28-Jul-1997 |
thorpej | branches: 1.31.2;
Garbage-collect some "extern"s.
|
| 1.30 | 28-Jul-1997 |
thorpej | Fix a rather severe bug in handling of incoming SYNs for peer/port values
which happen to have a TCB in TIME_WAIT, where an mbuf which had been
advanced past the IP+TCP headers and TCP options would be reused as if
it had not been advanced. Problem found by Juergen Hannken-Illjes, who
also suggested a work-around on which this fix is based.
|
| 1.29 | 23-Jul-1997 |
thorpej | Pull SYN_cache_branch down into the main line.
|
| 1.28 | 06-Jul-1997 |
thorpej | Fix an old and obscure TCP bug, brought to my attention by Bill Fenner,
fixed in FreeBSD by John Polstra:
Fix a bug (apparently very old) that can cause a TCP connection to
be dropped when it has an unusual traffic pattern. For full details
as well as a test case that demonstrates the failure, see the
referenced PR (FreeBSD's kern/3998).
Under certain circumstances involving the persist state, it is
possible for the receive side's tp->rcv_nxt to advance beyond its
tp->rcv_adv. This causes (tp->rcv_adv - tp->rcv_nxt) to become
negative. However, in the code affected by this fix, that difference
was interpreted as an unsigned number by max(). Since it was
negative, it was taken as a huge unsigned number. The effect was
to cause the receiver to believe that its receive window had negative
size, thereby rejecting all received segments including ACKs. As
the test case shows, this led to fruitless retransmissions and
eventually to a dropped connection. Even connections using the
loopback interface could be dropped. The fix substitutes the signed
imax() for the unsigned max() function.
Bill informs me that his research indicates this bug appeared in Reno.
|
| 1.27 | 10-Dec-1996 |
mycroft | branches: 1.27.8;
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.26 | 15-Sep-1996 |
mycroft | Hash unconnected PCBs.
|
| 1.25 | 10-Sep-1996 |
mycroft | If we're in SYN-SENT or SYN-RECEIVED state, don't reset the keepalive
timer until we transition to ESTABLISHED state. Suggested by TCP/IP
vol 3.
|
| 1.24 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.23 | 13-Feb-1996 |
christos | branches: 1.23.4;
netinet prototypes
|
| 1.22 | 31-Jan-1996 |
mycroft | Ignore FIN if not yet connected.
|
| 1.21 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.20 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.19 | 04-Aug-1995 |
mycroft | branches: 1.19.2;
Encapsulate the test for sending a notification in a macro, sb_notify().
|
| 1.18 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.17 | 11-Jun-1995 |
mycroft | Oops. Decrement rtt earlier.
|
| 1.16 | 11-Jun-1995 |
mycroft | As suggested by Brakmo and Peterson:
* Don't add the extra 1/8 of the mss when ramping up the congestion window.
* Scale the RTT values slightly to adjust for rounding errors.
* Set the lower bound of the RTO to RTT+2.
|
| 1.15 | 11-Jun-1995 |
mycroft | Check for inflated congestion window during header prediction, per Bramko and
Peterson.
|
| 1.14 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.13 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.12 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.11 | 14-Oct-1994 |
mycroft | Don't return received data to the user until the initial handshake is complete.
Also use TCPS_HAVEESTABLISHED() in a few other places.
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 25-Apr-1994 |
mycroft | As I described this on comp.protocols.tcp-ip:
I've found a problem with the TCP delayed ack algorithm. If the writer's
buffer becomes full before sending an entire window, the writer will stop
and the ack will be delayed and the transmission will be stalled pending
a timeout on (and transmission of) the delayed ack.
As an experiment, I've applied the following patch to my (NetBSD) kernel,
and it alleviates the problem.
The worst case for this change is that the writer sets the PSH bit on
every outgoing packet, in which case delayed ack is effectively disabled.
This is not an issue of correctness, however, and since most vendors use
the PSH bit a bit more intelligently, it doesn't seem like a serious
problem.
|
| 1.7 | 12-Apr-1994 |
mycroft | Patch from James Carlson to fix TCP stalls.
|
| 1.6 | 08-Jan-1994 |
mycroft | Remove some extra prototypes.
|
| 1.5 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.4 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.3 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.19.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.23.4.2 | 11-Dec-1996 |
mycroft | From trunk:
If we're in SYN-SENT or SYN-RECEIVED state, don't reset the keepalive
timer until we transition to ESTABLISHED state.
|
| 1.23.4.1 | 10-Dec-1996 |
mycroft | From trunk:
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.27.8.26 | 16-Jul-1997 |
thorpej | Rearrange things a bit so that TUBA compiles again.
|
| 1.27.8.25 | 13-Jul-1997 |
thorpej | Pay attention to tcp_do_rfc1323 when creating compressed state for a SYN.
|
| 1.27.8.24 | 11-Jul-1997 |
thorpej | If we have to abort the connection after the 3-way handshake is completed,
send an RST to the peer.
|
| 1.27.8.23 | 11-Jul-1997 |
thorpej | In the "received SYN for listen socket" case, correct an off-by-one
botch in the backlog test, which could cause our SYN to not be put in
compressed state.
XXX In a perfect world, we want to create compressed state for _all_ SYNs,
XXX even if the socket queue is full of legit connections, but doing so
XXX would break backlog semantics (because we'd have to drop the connection
XXX before the last ACK that completes the 3-way handshake, and we don't
XXX currently have a way to do that). Needs more thought.
|
| 1.27.8.22 | 06-Jul-1997 |
thorpej | Update from trunk.
|
| 1.27.8.21 | 30-Jun-1997 |
thorpej | Clean up the TODO list a bit.
|
| 1.27.8.20 | 30-Jun-1997 |
thorpej | Correct and document the semantics of error handling in syn_cache_get()
when creating the connection. syn_cache_get() now returns the following:
* NULL: We don't have a SYN for this ACK; send the peer an RST.
* -1: We are unable to create a socket for this connection. We have
sent an ACK,RST to the peer. Since the mbuf is being used to send
the response, caller should not free it.
* -2: Some other error occured while creating the connection, but we were
able to create a socket. In this case, caller should simply drop the
packet, and let the peer resend the ACK (this is the "abort and retry"
case).
* Else, return value is a pointer to the socket created for the new
connection.
|
| 1.27.8.19 | 29-Jun-1997 |
thorpej | In tcp_input():
- Perform the DIAGNOSTIC check for TCPS_LISTEN a bit earlier, and remove
a now-unnecessary check for TCPS_LISTEN state in the TCB.
|
| 1.27.8.18 | 29-Jun-1997 |
thorpej | In syn_cache_get():
- If we abort the connection (due to resource shortage or other error),
make sure notify tcp_input() upon return.
|
| 1.27.8.17 | 29-Jun-1997 |
thorpej | In tcp_input():
- Garbage-collect TCPS_LISTEN state from TCP input processing now that
all SYNs are handled with compressed state. Since we should never see
TCPS_LISTEN here, add a DIAGNOSTIC panic() if we encounter it.
In syn_cache_insert():
- Instrument hash collisions.
In syn_cache_get():
- Actually do all of the processing necessary to complete the connection
(not as far as the foreign host is concerned, but rather internal state
housekeeping). This fixes the last of the "weird" problems I have
encountered during normal use and under attack from multiple 10Mb/s
syn floods.
Also, add a few comments to clarify a few bits of code.
|
| 1.27.8.16 | 28-Jun-1997 |
thorpej | SYN cache can't ever be disabled since we use it for all incoming SYNs.
|
| 1.27.8.15 | 28-Jun-1997 |
thorpej | KNF.
|
| 1.27.8.14 | 28-Jun-1997 |
thorpej | In tcp_input(), always create compressed state when we receive a SYN,
rather than attempting to create a TCB, and using compressed state if
that fails. This makes the logic much simpler, removes some code
duplication, and kills the "temporary socket" hack that has historically
existed in SYN handling.
This change eliminates the need for SS_FORCE, since we only create
TCBs on legitimate connections. Previously, SS_FORCE was needed to
bypass the socket queue limit because legitimate connections might
have bee blocked out by bogus SYNs that weren't in compressed state.
Update TODO list appropriately. (This was a BIG one.)
|
| 1.27.8.13 | 28-Jun-1997 |
thorpej | Change some spaces -> tabs in indentation. (Hmm, looks like someone
uses _emacs_ :-)
|
| 1.27.8.12 | 28-Jun-1997 |
thorpej | In syn_cache_get(), set the state of the new inpcb to INP_BOUND.
If we don't do this, the PCB lookup will fail on subsequent packets
from the peer, causing an RST to be generated.
Couple of minor stylistic changes while I'm here.
|
| 1.27.8.11 | 26-Jun-1997 |
thorpej | KNF sweep over the syn cache code.
|
| 1.27.8.10 | 26-Jun-1997 |
thorpej | Remove unnecessary (and incorrect) htonl()s in the multicast test in
syn_cache_add().
|
| 1.27.8.9 | 26-Jun-1997 |
thorpej | Knock tcp_mss() lossage and type-size problems off the TODO list.
|
| 1.27.8.8 | 26-Jun-1997 |
thorpej | In sys_cache_respond():
- Fix type size problems, especially in creation of the timestamp option.
- Fix byte order problems in creation of the MSS option.
|
| 1.27.8.7 | 26-Jun-1997 |
thorpej | tcp_mss() needs to take a u_int, not a u_int16_t.
|
| 1.27.8.6 | 26-Jun-1997 |
thorpej | Casting pointers to u_int64_t isn't correct. Casting to u_long is fine
for arithmetic operations.
|
| 1.27.8.5 | 26-Jun-1997 |
thorpej | Oops, remove nested comment in the TODO list.
|
| 1.27.8.4 | 26-Jun-1997 |
thorpej | Add a TODO list, from Charles M. Hannum.
|
| 1.27.8.3 | 28-May-1997 |
mellon | Pointers are 64 bits on alpha - fix warning.
|
| 1.27.8.2 | 14-May-1997 |
mellon | SS_PRIV -> SS_FORCE
|
| 1.27.8.1 | 14-May-1997 |
mellon | Incorporate David Borman of BSDI's tcp SYN caching patches for
4.4BSD-lite2:
- define non-global syn cache variables
- define syn cache hashing algorithm
in tcp_input():
- package ts_val, ts_ecr and ts_present in a tcp_opt_info
structure so that they can be passed en masse to the
syn_cache code.
if the packet matched a socket that's in the ACCEPTING state:
- if an incoming connection does not yet have a tcpcb, but
it's not a SYN packet, check in the syn cache to see if we
cached the initial SYN. If not, send an RST packet. If
so, and if it's an RST packet, though, just blow away the
cache entry. If there was a cache entry and we aren't
processing an RST packet, create the full-blown connection
now and jump into the part of tcp_input() that deals with
connected sockets.
- if it is a SYN, and sonewconn() wouldn't queue it because
the limit for incoming half-up connections has been
exceeded, but the limit for established connections hasn't
yet been exceeded, then put this connection into the syn
cache.
after we've handled the accepting state:
- call tcp_dooptions with tcp_opt_info structure rather than
discrete option state variables.
- If the connection is half-up, and we get an ACK packet, but
it's not for the SYN we sent, drop the connection and send
an RST, per rfc793, p. 36.
in tcp_dooptions:
- combine all the option state passed as arguments into one
tcp_opt_info structure.
add syn cache management functions, verbatim from David's
patch:
syn_cache_insert: insert a connection into the SYN cache. If
we reach the per-bucket or cache size limit, toss the oldest
entry in the bucket, or if there are no entries in this
bucket yet, go looking for an entry to toss.
syn_cache_timer: blow away aging cache entries.
syn_cache_lookup: find the syn cache entry matching a
particular tcp packet, if any.
syn_cache_get: take an entry out of the cache and make a
socket for it.
syn_cache_reset: zap a connection in the syn cache based on
receipt of an RST packet.
syn_cache_unreach: zap a connection in the syn cache based on
an ICMP unreachable message.
syn_cache_add: given a LISTEN socket and an inbound SYN
request, add an entry to the syn cache and send a SYN,ACK to
the source.
syn_cache_respond: actually sends the SYN,ACK.
|
| 1.31.2.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.31.2.1 | 29-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.33.2.6 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.33.2.5 | 05-May-1998 |
mycroft | Pull up 1.45, per request of kml.
|
| 1.33.2.4 | 29-Jan-1998 |
mellon | Pull up 1.37-1.37 (thorpej). Pull up 1.40-1.41 and 1.43 (mellon) Pull up 1.42 (mycroft)
|
| 1.33.2.3 | 21-Nov-1997 |
thorpej | Pull up from trunk: slight change to previous: don't send RST in the
self-connect case.
|
| 1.33.2.2 | 21-Nov-1997 |
thorpej | Pull up from trunk: send RST on SYNs that come from themselves, and
if we receive SYN,ACK on a LISTEN socket.
|
| 1.33.2.1 | 08-Nov-1997 |
thorpej | Pull up from trunk: TCP MSS fixes to provide cleaner slow-start and recovery.
(kml)
|
| 1.77.2.3 | 03-May-1999 |
perry | branches: 1.77.2.3.2; 1.77.2.3.4;
pullup 1.80->1.81 (thorpej)
|
| 1.77.2.2 | 29-Apr-1999 |
perry | sync to 1.80 (thorpej)
|
| 1.77.2.1 | 09-Apr-1999 |
kml | Pullup of 1.78, which fixes the stack so that out of window SYNs are ACKed.
|
| 1.77.2.3.4.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.77.2.3.4.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.77.2.3.4.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.77.2.3.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.77.2.3.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.77.2.3.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.96.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.96.2.4 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.96.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.96.2.2 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.96.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.108.4.15 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #143)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.108.4.14 | 18-Oct-2002 |
itojun | sys/netinet/tcp_input.c 1.156 via patch
Correct log_refused check (TH_SYN, !TH_RST and !TH_ACK). Fixes PR#18669.
(itojun)
|
| 1.108.4.13 | 05-Sep-2002 |
itojun | sys/netinet/tcp_input.c 1.154
always consult SS_CANTRCVMORE. PR 18185
(itojun)
|
| 1.108.4.12 | 03-Apr-2002 |
he | Pull up revisions 1.138-1.140 (via patch, requested by itojun):
Reject TCP SYN packets sent to the broadcast address.
|
| 1.108.4.11 | 20-Mar-2002 |
he | Pull up revision 1.136 (requested by itojun):
Support tcp_log_refused for IPv6.
|
| 1.108.4.10 | 24-Jan-2002 |
he | Pull up revision 1.134 (requested by itojun):
Clean up the NRL copyright.
|
| 1.108.4.9 | 09-May-2001 |
he | Pull up revision 1.124 (requested by itojun):
Correct faith prefix determintaion.
|
| 1.108.4.8 | 06-Apr-2001 |
he | Pull up revision 1.122 (requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.108.4.7 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.108.4.6 | 28-Jul-2000 |
itojun | pullup 1.114 -> 1.115 (approved by releng-1-5)
> be proactive about unspecified IPv6 source address. pcb layer uses
> unspecified address (::) to mean "unbounded" or "unconnected",
> and can be confused by packets from outside.
>
> use of :: as source is not documented well in IPv6 specification.
>
> not sure if it presents a real threat. the worst case scenario is a DoS
> against TCP listening socket:
> - - outsider transmit TCP SYN with :: as IPv6 source
> - - receiving side creates TCP control block with:
> local address = my addres
> remote address = :: (meaning "unconnected")
> state = SYN_RCVD
> note that SYN ACK will not be sent due to ip6_output() filter.
> this stays until it timeouts.
> - - the TCP control block prevents listening TCP control block from
> being contacted (DoS).
>
> udp6/raw6 socket may have similar problem, but as they are connectionless,
> it may too much to filter it out.
|
| 1.108.4.5 | 23-Jul-2000 |
itojun | pullup 1.113 -> 1.114 (approved by releng-1-5)
add an DIAGNOSTIC case for MCLBYTES assumption
|
| 1.108.4.4 | 23-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
remove old mbuf assumption (ip header and tcp header are on the same mbuf).
this is for m_pulldown use. (sync with kame)
1.108 -> 1.109 syssrc/sys/netinet/tcp_input.c
1.56 -> 1.57 syssrc/sys/netinet/tcp_output.c
1.91 -> 1.92 syssrc/sys/netinet/tcp_subr.c
|
| 1.108.4.3 | 20-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
- add protection mechanism against ND cache corruption due to bad NUD hints.
this is part of:
sys/netinet/icmp6.h 1.9 -> 1.10
sys/netinet/tcp_input.c 1.111 -> 1.112
sys/netinet6/icmp6.c 1.34 -> 1.35
sys/netinet6/nd6.c 1.30 -> 1.31
sys/netinet6/nd6.h 1.14 -> 1.15
|
| 1.108.4.2 | 16-Jul-2000 |
itojun | pullup 1.112 -> 1.113 (approved by releng-1-5)
date: 2000/07/09 12:49:08; author: itojun; state: Exp; lines: +4 -2
be more cautious about tcp option length field. drop bogus ones earlier.
not sure if there is a real threat or not, but it seems that there's
possibility for overrun/underrun (like non-NOP option with optlen > cnt).
|
| 1.108.4.1 | 03-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
repair kernel faithd(8) support. there were two mistakes:
(1) tcp6_input dropped packets for translation
(2) in6_pcblookup_connect was too strict
|
| 1.122.2.16 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.122.2.15 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.122.2.14 | 22-Oct-2002 |
thorpej | Sync with HEAD.
|
| 1.122.2.13 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.122.2.12 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.122.2.11 | 28-Aug-2002 |
thorpej | Sync with -current.
|
| 1.122.2.10 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.122.2.9 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.122.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.122.2.7 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.122.2.6 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.122.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.122.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.122.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.122.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.122.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.127.4.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.127.2.7 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.127.2.6 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.127.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.127.2.4 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.127.2.3 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.127.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.127.2.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.131.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.141.4.8 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #1680)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.141.4.7 | 22-Oct-2003 |
jmc | Pullup rev 1.173 (requested by he in ticket #1530)
Introduce a new INVOKING status for callouts, and use it to close
a race condition in the TCP code. Fixes PR#20390.
|
| 1.141.4.6 | 17-Jun-2003 |
msaitoh | Pull up revisions 1.168 (requested by itojun in ticket #1304):
inherit IPV6_V6ONLY from listening socket. Fixes PR#21713.
|
| 1.141.4.5 | 23-Oct-2002 |
lukem | Pull up revision 1.158 (requested by thorpej in ticket #938):
Make sure TF_REQ_TSTMP and TF_REQ_SCALE get set correctly in the new
TCPCB in the passive-open case.
Fixes PR 18677.
|
| 1.141.4.4 | 21-Oct-2002 |
lukem | Pull up revision 1.156 via patch (requested by itojun in ticket #915):
correct log_refused check (TH_SYN, !TH_RST and !TH_ACK). PR 18669
|
| 1.141.4.3 | 06-Sep-2002 |
lukem | Pull up revision 1.154 via patch (requested by itojun in ticket #775):
always consult SS_CANTRCVMORE. PR 18185
|
| 1.141.4.2 | 28-Aug-2002 |
lukem | Pull up revision 1.153 (requested by thorpej in ticket #738):
Fix a problem introduced in rev 1.103, where we recycle a TIME_WAIT
TCPCB .. the fields need to be converted back to net-order, because
the packet is checksummed after the TCPCB lookup happens.
From YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>.
|
| 1.141.4.1 | 21-Jul-2002 |
lukem | Pull up revision 1.149 (requested by wrstuden in ticket #517):
When a new connection arrives on a listening port, copy over the
value of the TCP_NODELAY socket option from the listener to the
newly connected connection. Agrees with how Linux & FreeBSD behave,
and goes more with the spirit of accept(2) creating a socket with
the same properties as the listener.
Analysis by Kevin Lahey. Closes PR 17616 by myself.
|
| 1.141.2.5 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.141.2.4 | 20-Jul-2002 |
gehenna | catch up with -current.
|
| 1.141.2.3 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.141.2.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.141.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.171.2.11 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.171.2.10 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.171.2.9 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.171.2.8 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.171.2.7 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.171.2.6 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.171.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.171.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.171.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.171.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.171.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.190.2.8 | 22-Apr-2005 |
tron | Apply patch (requested by christos in ticket #1445):
Fix TCP performance problems introduced in ticket 1401.
|
| 1.190.2.7 | 06-Apr-2005 |
tron | Apply patch (requested by christos in ticket #1401):
If an echoed RFC 1323 time stamp appears to be later than the current time,
ignore it and fall back to old-style RTT calculation. This prevents ending
up with a negative RTT and panicking later.
|
| 1.190.2.6 | 19-Sep-2004 |
he | branches: 1.190.2.6.2;
Apply patch (requested by yamt in ticket #861):
Fix this so it compiles again; we cannot use the link
set macros for pool initialization on this release branch.
|
| 1.190.2.5 | 18-Sep-2004 |
he | Pull up revision 1.209 (requested by yamt in ticket #861):
Fix ipqent pool corruption problems. Make the TCP reassembly
code use its own pool of ipqent rather than sharing it with
the IP reassembly code. Fixes PR#24782.
|
| 1.190.2.4 | 10-May-2004 |
tron | Pull up revision 1.205 (requested by jonathan in ticket #280):
Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.190.2.3 | 29-Apr-2004 |
jmc | Pullup rev 1.204 (requested by matt in ticket #213)
When a packet is received that overlaps the left side of the window,
check for RST *before* trimming data and adjust its sequence number.
|
| 1.190.2.2 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #169)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.190.2.1 | 16-Apr-2004 |
tron | Pull up revision 1.192 (requested by ragge in ticket #140):
Add back one line which was accidentially removed (by me) a while ago.
Spotted by Markus Friedl (markus at openbsd.org).
|
| 1.190.2.6.2.2 | 22-Apr-2005 |
tron | Apply patch (requested by christos in ticket #1445):
Fix TCP performance problems introduced in ticket 1401.
|
| 1.190.2.6.2.1 | 06-Apr-2005 |
tron | Apply patch (requested by christos in ticket #1401):
If an echoed RFC 1323 time stamp appears to be later than the current time,
ignore it and fall back to old-style RTT calculation. This prevents ending
up with a negative RTT and panicking later.
|
| 1.212.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.212.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.212.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.224.2.3 | 26-Oct-2006 |
ghen | Pull up following revision(s) (requested by rpaulo in ticket #1547):
sys/netinet/tcp_input.c: revision 1.250
PR 34776: don't accept TCP connections to broadcast addresses.
Move the multicast/broadcast check above (before creating a
syn_cache entry)
By Yasuoka Yasuoka.
|
| 1.224.2.2 | 28-Apr-2005 |
tron | branches: 1.224.2.2.2;
Pull up revision 1.227 (requested by manu in ticket #206):
Fix build problem after recent NAT-T changes
|
| 1.224.2.1 | 04-Apr-2005 |
tron | Pull up revision 1.225 (requested by yamt in ticket #90):
protect tcpipqent with splvm.
|
| 1.224.2.2.2.1 | 26-Oct-2006 |
ghen | Pull up following revision(s) (requested by rpaulo in ticket #1547):
sys/netinet/tcp_input.c: revision 1.250
PR 34776: don't accept TCP connections to broadcast addresses.
Move the multicast/broadcast check above (before creating a
syn_cache entry)
By Yasuoka Yasuoka.
|
| 1.230.2.10 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.230.2.9 | 27-Feb-2008 |
yamt | sync with head.
|
| 1.230.2.8 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.230.2.7 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.230.2.6 | 21-Jan-2008 |
yamt | sync with head
|
| 1.230.2.5 | 15-Nov-2007 |
yamt | sync with head.
|
| 1.230.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.230.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.230.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.230.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.236.6.1 | 22-Nov-2005 |
yamt | sync with head.
|
| 1.237.4.3 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.237.4.2 | 05-Feb-2006 |
rpaulo | Adapt to in6pcb -> inpcb changes.
|
| 1.237.4.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.237.2.2 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.237.2.1 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.238.2.2 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.238.2.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time(), use bintime's for timeout
calculations and use "time_second" instead of "time.tv_sec".
|
| 1.239.6.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.239.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.239.2.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.239.2.2 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.239.2.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.240.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.244.4.3 | 18-Dec-2006 |
yamt | sync with head.
|
| 1.244.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.244.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.244.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.244.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.255.4.1 | 04-Jun-2007 |
wrstuden | Update to today's netbsd-4.
|
| 1.255.2.2 | 25-May-2007 |
pavel | Pull up following revision(s) (requested by riz in ticket #670):
sys/netinet/tcp_input.c: revision 1.265
sys/netinet/tcp_input.c: revision 1.266
sys/arch/i386/conf/ALL: revision 1.98
Revert a small part of revision 1.254 - remove const qualifier from
the struct tcphdr * argument of tcp_dooptions(). RFC2385 support
(options TCP_SIGNATURE) needs to modify the header during options
processing, and this revision broke it.
OK yamt@.
Fix compilation in the TCP_SIGNATURE case:
- don't use void * for pointer arithmetic
- don't try to modify const parameters
A kernel with 'options TCP_SIGNATURE' works as well as it ever did, now.
(ie, clunky, but passable)
Add 'options TCP_SIGNATURE' to hopefully keep this code from
invisibly breaking periodically, as it's done a couple times.
|
| 1.255.2.1 | 24-May-2007 |
pavel | Pull up following revision(s) (requested by degroote in ticket #667):
sys/netinet/tcp_input.c: revision 1.260
sys/netinet/tcp_output.c: revision 1.154
sys/netinet/tcp_subr.c: revision 1.210
sys/netinet6/icmp6.c: revision 1.129
sys/netinet6/in6_proto.c: revision 1.70
sys/netinet6/ip6_forward.c: revision 1.54
sys/netinet6/ip6_input.c: revision 1.94
sys/netinet6/ip6_output.c: revision 1.114
sys/netinet6/raw_ip6.c: revision 1.81
sys/netipsec/ipcomp_var.h: revision 1.4
sys/netipsec/ipsec.c: revision 1.26 via patch,1.31-1.32
sys/netipsec/ipsec6.h: revision 1.5
sys/netipsec/ipsec_input.c: revision 1.14
sys/netipsec/ipsec_netbsd.c: revision 1.18,1.26
sys/netipsec/ipsec_output.c: revision 1.21 via patch
sys/netipsec/key.c: revision 1.33,1.44
sys/netipsec/xform_ipcomp.c: revision 1.9
sys/netipsec/xform_ipip.c: revision 1.15
sys/opencrypto/deflate.c: revision 1.8
Commit my SoC work
Add ipv6 support for fast_ipsec
Note that currently, packet with extensions headers are not correctly
supported
Change the ipcomp logic
Add sysctl tree to modify the fast_ipsec options related to ipv6. Similar
to the sysctl kame interface.
Choose the good default policy, depending of the adress family of the
desired policy
Increase the refcount for the default ipv6 policy so nobody can reclaim it
Always compute the sp index even if we don't have any sp in spd. It will
let us to choose the right default policy (based on the adress family
requested).
While here, fix an error message
Use dynamic array instead of an static array to decompress. It lets us to
decompress any data, whatever is the radio decompressed data / compressed
data.
It fixes the last issues with fast_ipsec and ipcomp.
While here, bzero -> memset, bcopy -> memcpy, FREE -> free
Reviewed a long time ago by sam@
|
| 1.260.2.4 | 07-May-2007 |
yamt | sync with head.
|
| 1.260.2.3 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.260.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.260.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.262.2.5 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.262.2.4 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.262.2.3 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.262.2.2 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.262.2.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.263.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.268.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.270.10.2 | 02-Aug-2007 |
yamt | our tcp timestamps are in PR_SLOWHZ, not HZ.
|
| 1.270.10.1 | 02-Aug-2007 |
yamt | file tcp_input.c was added on branch matt-mips64 on 2007-08-02 13:06:31 +0000
|
| 1.270.8.1 | 13-Nov-2007 |
bouyer | Sync with HEAD
|
| 1.270.4.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.270.4.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.270.4.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.270.2.2 | 11-Nov-2007 |
joerg | Sync with HEAD.
|
| 1.270.2.1 | 04-Nov-2007 |
jmcneill | Sync with HEAD.
|
| 1.271.2.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.271.2.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.271.2.1 | 19-Nov-2007 |
mjf | Sync with HEAD.
|
| 1.272.6.2 | 19-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.272.6.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.272.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.280.6.3 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.280.6.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.280.6.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.280.2.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.284.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.286.2.7 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.286.2.6 | 11-Mar-2010 |
yamt | sync with head
|
| 1.286.2.5 | 16-Sep-2009 |
yamt | sync with head
|
| 1.286.2.4 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.286.2.3 | 18-Jul-2009 |
yamt | sync with head.
|
| 1.286.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.286.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.287.4.3 | 31-Jul-2008 |
simonb | Sync with head.
|
| 1.287.4.2 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.287.4.1 | 03-Jul-2008 |
simonb | Sync with head.
|
| 1.287.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.289.2.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.291.8.1 | 18-Jul-2009 |
snj | branches: 1.291.8.1.2;
Pull up following revision(s) (requested by is in ticket #859):
sys/netinet/tcp_input.c: revision 1.296
Follow exactly the recommendation of draft-ietf-tcpm-tcpsecure-11.txt:
Don't check gainst the last ack received, but the expected sequence number.
This makes RST handling independent of delayed ACK. From Joanne M Mikkelson.
|
| 1.291.8.1.2.2 | 20-May-2011 |
matt | bring matt-nb5-mips64 up to date with netbsd-5-1-RELEASE (except compat).
|
| 1.291.8.1.2.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.291.4.6 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1973):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.291.4.5 | 11-Jun-2010 |
riz | branches: 1.291.4.5.2; 1.291.4.5.6;
Pull up following revision(s) (requested by bouyer in ticket #1382):
sys/netinet/tcp_input.c: revision 1.304
sys/netinet/tcp_input.c: revision 1.305
syn_cache_put(): defer all pool_put() to the callout. Reschedule
the callout if needed so frees are not delayed too much.
syn_cache_timer(): we can't call syn_cache_put() here any more,
so move code deleted from syn_cache_put() here.
Avoid KASSERT() in kern_timeout.c because pool_put() is called from
ipintr context, as reported in
http://mail-index.netbsd.org/tech-kern/2010/03/19/msg007762.html
Thanks to Andrew Doran and Mindaugas Rasiukevicius for help and review.
Make sure SYN_CACHE_TIMER_ARM() has been run before calling syn_cache_put()
as it will reschedule the timer. Fixes PR kern/43318.
|
| 1.291.4.4 | 20-May-2010 |
snj | Revert ticket 1382.
|
| 1.291.4.3 | 20-May-2010 |
snj | Pull up following revision(s) (requested by bouyer in ticket #1382):
sys/netinet/tcp_input.c: revision 1.304
syn_cache_put(): defer all pool_put() to the callout. Reschedule
the callout if needed so frees are not delayed too much.
syn_cache_timer(): we can't call syn_cache_put() here any more,
so move code deleted from syn_cache_put() here.
Avoid KASSERT() in kern_timeout.c because pool_put() is called from
ipintr context, as reported in
http://mail-index.netbsd.org/tech-kern/2010/03/19/msg007762.html
Thanks to Andrew Doran and Mindaugas Rasiukevicius for help and review.
|
| 1.291.4.2 | 26-Sep-2009 |
snj | Pull up following revision(s) (requested by darran in ticket #950):
sys/netinet/tcp_input.c: revision 1.299
sys/netinet/tcp_usrreq.c: revision 1.156
sys/netinet/tcp_var.h: revision 1.161
Make tcp msl (max segment life) tunable via sysctl net.inet.tcp.msl.
Okayed by tls@.
|
| 1.291.4.1 | 18-Jul-2009 |
snj | Pull up following revision(s) (requested by is in ticket #859):
sys/netinet/tcp_input.c: revision 1.296
Follow exactly the recommendation of draft-ietf-tcpm-tcpsecure-11.txt:
Don't check gainst the last ack received, but the expected sequence number.
This makes RST handling independent of delayed ACK. From Joanne M Mikkelson.
|
| 1.291.4.5.6.1 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1973):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.291.4.5.2.1 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1973):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.291.2.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.291.2.1 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.292.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.292.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.300.4.4 | 31-May-2011 |
rmind | sync with head
|
| 1.300.4.3 | 21-Apr-2011 |
rmind | sync with head
|
| 1.300.4.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.300.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.300.2.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.300.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.306.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.317.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.317.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.318.2.3 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.318.2.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.318.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.321.8.1 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1315):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.321.6.1 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1315):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.321.2.1 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1315):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.325.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.325.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.325.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.327.2.3 | 18-May-2014 |
rmind | sync with head
|
| 1.327.2.2 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.327.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.331.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.334.4.8 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.334.4.7 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.334.4.6 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.334.4.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.334.4.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.334.4.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.334.4.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.334.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.334.2.2 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #886):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.334.2.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by he in ticket #530):
sys/netinet/tcp_output.c: revision 1.180
sys/netinet/tcp_input.c: revision 1.336
sys/netinet/tcp_usrreq.c: revision 1.203
share/man/man4/tcp.4: revision 1.30
sys/netinet/tcp.h: revision 1.31
sys/netinet/tcp_subr.c: revision 1.258
sys/netinet/tcp_var.h: revision 1.176
sys/netinet/tcp_var.h: revision 1.177
sys/sys/param.h: bump revision
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.347.2.3 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.347.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.347.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.353.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.357.4.7 | 13-Sep-2020 |
martin | Pull up following revision(s) (requested by kardel in ticket #1604):
sys/netinet/tcp_input.c: revision 1.420
PR/kern 55567
fix the data-only fast path. RCV.UP and SND.WL1 could be left behind
on long sequences of data only packets. pull them along to avoid relative
sequence wraps.
consistent with FreeBSD
addresses second failure mode of PR/kern 55567.
pullup to netbsd-8
pullup to netbsd-9
|
| 1.357.4.6 | 03-Sep-2020 |
martin | Pull up following revision(s) (requested by kardel in ticket #1602):
sys/netinet/tcp_input.c: revision 1.419
Fix fast path for uni directional transfers
pure ACK case:
drag snd_wl2 along so only newer
ACKs can update the window size.
also avoids the state where snd_wl2
is eventually larger than th_ack and thus
blocking the window update mechanism and
the connection gets stuck for a loooong
time in the zero sized send window state.
see PR/kern 55567
ok thorpej@, also found in FreeBSD
|
| 1.357.4.5 | 08-Jul-2020 |
martin | Apply patch, requested by christos in ticket #1566:
Deduplicate more code and avoid use of uninitialized variables.
|
| 1.357.4.4 | 07-Jul-2020 |
martin | Pull up following revision(s) (requested by christos in ticket #1566):
sys/netinet/tcp_input.c: revision 1.418 (via patch)
- always set both ip and ip6, otherwise a kernel assertion can be triggered
- move alignment early so that we do less work
|
| 1.357.4.3 | 30-Mar-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #662):
sys/netinet/tcp_input.c: revision 1.383 (via patch)
Revert rev1.183 (2003).
It was intended as an optimization, but it increases the attack surface:
the IPsec policy is not enforced on RST packets when the socket is in the
LISTEN state, and an (unauthenticated) attacker could jam the connection
between two IPsec hosts by sending RST packets between the client's SYN
and ACK packets.
Discussed with ozaki-r@.
|
| 1.357.4.2 | 21-Oct-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #300):
crypto/dist/ipsec-tools/src/setkey/parse.y: 1.19
crypto/dist/ipsec-tools/src/setkey/token.l: 1.20
distrib/sets/lists/tests/mi: 1.754, 1.757, 1.759
doc/TODO.smpnet: 1.12-1.13
sys/net/pfkeyv2.h: 1.32
sys/net/raw_cb.c: 1.23-1.24, 1.28
sys/net/raw_cb.h: 1.28
sys/net/raw_usrreq.c: 1.57-1.58
sys/net/rtsock.c: 1.228-1.229
sys/netinet/in_proto.c: 1.125
sys/netinet/ip_input.c: 1.359-1.361
sys/netinet/tcp_input.c: 1.359-1.360
sys/netinet/tcp_output.c: 1.197
sys/netinet/tcp_var.h: 1.178
sys/netinet6/icmp6.c: 1.213
sys/netinet6/in6_proto.c: 1.119
sys/netinet6/ip6_forward.c: 1.88
sys/netinet6/ip6_input.c: 1.181-1.182
sys/netinet6/ip6_output.c: 1.193
sys/netinet6/ip6protosw.h: 1.26
sys/netipsec/ipsec.c: 1.100-1.122
sys/netipsec/ipsec.h: 1.51-1.61
sys/netipsec/ipsec6.h: 1.18-1.20
sys/netipsec/ipsec_input.c: 1.44-1.51
sys/netipsec/ipsec_netbsd.c: 1.41-1.45
sys/netipsec/ipsec_output.c: 1.49-1.64
sys/netipsec/ipsec_private.h: 1.5
sys/netipsec/key.c: 1.164-1.234
sys/netipsec/key.h: 1.20-1.32
sys/netipsec/key_debug.c: 1.18-1.21
sys/netipsec/key_debug.h: 1.9
sys/netipsec/keydb.h: 1.16-1.20
sys/netipsec/keysock.c: 1.59-1.62
sys/netipsec/keysock.h: 1.10
sys/netipsec/xform.h: 1.9-1.12
sys/netipsec/xform_ah.c: 1.55-1.74
sys/netipsec/xform_esp.c: 1.56-1.72
sys/netipsec/xform_ipcomp.c: 1.39-1.53
sys/netipsec/xform_ipip.c: 1.50-1.54
sys/netipsec/xform_tcp.c: 1.12-1.16
sys/rump/librump/rumpkern/Makefile.rumpkern: 1.170
sys/rump/librump/rumpnet/net_stub.c: 1.27
sys/sys/protosw.h: 1.67-1.68
tests/net/carp/t_basic.sh: 1.7
tests/net/if_gif/t_gif.sh: 1.11
tests/net/if_l2tp/t_l2tp.sh: 1.3
tests/net/ipsec/Makefile: 1.7-1.9
tests/net/ipsec/algorithms.sh: 1.5
tests/net/ipsec/common.sh: 1.4-1.6
tests/net/ipsec/t_ipsec_ah_keys.sh: 1.2
tests/net/ipsec/t_ipsec_esp_keys.sh: 1.2
tests/net/ipsec/t_ipsec_gif.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_l2tp.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_misc.sh: 1.8-1.18
tests/net/ipsec/t_ipsec_sockopt.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tcp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_transport.sh: 1.5-1.6
tests/net/ipsec/t_ipsec_tunnel.sh: 1.9
tests/net/ipsec/t_ipsec_tunnel_ipcomp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tunnel_odd.sh: 1.3
tests/net/mcast/t_mcast.sh: 1.6
tests/net/net/t_ipaddress.sh: 1.11
tests/net/net_common.sh: 1.20
tests/net/npf/t_npf.sh: 1.3
tests/net/route/t_flags.sh: 1.20
tests/net/route/t_flags6.sh: 1.16
usr.bin/netstat/fast_ipsec.c: 1.22
Do m_pullup before mtod
It may fix panicks of some tests on anita/sparc and anita/GuruPlug.
---
KNF
---
Enable DEBUG for babylon5
---
Apply C99-style struct initialization to xformsw
---
Tweak outputs of netstat -s for IPsec
- Get rid of "Fast"
- Use ipsec and ipsec6 for titles to clarify protocol
- Indent outputs of sub protocols
Original outputs were organized like this:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
New outputs are organized like this:
ipsec:
ah:
esp:
ipip:
ipcomp:
ipsec6:
ah:
esp:
ipip:
ipcomp:
---
Add test cases for IPComp
---
Simplify IPSEC_OSTAT macro (NFC)
---
KNF; replace leading whitespaces with hard tabs
---
Introduce and use SADB_SASTATE_USABLE_P
---
KNF
---
Add update command for testing
Updating an SA (SADB_UPDATE) requires that a process issuing
SADB_UPDATE is the same as a process issued SADB_ADD (or SADB_GETSPI).
This means that update command must be used with add command in a
configuration of setkey. This usage is normally meaningless but
useful for testing (and debugging) purposes.
---
Add test cases for updating SA/SP
The tests require newly-added udpate command of setkey.
---
PR/52346: Frank Kardel: Fix checksumming for NAT-T
See XXX for improvements.
---
Remove codes for PACKET_TAG_IPSEC_IN_CRYPTO_DONE
It seems that PACKET_TAG_IPSEC_IN_CRYPTO_DONE is for network adapters
that have IPsec accelerators; a driver sets the mtag to a packet
when its device has already encrypted the packet.
Unfortunately no driver implements such offload features for long
years and seems unlikely to implement them soon. (Note that neither
FreeBSD nor Linux doesn't have such drivers.) Let's remove related
(unused) codes and simplify the IPsec code.
---
Fix usages of sadb_msg_errno
---
Avoid updating sav directly
On SADB_UPDATE a target sav was updated directly, which was unsafe.
Instead allocate another sav, copy variables of the old sav to
the new one and replace the old one with the new one.
---
Simplify; we can assume sav->tdb_xform cannot be NULL while it's valid
---
Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
---
Use explicit_memset to surely zero-clear key_auth and key_enc
---
Make sure to clear keys on error paths of key_setsaval
---
Add missing KEY_FREESAV
---
Make sure a sav is inserted to a sah list after its initialization completes
---
Remove unnecessary zero-clearing codes from key_setsaval
key_setsaval is now used only for a newly-allocated sav. (It was
used to reset variables of an existing sav.)
---
Correct wrong assumption of sav->refcnt in key_delsah
A sav in a list is basically not to be sav->refcnt == 0. And also
KEY_FREESAV assumes sav->refcnt > 0.
---
Let key_getsavbyspi take a reference of a returning sav
---
Use time_mono_to_wall (NFC)
---
Separate sending message routine (NFC)
---
Simplify; remove unnecessary zero-clears
key_freesaval is used only when a target sav is being destroyed.
---
Omit NULL checks for sav->lft_c
sav->lft_c can be NULL only when initializing or destroying sav.
---
Omit unnecessary NULL checks for sav->sah
---
Omit unnecessary check of sav->state
key_allocsa_policy picks a sav of either MATURE or DYING so we
don't need to check its state again.
---
Simplify; omit unnecessary saidx passing
- ipsec_nextisr returns a saidx but no caller uses it
- key_checkrequest is passed a saidx but it can be gotton by
another argument (isr)
---
Fix splx isn't called on some error paths
---
Fix header size calculation of esp where sav is NULL
---
Fix header size calculation of ah in the case sav is NULL
This fix was also needed for esp.
---
Pass sav directly to opencrypto callback
In a callback, use a passed sav as-is by default and look up a sav
only if the passed sav is dead.
---
Avoid examining freshness of sav on packet processing
If a sav list is sorted (by lft_c->sadb_lifetime_addtime) in advance,
we don't need to examine each sav and also don't need to delete one
on the fly and send up a message. Fortunately every sav lists are sorted
as we need.
Added key_validate_savlist validates that each sav list is surely sorted
(run only if DEBUG because it's not cheap).
---
Add test cases for SAs with different SPIs
---
Prepare to stop using isr->sav
isr is a shared resource and using isr->sav as a temporal storage
for each packet processing is racy. And also having a reference from
isr to sav makes the lifetime of sav non-deterministic; such a reference
is removed when a packet is processed and isr->sav is overwritten by
new one. Let's have a sav locally for each packet processing instead of
using shared isr->sav.
However this change doesn't stop using isr->sav yet because there are
some users of isr->sav. isr->sav will be removed after the users find
a way to not use isr->sav.
---
Fix wrong argument handling
---
fix printf format.
---
Don't validate sav lists of LARVAL or DEAD states
We don't sort the lists so the validation will always fail.
Fix PR kern/52405
---
Make sure to sort the list when changing the state by key_sa_chgstate
---
Rename key_allocsa_policy to key_lookup_sa_bysaidx
---
Separate test files
---
Calculate ah_max_authsize on initialization as well as esp_max_ivlen
---
Remove m_tag_find(PACKET_TAG_IPSEC_PENDING_TDB) because nobody sets the tag
---
Restore a comment removed in previous
The comment is valid for the below code.
---
Make tests more stable
sleep command seems to wait longer than expected on anita so
use polling to wait for a state change.
---
Add tests that explicitly delete SAs instead of waiting for expirations
---
Remove invalid M_AUTHIPDGM check on ESP isr->sav
M_AUTHIPDGM flag is set to a mbuf in ah_input_cb. An sav of ESP can
have AH authentication as sav->tdb_authalgxform. However, in that
case esp_input and esp_input_cb are used to do ESP decryption and
AH authentication and M_AUTHIPDGM never be set to a mbuf. So
checking M_AUTHIPDGM of a mbuf on isr->sav of ESP is meaningless.
---
Look up sav instead of relying on unstable sp->req->sav
This code is executed only in an error path so an additional lookup
doesn't matter.
---
Correct a comment
---
Don't release sav if calling crypto_dispatch again
---
Remove extra KEY_FREESAV from ipsec_process_done
It should be done by the caller.
---
Don't bother the case of crp->crp_buf == NULL in callbacks
---
Hold a reference to an SP during opencrypto processing
An SP has a list of isr (ipsecrequest) that represents a sequence
of IPsec encryption/authentication processing. One isr corresponds
to one opencrypto processing. The lifetime of an isr follows its SP.
We pass an isr to a callback function of opencrypto to continue
to a next encryption/authentication processing. However nobody
guaranteed that the isr wasn't freed, i.e., its SP wasn't destroyed.
In order to avoid such unexpected destruction of isr, hold a reference
to its SP during opencrypto processing.
---
Don't make SAs expired on tests that delete SAs explicitly
---
Fix a debug message
---
Dedup error paths (NFC)
---
Use pool to allocate tdb_crypto
For ESP and AH, we need to allocate an extra variable space in addition
to struct tdb_crypto. The fixed size of pool items may be larger than
an actual requisite size of a buffer, but still the performance
improvement by replacing malloc with pool wins.
---
Don't use unstable isr->sav for header size calculations
We may need to optimize to not look up sav here for users that
don't need to know an exact size of headers (e.g., TCP segmemt size
caclulation).
---
Don't use sp->req->sav when handling NAT-T ESP fragmentation
In order to do this we need to look up a sav however an additional
look-up degrades performance. A sav is later looked up in
ipsec4_process_packet so delay the fragmentation check until then
to avoid an extra look-up.
---
Don't use key_lookup_sp that depends on unstable sp->req->sav
It provided a fast look-up of SP. We will provide an alternative
method in the future (after basic MP-ification finishes).
---
Stop setting isr->sav on looking up sav in key_checkrequest
---
Remove ipsecrequest#sav
---
Stop setting mtag of PACKET_TAG_IPSEC_IN_DONE because there is no users anymore
---
Skip ipsec_spi_*_*_preferred_new_timeout when running on qemu
Probably due to PR 43997
---
Add localcount to rump kernels
---
Remove unused macro
---
Fix key_getcomb_setlifetime
The fix adjusts a soft limit to be 80% of a corresponding hard limit.
I'm not sure the fix is really correct though, at least the original
code is wrong. A passed comb is zero-cleared before calling
key_getcomb_setlifetime, so
comb->sadb_comb_soft_addtime = comb->sadb_comb_soft_addtime * 80 / 100;
is meaningless.
---
Provide and apply key_sp_refcnt (NFC)
It simplifies further changes.
---
Fix indentation
Pointed out by knakahara@
---
Use pslist(9) for sptree
---
Don't acquire global locks for IPsec if NET_MPSAFE
Note that the change is just to make testing easy and IPsec isn't MP-safe yet.
---
Let PF_KEY socks hold their own lock instead of softnet_lock
Operations on SAD and SPD are executed via PF_KEY socks. The operations
include deletions of SAs and SPs that will use synchronization mechanisms
such as pserialize_perform to wait for references to SAs and SPs to be
released. It is known that using such mechanisms with holding softnet_lock
causes a dead lock. We should avoid the situation.
---
Make IPsec SPD MP-safe
We use localcount(9), not psref(9), to make the sptree and secpolicy (SP)
entries MP-safe because SPs need to be referenced over opencrypto
processing that executes a callback in a different context.
SPs on sockets aren't managed by the sptree and can be destroyed in softint.
localcount_drain cannot be used in softint so we delay the destruction of
such SPs to a thread context. To do so, a list to manage such SPs is added
(key_socksplist) and key_timehandler_spd deletes dead SPs in the list.
For more details please read the locking notes in key.c.
Proposed on tech-kern@ and tech-net@
---
Fix updating ipsec_used
- key_update_used wasn't called in key_api_spddelete2 and key_api_spdflush
- key_update_used wasn't called if an SP had been added/deleted but
a reply to userland failed
---
Fix updating ipsec_used; turn on when SPs on sockets are added
---
Add missing IPsec policy checks to icmp6_rip6_input
icmp6_rip6_input is quite similar to rip6_input and the same checks exist
in rip6_input.
---
Add test cases for setsockopt(IP_IPSEC_POLICY)
---
Don't use KEY_NEWSP for dummy SP entries
By the change KEY_NEWSP is now not called from softint anymore
and we can use kmem_zalloc with KM_SLEEP for KEY_NEWSP.
---
Comment out unused functions
---
Add test cases that there are SPs but no relevant SAs
---
Don't allow sav->lft_c to be NULL
lft_c of an sav that was created by SADB_GETSPI could be NULL.
---
Clean up clunky eval strings
- Remove unnecessary \ at EOL
- This allows to omit ; too
- Remove unnecessary quotes for arguments of atf_set
- Don't expand $DEBUG in eval
- We expect it's expanded on execution
Suggested by kre@
---
Remove unnecessary KEY_FREESAV in an error path
sav should be freed (unreferenced) by the caller.
---
Use pslist(9) for sahtree
---
Use pslist(9) for sah->savtree
---
Rename local variable newsah to sah
It may not be new.
---
MP-ify SAD slightly
- Introduce key_sa_mtx and use it for some list operations
- Use pserialize for some list iterations
---
Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
---
Remove out-of-date log output
Pointed out by riastradh@
---
Use KDASSERT instead of KASSERT for mutex_ownable
Because mutex_ownable is too heavy to run in a fast path
even for DIAGNOSTIC + LOCKDEBUG.
Suggested by riastradh@
---
Assemble global lists and related locks into cache lines (NFCI)
Also rename variable names from *tree to *list because they are
just lists, not trees.
Suggested by riastradh@
---
Move locking notes
---
Update the locking notes
- Add locking order
- Add locking notes for misc lists such as reglist
- Mention pserialize, key_sp_ref and key_sp_unref on SP operations
Requested by riastradh@
---
Describe constraints of key_sp_ref and key_sp_unref
Requested by riastradh@
---
Hold key_sad.lock on SAVLIST_WRITER_INSERT_TAIL
---
Add __read_mostly to key_psz
Suggested by riastradh@
---
Tweak wording (pserialize critical section => pserialize read section)
Suggested by riastradh@
---
Add missing mutex_exit
---
Fix setkey -D -P outputs
The outputs were tweaked (by me), but I forgot updating libipsec
in my local ATF environment...
---
MP-ify SAD (key_sad.sahlist and sah entries)
localcount(9) is used to protect key_sad.sahlist and sah entries
as well as SPD (and will be used for SAD sav).
Please read the locking notes of SAD for more details.
---
Introduce key_sa_refcnt and replace sav->refcnt with it (NFC)
---
Destroy sav only in the loop for DEAD sav
---
Fix KASSERT(solocked(sb->sb_so)) failure in sbappendaddr that is called eventually from key_sendup_mbuf
If key_sendup_mbuf isn't passed a socket, the assertion fails.
Originally in this case sb->sb_so was softnet_lock and callers
held softnet_lock so the assertion was magically satisfied.
Now sb->sb_so is key_so_mtx and also softnet_lock isn't always
held by callers so the assertion can fail.
Fix it by holding key_so_mtx if key_sendup_mbuf isn't passed a socket.
Reported by knakahara@
Tested by knakahara@ and ozaki-r@
---
Fix locking notes of SAD
---
Fix deadlock between key_sendup_mbuf called from key_acquire and localcount_drain
If we call key_sendup_mbuf from key_acquire that is called on packet
processing, a deadlock can happen like this:
- At key_acquire, a reference to an SP (and an SA) is held
- key_sendup_mbuf will try to take key_so_mtx
- Some other thread may try to localcount_drain to the SP with
holding key_so_mtx in say key_api_spdflush
- In this case localcount_drain never return because key_sendup_mbuf
that has stuck on key_so_mtx never release a reference to the SP
Fix the deadlock by deferring key_sendup_mbuf to the timer
(key_timehandler).
---
Fix that prev isn't cleared on retry
---
Limit the number of mbufs queued for deferred key_sendup_mbuf
It's easy to be queued hundreds of mbufs on the list under heavy
network load.
---
MP-ify SAD (savlist)
localcount(9) is used to protect savlist of sah. The basic design is
similar to MP-ifications of SPD and SAD sahlist. Please read the
locking notes of SAD for more details.
---
Simplify ipsec_reinject_ipstack (NFC)
---
Add per-CPU rtcache to ipsec_reinject_ipstack
It reduces route lookups and also reduces rtcache lock contentions
when NET_MPSAFE is enabled.
---
Use pool_cache(9) instead of pool(9) for tdb_crypto objects
The change improves network throughput especially on multi-core systems.
---
Update
ipsec(4), opencrypto(9) and vlan(4) are now MP-safe.
---
Write known issues on scalability
---
Share a global dummy SP between PCBs
It's never be changed so it can be pre-allocated and shared safely between PCBs.
---
Fix race condition on the rawcb list shared by rtsock and keysock
keysock now protects itself by its own mutex, which means that
the rawcb list is protected by two different mutexes (keysock's one
and softnet_lock for rtsock), of course it's useless.
Fix the situation by having a discrete rawcb list for each.
---
Use a dedicated mutex for rt_rawcb instead of softnet_lock if NET_MPSAFE
---
fix localcount leak in sav. fixed by ozaki-r@n.o.
I commit on behalf of him.
---
remove unnecessary comment.
---
Fix deadlock between pserialize_perform and localcount_drain
A typical ussage of localcount_drain looks like this:
mutex_enter(&mtx);
item = remove_from_list();
pserialize_perform(psz);
localcount_drain(&item->localcount, &cv, &mtx);
mutex_exit(&mtx);
This sequence can cause a deadlock which happens for example on the following
situation:
- Thread A calls localcount_drain which calls xc_broadcast after releasing
a specified mutex
- Thread B enters the sequence and calls pserialize_perform with holding
the mutex while pserialize_perform also calls xc_broadcast
- Thread C (xc_thread) that calls an xcall callback of localcount_drain tries
to hold the mutex
xc_broadcast of thread B doesn't start until xc_broadcast of thread A
finishes, which is a feature of xcall(9). This means that pserialize_perform
never complete until xc_broadcast of thread A finishes. On the other hand,
thread C that is a callee of xc_broadcast of thread A sticks on the mutex.
Finally the threads block each other (A blocks B, B blocks C and C blocks A).
A possible fix is to serialize executions of the above sequence by another
mutex, but adding another mutex makes the code complex, so fix the deadlock
by another way; the fix is to release the mutex before pserialize_perform
and instead use a condvar to prevent pserialize_perform from being called
simultaneously.
Note that the deadlock has happened only if NET_MPSAFE is enabled.
---
Add missing ifdef NET_MPSAFE
---
Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
---
- sanitize key debugging so that we don't print extra newlines or unassociated
debugging messages.
- remove unused functions and make internal ones static
- print information in one line per message
---
humanize printing of ip addresses
---
cast reduction, NFC.
---
Fix typo in comment
---
Pull out ipsec_fill_saidx_bymbuf (NFC)
---
Don't abuse key_checkrequest just for looking up sav
It does more than expected for example key_acquire.
---
Fix SP is broken on transport mode
isr->saidx was modified accidentally in ipsec_nextisr.
Reported by christos@
Helped investigations by christos@ and knakahara@
---
Constify isr at many places (NFC)
---
Include socketvar.h for softnet_lock
---
Fix buffer length for ipsec_logsastr
|
| 1.357.4.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #51):
sys/netinet/tcp_input.c: revision 1.358
tests/net/ipsec/t_ipsec_misc.sh: revision 1.7
Fix KASSERT in tcp_input
inp can be NULL when receiving an IPv4 packet on an IPv4-mapped IPv6
address. In that case KASSERT(sotoinpcb(so) == inp) always fails.
Should fix PR kern/52304 (at least it fixes the same panic as the
report)
--
Add test cases of TCP/IPsec on an IPv4-mapped IPv6 address
It reproduces the same panic reported in PR kern/52304
(but not sure that its cause is also same).
|
| 1.383.2.8 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.383.2.7 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.383.2.6 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.383.2.5 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.383.2.4 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.383.2.3 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.383.2.2 | 07-Apr-2018 |
pgoyette | Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
|
| 1.383.2.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.408.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.408.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.414.2.5 | 08-Oct-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #1894):
sys/netinet/tcp_input.c: revision 1.441
tcp_reass: Mitigate CVE-2018-6922 (SegmentSmack)
at a level of FreeBSD, by introducing an arbitrary (100) limit to
the length of TCP reassembly queues:
https://github.com/freebsd/freebsd-src/commit/95a914f6316874f5b0c45d491f2843dc810071ef
Originally authored by ryo@.
We thank Tomoyuki Sahara <tsahara at iij>, who has analyzed the
problem again, updated the patch, and carried out experiments for
vulnerability scenarios. The confidential PR below is based on
his work.
PR security/58708
|
| 1.414.2.4 | 13-Sep-2020 |
martin | Pull up following revision(s) (requested by kardel in ticket #1081):
sys/netinet/tcp_input.c: revision 1.420
PR/kern 55567
fix the data-only fast path. RCV.UP and SND.WL1 could be left behind
on long sequences of data only packets. pull them along to avoid relative
sequence wraps.
consistent with FreeBSD
addresses second failure mode of PR/kern 55567.
pullup to netbsd-8
pullup to netbsd-9
|
| 1.414.2.3 | 03-Sep-2020 |
martin | Pull up following revision(s) (requested by kardel in ticket #1074):
sys/netinet/tcp_input.c: revision 1.419
Fix fast path for uni directional transfers
pure ACK case:
drag snd_wl2 along so only newer
ACKs can update the window size.
also avoids the state where snd_wl2
is eventually larger than th_ack and thus
blocking the window update mechanism and
the connection gets stuck for a loooong
time in the zero sized send window state.
see PR/kern 55567
ok thorpej@, also found in FreeBSD
|
| 1.414.2.2 | 07-Jul-2020 |
martin | Pull up following revision(s) (requested by christos in ticket #985):
sys/netinet/tcp_input.c: revision 1.418
- always set both ip and ip6, otherwise a kernel assertion can be triggered
- move alignment early so that we do less work
|
| 1.414.2.1 | 10-Sep-2019 |
martin | Pull up following revision(s) (requested by maxv in ticket #193):
sys/netinet/tcp_timer.h: revision 1.30
sys/netinet/tcp_input.c: revision 1.415
sys/netinet/tcp_usrreq.c: revision 1.225
sys/netinet/tcp_subr.c: revision 1.283
Clamp tcp timer quantities to reasonable ranges.
|
| 1.424.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.428.4.1 | 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.438.2.1 | 08-Oct-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #932):
sys/netinet/tcp_input.c: revision 1.441
tcp_reass: Mitigate CVE-2018-6922 (SegmentSmack)
at a level of FreeBSD, by introducing an arbitrary (100) limit to
the length of TCP reassembly queues:
https://github.com/freebsd/freebsd-src/commit/95a914f6316874f5b0c45d491f2843dc810071ef
Originally authored by ryo@.
We thank Tomoyuki Sahara <tsahara at iij>, who has analyzed the
problem again, updated the patch, and carried out experiments for
vulnerability scenarios. The confidential PR below is based on
his work.
PR security/58708
|
| 1.439.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.222 | 08-Sep-2024 |
rillig | fix a/an grammar in obvious cases
|
| 1.221 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.220 | 29-Jun-2024 |
riastradh | branches: 1.220.2;
netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.219 | 13-Sep-2023 |
bouyer | Handle EHOSTDOWN the same way as EHOSTUNREACH and ENETDOWN for established
connections. Avoid premature end of tcp connection with "Host is down" error
in case of transient link-layer failure.
Discussed and patch proposed in
http://mail-index.netbsd.org/tech-net/2023/09/11/msg008610.html
and followups.
|
| 1.218 | 04-Nov-2022 |
ozaki-r | branches: 1.218.2;
inpcb: rename functions to in6pcb_*
|
| 1.217 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.216 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.215 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.214 | 30-Dec-2021 |
andvar | s/bandwith/bandwidth/
|
| 1.213 | 12-Jun-2020 |
roy | Remove in-kernel handling of Router Advertisements
This is much better handled by a user-land tool.
Proposed on tech-net here:
https://mail-index.netbsd.org/tech-net/2020/04/22/msg007766.html
Note that the ioctl SIOCGIFINFO_IN6 no longer sets flags. That now
needs to be done using the pre-existing SIOCSIFINFO_FLAGS ioctl.
Compat is fully provided where it makes sense, but trying to turn on
RA handling will obviously throw an error as it no longer exists.
Note that if you use IPv6 temporary addresses, this now needs to be
turned on in dhcpcd.conf(5) rather than in sysctl.conf(5).
|
| 1.212 | 17-Nov-2019 |
mlelstv | Don't allow zero sized segments that will panic the stack.
Reported-by: syzbot+5542516fa4afe7a101e6@syzkaller.appspotmail.com
|
| 1.211 | 25-Feb-2019 |
maxv | Improve panic messages.
|
| 1.210 | 27-Dec-2018 |
maxv | Remove unused arguments.
|
| 1.209 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.208 | 17-May-2018 |
maxv | branches: 1.208.2;
Remove reference to tcpiphdr in comment.
|
| 1.207 | 07-May-2018 |
uwe | Fix unsigned wraparound on window size calculations.
This is another instance where tp->rcv_adv - tp->rcv_nxt can wrap
around after successful zero-window probe from the peer. The first
one was fixed by chs@ in revision 1.112 on 2004-05-08.
While here, CSE and de-obfuscate the code a bit.
|
| 1.206 | 03-May-2018 |
maxv | Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.205 | 03-Apr-2018 |
maxv | bcopy -> memcpy, it's obvious the areas don't overlap.
|
| 1.204 | 01-Apr-2018 |
maxv | Change the check to be <= instead of <. This fixes one occurrence of an
apparently widespread division-by-zero bug in our TCP code: if a user adds
huge IPv6 options with setsockopt, and if the total size of the options
happens to be equal to the available space calculated for the TCP payload,
t_segsz gets set to zero, and given that we then divide several things by
it, the kernel crashes.
|
| 1.203 | 01-Apr-2018 |
maxv | Reorder and style, for clarity.
|
| 1.202 | 30-Mar-2018 |
maxv | Remove dead code. It was introduced in rev1 (25 years ago), and is
irrelevant today.
|
| 1.201 | 30-Mar-2018 |
maxv | Style, use NULL for pointers, use KASSERT, and don't inline huge functions,
we want to debug them with DDB (and not just with GPROF).
|
| 1.200 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.199 | 10-Mar-2018 |
khorben | Fix spello in a comment
|
| 1.198 | 12-Feb-2018 |
maxv | branches: 1.198.2;
Remove unused argument from tcp_signature_getsav.
|
| 1.197 | 03-Aug-2017 |
ozaki-r | Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
|
| 1.196 | 02-Jun-2017 |
ozaki-r | branches: 1.196.2;
Assert inph_locked on ipsec_pcb_skip_ipsec (was IPSEC_PCB_SKIP_IPSEC)
The assertion confirms SP caches are accessed under inph lock (solock).
|
| 1.195 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.194 | 04-Jan-2017 |
martin | branches: 1.194.2;
Fix optlen calculation for the SACK block - 2 bytes too few were
calculated, causing corruption in PR kern/51767.
|
| 1.193 | 04-Jan-2017 |
kre | Remove redundant tests: if optlen === 0, then optlen % 4 != 2 (it is 0)
so there is no need to test both.
|
| 1.192 | 03-Jan-2017 |
christos | use symbolic constants; no functional change.
|
| 1.191 | 03-Jan-2017 |
christos | put it the way we had it before; since we check for the resulting size after
we added the extra space we can be equal to the size of the buffer.
|
| 1.190 | 03-Jan-2017 |
christos | fix off-by-one
|
| 1.189 | 02-Jan-2017 |
christos | make sure that the reset label is defined without TCP_SIGNATURE.
|
| 1.188 | 02-Jan-2017 |
christos | Fix TCP signature code:
1. pack options more tightly instead of being generous with no/op
2. put TCP_SIGNATURE option before SACK
3. fix computation of options length, by deferring it
XXX: Really we should move the options setting code in one place instead
of having two copies one for input and one for output.
XXX: tcp_optlen/tcp_hdrsiz need to be fixed; they were wrong before too.
|
| 1.187 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.186 | 10-Jun-2016 |
ozaki-r | branches: 1.186.2;
Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.185 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.184 | 24-Jul-2015 |
matt | If we are sending a window probe and there's unacked data in the socket, make
sure at least the persist timer is running.
|
| 1.183 | 16-May-2015 |
kefren | Don't put segment on the wire if security request can't be fulfilled
|
| 1.182 | 27-Apr-2015 |
christos | Apply Revision 220794 from FreeBSD to avoid dup ACKs:
When checking to see if a window update should be sent to the remote peer,
don't force a window update if the window would not actually grow due to
window scaling. Specifically, if the window scaling factor is larger than
2 * MSS, then after the local reader has drained 2 * MSS bytes from the
socket, a window update can end up advertising the same window. If this
happens, the supposed window update actually ends up being a duplicate ACK.
This can result in an excessive number of duplicate ACKs when using a
higher maximum socket buffer size.
Pointed out by Ricky Charlet, in tech-net.
|
| 1.181 | 27-Apr-2015 |
ozaki-r | Introduce in6_selecthlim_rt to consolidate an idiom for rt->rt_ifp
It consolidates a scattered routine:
(rt = rtcache_validate(&in6p->in6p_route)) != NULL ? rt->rt_ifp : NULL
|
| 1.180 | 14-Feb-2015 |
he | Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
|
| 1.179 | 10-Nov-2014 |
maxv | branches: 1.179.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.178 | 25-Oct-2014 |
christos | Avoid stack overflow when SACK and TCP_SIGNATURE are both present. Thanks
to Jonathan Looney for pointing this out.
|
| 1.177 | 21-Oct-2014 |
hikaru | Fix wrong condition checking TSO capability.
ipsec_used is not necessary condition.
IPsec outbound policy will not be checked when ipsec_used is false.
|
| 1.176 | 30-May-2014 |
christos | branches: 1.176.2;
Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.175 | 05-Jun-2013 |
christos | branches: 1.175.2; 1.175.6;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.174 | 22-Mar-2012 |
drochner | branches: 1.174.2;
remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.173 | 31-Dec-2011 |
christos | branches: 1.173.2; 1.173.6; 1.173.8;
- fix offsetof usage, and redundant defines
- kill pointer casts to 0
|
| 1.172 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.171 | 14-Apr-2011 |
yamt | branches: 1.171.4; 1.171.8;
simplify a compile-time assertion
|
| 1.170 | 21-Mar-2011 |
matt | Clean up setting ECN bit in TOS. Fixes PR 44742
|
| 1.169 | 26-Jan-2010 |
pooka | branches: 1.169.4; 1.169.6;
tcp sockbuf autoscaling was initially added turned off because it
was experimental. People (including myself) have been running with
it turned on for eons now, so flip the default to enabled.
|
| 1.168 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.167 | 28-Apr-2008 |
martin | branches: 1.167.8; 1.167.10; 1.167.14; 1.167.16; 1.167.20;
Remove clause 3 and 4 from TNF licenses
|
| 1.166 | 12-Apr-2008 |
thorpej | branches: 1.166.2; 1.166.4;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.165 | 08-Apr-2008 |
thorpej | Change TCP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old tcpstat structure; old netstat
binaries will continue to work properly.
|
| 1.164 | 14-Jan-2008 |
dyoung | branches: 1.164.6;
Use rtcache_validate() instead of rtcache_getrt(). Shorten staircase
in in_losing().
|
| 1.163 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.162 | 02-Sep-2007 |
dyoung | branches: 1.162.6; 1.162.8; 1.162.12;
m_copy() was deprecated, apparently, long ago. m_copy(...) ->
m_copym(..., M_DONTWAIT).
|
| 1.161 | 02-Aug-2007 |
yamt | branches: 1.161.2; 1.161.4; 1.161.6;
make rfbuf_ts a tcp timestamp so that calculations in tcp_input make sense.
|
| 1.160 | 02-Aug-2007 |
rmind | TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.159 | 18-May-2007 |
riz | branches: 1.159.2;
Fix compilation in the TCP_SIGNATURE case:
- don't use void * for pointer arithmetic
- don't try to modify const parameters
A kernel with 'options TCP_SIGNATURE' works as well as it ever did, now.
(ie, clunky, but passable)
|
| 1.158 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.157 | 04-Mar-2007 |
christos | branches: 1.157.2; 1.157.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.156 | 22-Feb-2007 |
thorpej | TRUE -> true, FALSE -> false
|
| 1.155 | 21-Feb-2007 |
thorpej | Replace the Mach-derived boolean_t type with the C99 bool type. A
future commit will replace use of TRUE and FALSE with true and false.
|
| 1.154 | 10-Feb-2007 |
degroote | branches: 1.154.2;
Commit my SoC work
Add ipv6 support for fast_ipsec
Note that currently, packet with extensions headers are not correctly
supported
Change the ipcomp logic
|
| 1.153 | 25-Nov-2006 |
yamt | branches: 1.153.2; 1.153.4;
move tso-by-software code to their own files. no functional changes.
|
| 1.152 | 23-Nov-2006 |
martin | Make it compile on IPv4-only kernels
|
| 1.151 | 23-Nov-2006 |
yamt | implement ipv6 TSO.
partly from Matthias Scheler. tested by him.
|
| 1.150 | 17-Oct-2006 |
yamt | tcp_output: as a comment in tcp_sack_newack says, actually send
one or two segments on partial acks. even if sack_bytes_rxmt==0,
if we are in fast recovory with sack, snd_cwnd has somewhat special
meaning here. PR/34749.
|
| 1.149 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.148 | 08-Oct-2006 |
yamt | tcp_output: don't make TSO duplicate CWR/ECE.
|
| 1.147 | 08-Oct-2006 |
yamt | tcp_output: don't try to send SACK option larger than txsegsize.
fix a panic like "panic: m_copydata: off 0, len -7".
|
| 1.146 | 07-Oct-2006 |
yamt | tcp_output: remove duplicated code and tweak indent. no functional changes.
|
| 1.145 | 01-Oct-2006 |
dbj | back out revision 1.144 calculating txsegsizep since it unmasks
other bugs. See PR kern/34674
|
| 1.144 | 28-Sep-2006 |
dbj | consider sb_lowat when limiting the transmit length to keep acks on the wire
|
| 1.143 | 05-Sep-2006 |
rpaulo | branches: 1.143.2; 1.143.4;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.142 | 25-Mar-2006 |
seanb | Slight simplification of hdr len calculation in tcp_segsize().
No functional change.
|
| 1.141 | 24-Dec-2005 |
perry | branches: 1.141.4; 1.141.6; 1.141.8; 1.141.10; 1.141.12;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.140 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.139 | 10-Aug-2005 |
yamt | wrap INET-only code by #if defined(INET).
|
| 1.138 | 10-Aug-2005 |
yamt | ipv6 tx checksum offloading. reviewed by Jason Thorpe.
|
| 1.137 | 19-Jul-2005 |
christos | Implement PMTU checks from:
http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.
From OpenBSD.
|
| 1.136 | 28-Jun-2005 |
drochner | branches: 1.136.2;
typo in comment
|
| 1.135 | 29-May-2005 |
christos | - add const
- remove bogus casts
- avoid nested variables
|
| 1.134 | 08-May-2005 |
yamt | tcp_output: account FIN when building sack option.
|
| 1.133 | 08-May-2005 |
yamt | tcp_output: don't try to send more data than we have. PR/30160.
|
| 1.132 | 08-May-2005 |
yamt | tcp_output: clear TH_FIN where appropriate. related to PR/30160.
|
| 1.131 | 18-Apr-2005 |
yamt | add a function to handle M_CSUM_TSOv4 by software.
|
| 1.130 | 18-Apr-2005 |
yamt | fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.129 | 29-Mar-2005 |
yamt | tcp_output: lock reass queue when building sack.
|
| 1.128 | 16-Mar-2005 |
yamt | branches: 1.128.2;
simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]),
build it directly from the segment list when needed.
|
| 1.127 | 16-Mar-2005 |
yamt | - use full sized segments unless we actually have SACKs to send.
- avoid TSO duplicate D-SACK.
- send SACKs regardless of TF_ACKNOW.
- don't clear rcv_sack_num when transmitting.
discussed on tech-net@.
|
| 1.126 | 12-Mar-2005 |
yamt | don't try to use TSO to transmit a single segment.
- there's no benefit.
- rtl8169 seems to be stuck with it.
|
| 1.125 | 09-Mar-2005 |
matt | For AF_INET, always set m->m_pkthdr.csum_data. Don't or TSOv4, just set it.
|
| 1.124 | 07-Mar-2005 |
yamt | tcp_sack_option: the max number of sack blocks in a packet is 4, not 3.
|
| 1.123 | 06-Mar-2005 |
thorpej | Add a /*CONSTCOND*/ to last.
|
| 1.122 | 06-Mar-2005 |
matt | Fix typo. Opposite of >= is <, not ==.
|
| 1.121 | 06-Mar-2005 |
matt | Replace some gotos with a do while (0) and breaks. No functional change.
|
| 1.120 | 06-Mar-2005 |
matt | Add IPv4/TCP hooks for TCP Segment Offload on transmit.
|
| 1.119 | 02-Mar-2005 |
mycroft | Copyright maintenance.
|
| 1.118 | 28-Feb-2005 |
jonathan | Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.117 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.116 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.115 | 15-Dec-2004 |
thorpej | branches: 1.115.2; 1.115.4;
Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.114 | 20-May-2004 |
jonathan | With FAST_IPSEC, include <netipsec/key.h>, as Itojun's recent changes
now require KEY_FREESAV() to be in scope.
|
| 1.113 | 18-May-2004 |
itojun | fix MD5 signature support to actually validate inbound signature, and
drop packet if fails.
|
| 1.112 | 08-May-2004 |
chs | work around an LP64 problem where we report an excessively large window
due to incorrect mixing of types.
|
| 1.111 | 26-Apr-2004 |
itojun | make TCP MD5 signature work with KAME IPSEC (#define IPSEC).
support IPv6 if KAME IPSEC (RFC is not explicit about how we make data stream
for checksum with IPv6, but i'm pretty sure using normal pseudo-header is the
right thing).
XXX
current TCP MD5 signature code has giant flaw:
it does not validate signature on input (can't believe it! what is the point?)
|
| 1.110 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.109 | 30-Mar-2004 |
christos | Make sure we disarm the persist timer before we arm the rexmit
timer, otherwise there is a tiny window where both timers are
active, and this is not correct according to the comments in the
code. I believe that this is the cause of the to_ticks <= 0 assertion
failure in callout_schedule() that I've been getting.
|
| 1.108 | 03-Mar-2004 |
thorpej | branches: 1.108.2;
Use IPSEC_PCB_SKIP_IPSEC() to short-circuit calls to ipsec{4,6}_hdrsiz_tcp().
|
| 1.107 | 04-Feb-2004 |
itojun | deal with IPv6 path MTU < 1280 (RFC2460 section 5 last paragraph).
check if there really is room for TCP data.
|
| 1.106 | 12-Nov-2003 |
ragge | Remove the FAST_MBSEARCH ifdef, send packet prediction is now default.
|
| 1.105 | 24-Oct-2003 |
ragge | Fix the bug in the tcp transmit prediction code.
During testing the prediction counters show a hit-rate on about 85% for
packets sent on a local LAN, and better than 99% for intercontinental
high-speed bulk traffic (!).
|
| 1.104 | 24-Oct-2003 |
enami | Make this file compile again when TCP_OUTPUT_COUNTERS defined.
|
| 1.103 | 23-Oct-2003 |
thorpej | Oops, FAST_MBSEARCH counters were swapped; fix it. Pointed out by yamt@.
|
| 1.102 | 21-Oct-2003 |
thorpej | Add event counters that measure FAST_MBSEARCH.
|
| 1.101 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.100 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.99 | 22-Aug-2003 |
jonathan | Replace the set_socket() method of passing an extra struct socket*
argument to ip6_output() with a new explicit struct in6pcb* argument.
(The underlying socket can be obtained via in6pcb->inp6_socket.)
In preparation for fast-ipsec. Reviewed by itojun.
|
| 1.98 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.97 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.96 | 02-Jul-2003 |
ragge | Make the fast-search stuff an option. There are still reports on
problem with it.
|
| 1.95 | 02-Jul-2003 |
ragge | Fix previous bug. Thanks to Enami for spotting the (obvious) error, and
to other people with much help with bug reports etc.
While fixing, change some of the code I added last time to make it
cleaner and simpler.
|
| 1.94 | 30-Jun-2003 |
ragge | branches: 1.94.2;
Disable the code I checked in yesterday; reports that samba (!) are crashing
machines with it. Will do some more tests.
|
| 1.93 | 29-Jun-2003 |
fvdl | Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.92 | 29-Jun-2003 |
ragge | Add code to remember where in the send queue of mbufs the last packet was
sent from. This change avoid a linear search through all mbufs when using
large TCP windows, and therefore permit high-speed connections on long
distances.
Tested on a 1 Gigabit connection between Lule� and San Francisco, a distance
of about 15000km. With TCP windows of just over 20 Mbytes it could keep up
with 950Mbit/s.
After discussions with Matt Thomas and Jason Thorpe.
|
| 1.91 | 17-May-2003 |
itojun | no need for ip_v recovery in output path too
(tcp_template includes ip_v setting)
|
| 1.90 | 01-Mar-2003 |
thorpej | Allow TCP connections to hosts on a local network to use a larger
slow start initial window. Default this larger initial window to
4 packets, allowing it to be adjusted with net.inet.tcp.init_win_local.
|
| 1.89 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.88 | 24-Nov-2002 |
scw | Fix a genuine uninitialised variable warning.
|
| 1.87 | 02-Nov-2002 |
itojun | cleanup ipsec.h dependency. commented by perry, sync w/kame
|
| 1.86 | 13-Sep-2002 |
mycroft | In the txsegsize bounding code, it is not necessary to adjust for the options
length.
|
| 1.85 | 20-Aug-2002 |
thorpej | Never send more than half a socket buffer of data. This insures that
we can always keep 2 packets on the wire, no matter what SO_SNDBUF is,
and therefore ACKs will never be delayed unless we run out of data to
transmit. The problem is quite easy to tickle when the MTU of the
outgoing interface is larger than the socket buffer size (e.g. loopback).
Fix from Charles Hannum.
|
| 1.84 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.83 | 13-Jun-2002 |
thorpej | Disable TCP Congestion Window Monitoring by default; there are
performance problems in the face of tinygrams.
|
| 1.82 | 09-Jun-2002 |
itojun | whitespace
|
| 1.81 | 29-May-2002 |
itojun | attach nd_ifinfo structure into if_afdata.
split IPv6 link MTU (advertised by RA) from real link MTU.
sync with kame
|
| 1.80 | 26-May-2002 |
itojun | path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.79 | 27-Apr-2002 |
thorpej | branches: 1.79.2; 1.79.4;
* Instrument tcp_build_datapkt().
* Remove the code that allocates a cluster if the packet would
fit in one; it totally defeats doing references to M_EXT mbufs
in the socket buffer. This drastically reduces the number of
data copies in the tcp_output() path for applications which use
large writes. Kudos to Matt Thomas for pointing me in the right
direction.
|
| 1.78 | 01-Mar-2002 |
thorpej | In tcp_segsize(), move a label so that option length is considered
when using the default TCP MSS as well. From Matt Thomas.
|
| 1.77 | 24-Jan-2002 |
itojun | place NRL copyright notice itself, not a reference to it.
|
| 1.76 | 03-Dec-2001 |
jmcneill | Fix TCP segment size computation. From Rick Byersm, PR kern/14799.
|
| 1.75 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.74 | 10-Sep-2001 |
thorpej | Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.
|
| 1.73 | 10-Sep-2001 |
thorpej | Change the way receive idle time and round trip time are measured.
Instead of incrementing t_idle and t_rtt in tcp_slowtimo(), we now
take a timstamp (via tcp_now) and use subtraction to compute the
delta when we actually need it (using unsigned arithmetic so that
tcp_now wrapping is handled correctly).
Based on similar changes in FreeBSD.
|
| 1.72 | 10-Sep-2001 |
thorpej | Enable Congestion Window Monitoring by default.
|
| 1.71 | 10-Sep-2001 |
thorpej | Use a callout for the delayed ACK timer, and delete tcp_fasttimo().
Expose the delayed ACK timer as net.inet.tcp.delack_ticks.
|
| 1.70 | 31-Jul-2001 |
thorpej | branches: 1.70.2;
Carve off the code that builds a TCP data packet into its own
function, and inline it, except when profiling... so we can
profile it.
|
| 1.69 | 31-Jul-2001 |
thorpej | Count the number of times we "self-quench" (ip_output() returns
ENOBUFS), and don't inline tcp_segsize() if profiling.
|
| 1.68 | 26-Jul-2001 |
thorpej | Slight cosmetic change.
|
| 1.67 | 08-Jul-2001 |
abs | branches: 1.67.2;
Rename TCPDEBUG to TCP_DEBUG, defopt TCP_DEBUG and TCP_NDEBUG, and
make all usage of tcp_trace dependent on TCP_DEBUG - resulting in
a 31K saving on an INET enabled i386 kernel.
|
| 1.66 | 02-Jun-2001 |
thorpej | Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.65 | 03-Apr-2001 |
itojun | check ip_mtudisc only for TCP over IPv4.
PMTUD is mandatory for TCP over IPv6 (if packets > 1280).
|
| 1.64 | 20-Mar-2001 |
thorpej | Two changes, designed to make us even more resilient against TCP
ISS attacks (which we already fend off quite well).
1. First-cut implementation of RFC1948, Steve Bellovin's cryptographic
hash method of generating TCP ISS values. Note, this code is experimental
and disabled by default (experimental enough that I don't export the
variable via sysctl yet, either). There are a couple of issues I'd
like to discuss with Steve, so this code should only be used by people
who really know what they're doing.
2. Per a recent thread on Bugtraq, it's possible to determine a system's
uptime by snooping the RFC1323 TCP timestamp options sent by a host; in
4.4BSD, timestamps are created by incrementing the tcp_now variable
at 2 Hz; there's even a company out there that uses this to determine
web server uptime. According to Newsham's paper "The Problem With
Random Increments", while NetBSD's TCP ISS generation method is much
better than the "random increment" method used by FreeBSD and OpenBSD,
it is still theoretically possible to mount an attack against NetBSD's
method if the attacker knows how many times the tcp_iss_seq variable
has been incremented. By not leaking uptime information, we can make
that much harder to determine. So, we avoid the leak by giving each
TCP connection a timebase of 0.
|
| 1.63 | 24-Jan-2001 |
itojun | branches: 1.63.2;
- record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.62 | 06-Nov-2000 |
itojun | fix IPv4 TTL selection with AF_INET6 API. sync with kame. From: jdc
|
| 1.61 | 19-Oct-2000 |
itojun | remove #ifdef TCP6. it is not likely for us to bring in sys/netinet6/tcp6*.c
(separate TCP/IPv6 stack) into netbsd-current.
|
| 1.60 | 17-Oct-2000 |
itojun | be more friendly with INET-less build.
XXX we need to do more to do a working INET-less build
|
| 1.59 | 17-Oct-2000 |
thorpej | Add an IP_MTUDISC flag to the flags that can be passed to
ip_output(). This flag, if set, causes ip_output() to set
DF in the IP header if the MTU in the route is not locked.
This allows a bunch of redundant code, which I was never
really all that happy about adding in the first place, to
be eliminated.
Inspired by a similar change made by provos@openbsd.org when
he integrated NetBSD's Path MTU Discovery code into OpenBSD.
|
| 1.58 | 28-Jul-2000 |
itojun | forgot to call tcp6_quench(). sync with kame.
|
| 1.57 | 30-Jun-2000 |
itojun | remove old mbuf assumption (ip header and tcp header are on the same mbuf).
this is for m_pulldown use. (sync with kame)
|
| 1.56 | 30-Mar-2000 |
augustss | branches: 1.56.4;
Remove register declarations.
|
| 1.55 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.54 | 09-Feb-2000 |
itojun | optimize mbuf allocation for ip/tcp/tcpopt part.
|
| 1.53 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.52 | 23-Sep-1999 |
itojun | branches: 1.52.2; 1.52.8;
cleanup and correct TCP MSS consideration with IPsec headers.
MSS advertisement must always be:
max(if mtu) - ip hdr siz - tcp hdr siz
We violated this in the previous code so it was fixed.
tcp_mss_to_advertise() now takes af (af on wire) as its argument,
to compute right ip hdr siz.
tcp_segsize() will take care of IPsec header size.
One thing I'm not really sure is how to handle IPsec header size in
*rxsegsizep (inbound segment size estimation).
The current code subtracts possible *outbound* IPsec size from *rxsegsizep,
hoping that the peer is using the same IPsec policy as me.
It may not be applicable, could TCP gulu please comment...
|
| 1.51 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.50 | 02-Jul-1999 |
fvdl | Fix for -Wunitialized warnings broke compiles without INET6, refix.
|
| 1.49 | 02-Jul-1999 |
itojun | avoid "variable not initialized" warnings on some of the platforms.
|
| 1.48 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.47 | 20-Jan-1999 |
thorpej | branches: 1.47.4; 1.47.6;
Fix a problem pointed out by Charles Hannum; DF wasn't being set in
SYN,ACK packets during Path MTU Discovery. Fix tcp_respond() to do the
appropriate route lookup and set DF as appropriate.
Also, fixup similar code in tcp_output() to relookup the route if it
is down.
|
| 1.46 | 16-Dec-1998 |
thorpej | Delay sending if SS_MORETOCOME is set in so_state. This avoids the case
where the user issued a write with a length greater than MLEN but less
than MINCLSIZE, thus causing two mbufs to be used. The loop in sosend()
would then call PRU_SEND twice, causing TCP to transmit 2 packets when
it could have transmitted one.
Suggested by Justin Walker <justin@apple.com> on the freebsd-net
mailing list.
|
| 1.45 | 06-Oct-1998 |
matt | Add a sysctl for newreno (default to off).
|
| 1.44 | 04-Oct-1998 |
matt | Adapt the NEWRENO changes from the UCSB diffs of BSDI 3.0's TCP
to NetBSD. Ignore the SACK & FACK stuff for now.
|
| 1.43 | 21-Jul-1998 |
mycroft | Implement a better fix for the `gratuitous FIN' problem, as
mentioned on tcp-impl but with a bit more commentary.
|
| 1.42 | 17-Jul-1998 |
thorpej | Add a comment wrt. a current issue w/ CWM.
|
| 1.41 | 17-Jul-1998 |
thorpej | Comment where the Restart Window is computed, and in the non-CWM case,
make sure it never _increases_ cwnd.
|
| 1.40 | 07-Jul-1998 |
sommerfe | Delete bogus (void) cast of m_freem (which is already a void function..)
|
| 1.39 | 11-May-1998 |
thorpej | Nuke TUBA per my note to tech-net; there's no reason to keep it around.
|
| 1.38 | 06-May-1998 |
thorpej | Use macros from tcp_timer.h to manipulate TCP timers, so that their
implementation can be changed easily.
|
| 1.37 | 02-May-1998 |
thorpej | Correct a comment related to Congestion Window Monitoring.
|
| 1.36 | 30-Apr-1998 |
thorpej | In the CWM code, don't use the Floyd initial window computation as
the burst size allowed, but rather a fixed number of packets, as
described in the Internet Draft. Default allowed burst is 4 packets,
per the Draft.
Make the use of CWM and the allowed burst size tunable via sysctl.
|
| 1.35 | 29-Apr-1998 |
kml | Add support for deletion of routes added by path MTU discovery;
uses new generic route timeout code. Add sysctl for timeout period.
|
| 1.34 | 13-Apr-1998 |
kml | Fix to ensure that the correct MSS is advertised for loopback
TCP connections by using the MTU of the interface. Also added
a knob, mss_ifmtu, to force all connections to use the MTU of
the interface to calculate the advertised MSS.
|
| 1.33 | 01-Apr-1998 |
thorpej | Implement Congestion Window Monitoring as described in the TCPIMPL
meeting of IETF #41 by Amy Hughes <ahughes@isi.edu>, and in an upcoming
internet draft from Hughes, Touch, and Heidemann.
CWM eliminates line-rate bursts after idle periods by counting pending
(unacknowledged) packets and limiting the congestion window to the
initial congestion window plus the pending packet count. This has the
effect of allowing us to use the window as long as we continue to transmit,
but as soon as we stop transmitting, we go back to a slow-start (also known
as `use it or lose it').
This is not enabled by default. You can enable this behavior by patching
the "tcp_cwm" global (set it to non-zero) or by building a kernel with the
TCP_CWM option.
|
| 1.32 | 31-Mar-1998 |
thorpej | Fix a potential-congestion case in the larger initial congestion window
code, as clarified in the TCPIMPL WG meeting at IETF #41: If the SYN
(active open) or SYN,ACK (passive open) was retransmitted, the initial
congestion window for the first slow start of that connection must be
one segment.
|
| 1.31 | 24-Mar-1998 |
kml | Ensure that we take the IP option length into account when we calculate
the effective maximum send size for TCP. ip_optlen() and tcp_optlen()
should probably be inlined for efficiency.
|
| 1.30 | 19-Mar-1998 |
kml | Fix a retransmission bug introduced by the Brakmo and Peterson
RTO estimation changes. Under some circumstances it would return a value
of 0, while the old Van Jacobson RTO code would return a minimum of 3.
This would result in 12 retransmissions, each 1 second apart.
This takes care of those instances, and ensures that t_rttmin is
used everywhere as a lower bound.
|
| 1.29 | 17-Mar-1998 |
kml | Ensure that the TCP segment size reflects the size of TCP options
in the packet. This fixes a bug that was resulting in extra packets
in retransmissions (the second packet would be 12 bytes long,
reflecting the RFC1323 timestamp option size).
|
| 1.28 | 19-Feb-1998 |
thorpej | Update copyright (sigh, should have done this long ago).
|
| 1.27 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.26 | 31-Dec-1997 |
thorpej | Implement a queue for delayed ACK processing. This queue is used in
tcp_fasttimo() in lieu of scanning all open TCP connections.
|
| 1.25 | 17-Dec-1997 |
thorpej | From 4.4BSD-Lite2:
- If we fail to allocate mbufs for the outgoing segment, free the header
and abort.
From Stevens:
- Ensure the persist timer is running if the send window reaches zero.
Part of the fix for kern/2335 (pete@daemon.net).
|
| 1.24 | 11-Dec-1997 |
thorpej | Implement an infrastructure to allow larger initial congestion windows.
The sysctl'able variable "tcp_init_win", when set to 0, selects an
auto-tuning algorithm for selecting the initial window, based on transmit
segment size, per discussion in the IETF tcpimpl working group.
Default initial window is still 1 segment, but will soon become 2 segments,
per discussion in tcpimpl.
|
| 1.23 | 11-Dec-1997 |
thorpej | Count delayed ACKs after they have been sucessfully transmitted.
|
| 1.22 | 20-Nov-1997 |
thorpej | Add missing (implied) int to a variable declaration.
|
| 1.21 | 08-Nov-1997 |
kml | TCP MSS fixes to provide cleaner slow-start and recovery.
|
| 1.20 | 18-Oct-1997 |
kml | branches: 1.20.2;
change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.19 | 17-Oct-1997 |
kml | Path MTU Discovery support. This is turned off by default.
Use sysctl -w net.inet.icmp.mtudisc=1 to turn on.
Still to come: path removal after some period, black hole detection
|
| 1.18 | 08-Oct-1997 |
thorpej | Fix an oversight in my previous MSS-related changes:
Basically, in silly window avoidance, don't use the raw MSS we advertised
to the peer. What we really want here is the _expected_ size of received
segments, so we need to account for the path MTU (eventually; right now,
the interface MTU for "local" addresses and loopback or tcp_mssdflt for
non-local addresses). Without this, silly window avoidance would never
kick in if we advertised a very large (e.g. ~64k) MSS to the peer.
|
| 1.17 | 22-Sep-1997 |
thorpej | Fix several annoyances related to MSS handling in BSD TCP:
- Don't overload t_maxseg. Previous behavior was to set it to the min
of the peer's advertised MSS, our advertised MSS, and tcp_mssdflt
(for non-local networks). This breaks PMTU discovery running on
either host. Instead, remember the MSS we advertise, and use it
as appropriate (in silly window avoidance).
- Per last bullet, split tcp_mss() into several functions for handling
MSS (ours and peer's), and performing various tasks when a connection
becomes ESTABLISHED.
- Introduce a new function, tcp_segsize(), which computes the max size
for every segment transmitted in tcp_output(). This will eventually
be used to hook in PMTU discovery.
|
| 1.16 | 03-Jun-1997 |
kml | branches: 1.16.4;
Fix urgent pointer overflow problems when used with large windows
|
| 1.15 | 10-Dec-1996 |
mycroft | Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.14 | 13-Feb-1996 |
christos | branches: 1.14.4;
netinet prototypes
|
| 1.13 | 13-Apr-1995 |
cgd | oops; missed the chance to fix a cast, that then becamse a compiler warning.
|
| 1.12 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.11 | 23-Jan-1995 |
mycroft | Fix a condition where we sometimes sent a FIN too early. Also, a small
optimization.
|
| 1.10 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 12-Apr-1994 |
mycroft | Acks with no data should have the highest sequence number sent.
|
| 1.7 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.6 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.3 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.14.4.1 | 10-Dec-1996 |
mycroft | From trunk:
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.16.4.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.16.4.1 | 29-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.20.2.6 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.20.2.5 | 05-May-1998 |
mycroft | Pull up 1.29, per request of kml.
|
| 1.20.2.4 | 05-May-1998 |
mycroft | Pull up 1.30, per request of kml.
|
| 1.20.2.3 | 29-Jan-1998 |
mellon | Pull up 1.24-1.27 (thorpej)
|
| 1.20.2.2 | 21-Nov-1997 |
thorpej | Sync w/ trunk: add a missing (previously implied) int.
|
| 1.20.2.1 | 08-Nov-1997 |
thorpej | Pull up from trunk: TCP MSS fixes to provide cleaner slow-start and recovery.
(kml)
|
| 1.47.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.47.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.47.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.47.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.47.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.52.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.52.2.5 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.52.2.4 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.52.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.52.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.52.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.56.4.5 | 24-Jan-2002 |
he | Pull up revision 1.77 (requested by itojun):
Clean up the NRL copyright.
|
| 1.56.4.4 | 06-Apr-2001 |
he | Pull up revision 1.63 (requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.56.4.3 | 10-Nov-2000 |
tv | Pullup 1.62 [itojun]:
fix IPv4 TTL selection with AF_INET6 API. sync with kame. From: jdc
|
| 1.56.4.2 | 15-Aug-2000 |
itojun | pullup 1.57 -> 1.58 (approved by releng-1-5)
> forgot to call tcp6_quench(). sync with kame.
|
| 1.56.4.1 | 23-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
remove old mbuf assumption (ip header and tcp header are on the same mbuf).
this is for m_pulldown use. (sync with kame)
1.108 -> 1.109 syssrc/sys/netinet/tcp_input.c
1.56 -> 1.57 syssrc/sys/netinet/tcp_output.c
1.91 -> 1.92 syssrc/sys/netinet/tcp_subr.c
|
| 1.63.2.14 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.63.2.13 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.63.2.12 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.63.2.11 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.63.2.10 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.63.2.9 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.63.2.8 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.63.2.7 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.63.2.6 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.63.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.63.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.63.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.63.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.63.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.67.2.8 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.67.2.7 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.67.2.6 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.67.2.5 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.67.2.4 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.67.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.67.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.67.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.70.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.79.4.5 | 07-Feb-2004 |
jmc | Pullup rev 1.107 (requested by itojun in ticket #1605)
Deal with IPv6 path MTU < 1280 (RFC2460 section 5 last paragraph)
Check if there really is room for TCP data.
|
| 1.79.4.4 | 05-Sep-2003 |
tron | Pull up revision 1.80 (requested by tls in ticket #1445):
path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.79.4.3 | 30-Nov-2002 |
he | Pull up revision 1.86 (requested by thorpej in ticket #795):
In the txsegsize bounding code, it is not necessary to adjust
for the options length.
|
| 1.79.4.2 | 21-Nov-2002 |
he | Pull up revision 1.85 (requested by thorpej in ticket #707):
Never send more than half a socket buffer of data in a
segment. This ensures that we can always keep 2 packets
on the wire, and we will therefore not cause any delayed
ACKs. Otherwise, this causes performance problems when
using large-MTU interfaces, such as the loopback interface.
|
| 1.79.4.1 | 14-Jun-2002 |
lukem | Pull up revision 1.83 (requested by thorpej in ticket #267):
Disable TCP Congestion Window Monitoring by default; there are
performance problems in the face of tinygrams.
|
| 1.79.2.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.79.2.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.79.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.94.2.9 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.94.2.8 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.94.2.7 | 08-Mar-2005 |
skrll | Sync with HEAD.
|
| 1.94.2.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.94.2.5 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.94.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.94.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.94.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.94.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.108.2.1 | 11-May-2004 |
tron | Pull up revision 1.112 (requested by chs in ticket #292):
work around an LP64 problem where we report an excessively large window
due to incorrect mixing of types.
|
| 1.115.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.115.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.115.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.128.2.5 | 11-May-2005 |
tron | Pull up revision 1.134 (requested by yamt in ticket #294):
tcp_output: account FIN when building sack option.
|
| 1.128.2.4 | 11-May-2005 |
tron | Pull up revision 1.133 (requested by yamt in ticket #293):
tcp_output: don't try to send more data than we have. PR/30160.
|
| 1.128.2.3 | 11-May-2005 |
tron | Pull up revision 1.132 (requested by yamt in ticket #293):
tcp_output: clear TH_FIN where appropriate. related to PR/30160.
|
| 1.128.2.2 | 06-May-2005 |
tron | Pull up revision 1.130 (requested by yamt in ticket #251):
fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.128.2.1 | 04-Apr-2005 |
tron | Pull up revision 1.129 (requested by yamt in ticket #89):
tcp_output: lock reass queue when building sack.
|
| 1.136.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.136.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.136.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.136.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.136.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.141.12.1 | 28-Mar-2006 |
tron | Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
|
| 1.141.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.141.8.2 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.141.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.141.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.141.4.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.141.4.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.143.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.143.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.143.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.143.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.153.4.1 | 04-Jun-2007 |
wrstuden | Update to today's netbsd-4.
|
| 1.153.2.2 | 03-Apr-2011 |
riz | Pull up following revision(s) (requested by spz in ticket #1424):
sys/netinet/tcp_output.c: revision 1.170
Clean up setting ECN bit in TOS. Fixes PR 44742
|
| 1.153.2.1 | 24-May-2007 |
pavel | branches: 1.153.2.1.4;
Pull up following revision(s) (requested by degroote in ticket #667):
sys/netinet/tcp_input.c: revision 1.260
sys/netinet/tcp_output.c: revision 1.154
sys/netinet/tcp_subr.c: revision 1.210
sys/netinet6/icmp6.c: revision 1.129
sys/netinet6/in6_proto.c: revision 1.70
sys/netinet6/ip6_forward.c: revision 1.54
sys/netinet6/ip6_input.c: revision 1.94
sys/netinet6/ip6_output.c: revision 1.114
sys/netinet6/raw_ip6.c: revision 1.81
sys/netipsec/ipcomp_var.h: revision 1.4
sys/netipsec/ipsec.c: revision 1.26 via patch,1.31-1.32
sys/netipsec/ipsec6.h: revision 1.5
sys/netipsec/ipsec_input.c: revision 1.14
sys/netipsec/ipsec_netbsd.c: revision 1.18,1.26
sys/netipsec/ipsec_output.c: revision 1.21 via patch
sys/netipsec/key.c: revision 1.33,1.44
sys/netipsec/xform_ipcomp.c: revision 1.9
sys/netipsec/xform_ipip.c: revision 1.15
sys/opencrypto/deflate.c: revision 1.8
Commit my SoC work
Add ipv6 support for fast_ipsec
Note that currently, packet with extensions headers are not correctly
supported
Change the ipcomp logic
Add sysctl tree to modify the fast_ipsec options related to ipv6. Similar
to the sysctl kame interface.
Choose the good default policy, depending of the adress family of the
desired policy
Increase the refcount for the default ipv6 policy so nobody can reclaim it
Always compute the sp index even if we don't have any sp in spd. It will
let us to choose the right default policy (based on the adress family
requested).
While here, fix an error message
Use dynamic array instead of an static array to decompress. It lets us to
decompress any data, whatever is the radio decompressed data / compressed
data.
It fixes the last issues with fast_ipsec and ipcomp.
While here, bzero -> memset, bcopy -> memcpy, FREE -> free
Reviewed a long time ago by sam@
|
| 1.153.2.1.4.1 | 03-Apr-2011 |
riz | Pull up following revision(s) (requested by spz in ticket #1424):
sys/netinet/tcp_output.c: revision 1.170
Clean up setting ECN bit in TOS. Fixes PR 44742
|
| 1.154.2.3 | 07-May-2007 |
yamt | sync with head.
|
| 1.154.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.154.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.157.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.157.2.3 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.157.2.2 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.157.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.159.2.2 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.159.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.161.6.2 | 02-Aug-2007 |
yamt | make rfbuf_ts a tcp timestamp so that calculations in tcp_input make sense.
|
| 1.161.6.1 | 02-Aug-2007 |
yamt | file tcp_output.c was added on branch matt-mips64 on 2007-08-02 13:12:36 +0000
|
| 1.161.4.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.161.4.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.161.4.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.161.2.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.162.12.2 | 19-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.162.12.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.162.8.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.162.6.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.164.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.166.4.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.166.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.166.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.166.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.167.20.2 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1973):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.167.20.1 | 29-Mar-2011 |
riz | Pull up following revision(s) (requested by spz in ticket #1586):
sys/netinet/tcp_output.c: revision 1.170
Clean up setting ECN bit in TOS. Fixes PR 44742
|
| 1.167.16.1 | 29-Mar-2011 |
riz | Pull up following revision(s) (requested by spz in ticket #1586):
sys/netinet/tcp_output.c: revision 1.170
Clean up setting ECN bit in TOS. Fixes PR 44742
|
| 1.167.14.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.167.10.2 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1973):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.167.10.1 | 29-Mar-2011 |
riz | branches: 1.167.10.1.2;
Pull up following revision(s) (requested by spz in ticket #1586):
sys/netinet/tcp_output.c: revision 1.170
Clean up setting ECN bit in TOS. Fixes PR 44742
|
| 1.167.10.1.2.1 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1973):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.167.8.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.169.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.169.4.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.171.8.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.171.8.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.171.4.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.171.4.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.173.8.2 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1315):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.173.8.1 | 03-Nov-2014 |
msaitoh | Pull up following revision(s) (requested by christos in ticket #1174):
sys/netinet/tcp_output.c: revision 1.178
Avoid stack overflow when SACK and TCP_SIGNATURE are both present. Thanks
to Jonathan Looney for pointing this out.
|
| 1.173.6.2 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1315):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.173.6.1 | 03-Nov-2014 |
msaitoh | Pull up following revision(s) (requested by christos in ticket #1174):
sys/netinet/tcp_output.c: revision 1.178
Avoid stack overflow when SACK and TCP_SIGNATURE are both present. Thanks
to Jonathan Looney for pointing this out.
|
| 1.173.2.2 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #1315):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.173.2.1 | 03-Nov-2014 |
msaitoh | Pull up following revision(s) (requested by christos in ticket #1174):
sys/netinet/tcp_output.c: revision 1.178
Avoid stack overflow when SACK and TCP_SIGNATURE are both present. Thanks
to Jonathan Looney for pointing this out.
|
| 1.174.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.174.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.174.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.175.6.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.175.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.176.2.5 | 24-Jul-2015 |
martin | Pull up following revision(s) (requested by matt in ticket #886):
sys/netinet/tcp_output.c: revision 1.184
sys/netinet/tcp_input.c: revision 1.343
If we are sending a window probe and there's unacked data in the
socket, make sure at least the persist timer is running.
Make sure that snd_win doesn't go negative.
|
| 1.176.2.4 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by he in ticket #530):
sys/netinet/tcp_output.c: revision 1.180
sys/netinet/tcp_input.c: revision 1.336
sys/netinet/tcp_usrreq.c: revision 1.203
share/man/man4/tcp.4: revision 1.30
sys/netinet/tcp.h: revision 1.31
sys/netinet/tcp_subr.c: revision 1.258
sys/netinet/tcp_var.h: revision 1.176
sys/netinet/tcp_var.h: revision 1.177
sys/sys/param.h: bump revision
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.176.2.3 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.176.2.2 | 26-Oct-2014 |
martin | Pull up following revision(s) (requested by christos in ticket #157):
sys/netinet/tcp_output.c: revision 1.178
Avoid stack overflow when SACK and TCP_SIGNATURE are both present. Thanks
to Jonathan Looney for pointing this out.
|
| 1.176.2.1 | 24-Oct-2014 |
martin | Pull up following revision(s) (requested by hikaru in ticket #154):
sys/netinet/tcp_output.c: revision 1.177
Fix wrong condition checking TSO capability.
ipsec_used is not necessary condition.
IPsec outbound policy will not be checked when ipsec_used is false.
|
| 1.179.2.6 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.179.2.5 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.179.2.4 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.179.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.179.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.179.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.186.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.186.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.194.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.196.2.1 | 21-Oct-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #300):
crypto/dist/ipsec-tools/src/setkey/parse.y: 1.19
crypto/dist/ipsec-tools/src/setkey/token.l: 1.20
distrib/sets/lists/tests/mi: 1.754, 1.757, 1.759
doc/TODO.smpnet: 1.12-1.13
sys/net/pfkeyv2.h: 1.32
sys/net/raw_cb.c: 1.23-1.24, 1.28
sys/net/raw_cb.h: 1.28
sys/net/raw_usrreq.c: 1.57-1.58
sys/net/rtsock.c: 1.228-1.229
sys/netinet/in_proto.c: 1.125
sys/netinet/ip_input.c: 1.359-1.361
sys/netinet/tcp_input.c: 1.359-1.360
sys/netinet/tcp_output.c: 1.197
sys/netinet/tcp_var.h: 1.178
sys/netinet6/icmp6.c: 1.213
sys/netinet6/in6_proto.c: 1.119
sys/netinet6/ip6_forward.c: 1.88
sys/netinet6/ip6_input.c: 1.181-1.182
sys/netinet6/ip6_output.c: 1.193
sys/netinet6/ip6protosw.h: 1.26
sys/netipsec/ipsec.c: 1.100-1.122
sys/netipsec/ipsec.h: 1.51-1.61
sys/netipsec/ipsec6.h: 1.18-1.20
sys/netipsec/ipsec_input.c: 1.44-1.51
sys/netipsec/ipsec_netbsd.c: 1.41-1.45
sys/netipsec/ipsec_output.c: 1.49-1.64
sys/netipsec/ipsec_private.h: 1.5
sys/netipsec/key.c: 1.164-1.234
sys/netipsec/key.h: 1.20-1.32
sys/netipsec/key_debug.c: 1.18-1.21
sys/netipsec/key_debug.h: 1.9
sys/netipsec/keydb.h: 1.16-1.20
sys/netipsec/keysock.c: 1.59-1.62
sys/netipsec/keysock.h: 1.10
sys/netipsec/xform.h: 1.9-1.12
sys/netipsec/xform_ah.c: 1.55-1.74
sys/netipsec/xform_esp.c: 1.56-1.72
sys/netipsec/xform_ipcomp.c: 1.39-1.53
sys/netipsec/xform_ipip.c: 1.50-1.54
sys/netipsec/xform_tcp.c: 1.12-1.16
sys/rump/librump/rumpkern/Makefile.rumpkern: 1.170
sys/rump/librump/rumpnet/net_stub.c: 1.27
sys/sys/protosw.h: 1.67-1.68
tests/net/carp/t_basic.sh: 1.7
tests/net/if_gif/t_gif.sh: 1.11
tests/net/if_l2tp/t_l2tp.sh: 1.3
tests/net/ipsec/Makefile: 1.7-1.9
tests/net/ipsec/algorithms.sh: 1.5
tests/net/ipsec/common.sh: 1.4-1.6
tests/net/ipsec/t_ipsec_ah_keys.sh: 1.2
tests/net/ipsec/t_ipsec_esp_keys.sh: 1.2
tests/net/ipsec/t_ipsec_gif.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_l2tp.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_misc.sh: 1.8-1.18
tests/net/ipsec/t_ipsec_sockopt.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tcp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_transport.sh: 1.5-1.6
tests/net/ipsec/t_ipsec_tunnel.sh: 1.9
tests/net/ipsec/t_ipsec_tunnel_ipcomp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tunnel_odd.sh: 1.3
tests/net/mcast/t_mcast.sh: 1.6
tests/net/net/t_ipaddress.sh: 1.11
tests/net/net_common.sh: 1.20
tests/net/npf/t_npf.sh: 1.3
tests/net/route/t_flags.sh: 1.20
tests/net/route/t_flags6.sh: 1.16
usr.bin/netstat/fast_ipsec.c: 1.22
Do m_pullup before mtod
It may fix panicks of some tests on anita/sparc and anita/GuruPlug.
---
KNF
---
Enable DEBUG for babylon5
---
Apply C99-style struct initialization to xformsw
---
Tweak outputs of netstat -s for IPsec
- Get rid of "Fast"
- Use ipsec and ipsec6 for titles to clarify protocol
- Indent outputs of sub protocols
Original outputs were organized like this:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
New outputs are organized like this:
ipsec:
ah:
esp:
ipip:
ipcomp:
ipsec6:
ah:
esp:
ipip:
ipcomp:
---
Add test cases for IPComp
---
Simplify IPSEC_OSTAT macro (NFC)
---
KNF; replace leading whitespaces with hard tabs
---
Introduce and use SADB_SASTATE_USABLE_P
---
KNF
---
Add update command for testing
Updating an SA (SADB_UPDATE) requires that a process issuing
SADB_UPDATE is the same as a process issued SADB_ADD (or SADB_GETSPI).
This means that update command must be used with add command in a
configuration of setkey. This usage is normally meaningless but
useful for testing (and debugging) purposes.
---
Add test cases for updating SA/SP
The tests require newly-added udpate command of setkey.
---
PR/52346: Frank Kardel: Fix checksumming for NAT-T
See XXX for improvements.
---
Remove codes for PACKET_TAG_IPSEC_IN_CRYPTO_DONE
It seems that PACKET_TAG_IPSEC_IN_CRYPTO_DONE is for network adapters
that have IPsec accelerators; a driver sets the mtag to a packet
when its device has already encrypted the packet.
Unfortunately no driver implements such offload features for long
years and seems unlikely to implement them soon. (Note that neither
FreeBSD nor Linux doesn't have such drivers.) Let's remove related
(unused) codes and simplify the IPsec code.
---
Fix usages of sadb_msg_errno
---
Avoid updating sav directly
On SADB_UPDATE a target sav was updated directly, which was unsafe.
Instead allocate another sav, copy variables of the old sav to
the new one and replace the old one with the new one.
---
Simplify; we can assume sav->tdb_xform cannot be NULL while it's valid
---
Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
---
Use explicit_memset to surely zero-clear key_auth and key_enc
---
Make sure to clear keys on error paths of key_setsaval
---
Add missing KEY_FREESAV
---
Make sure a sav is inserted to a sah list after its initialization completes
---
Remove unnecessary zero-clearing codes from key_setsaval
key_setsaval is now used only for a newly-allocated sav. (It was
used to reset variables of an existing sav.)
---
Correct wrong assumption of sav->refcnt in key_delsah
A sav in a list is basically not to be sav->refcnt == 0. And also
KEY_FREESAV assumes sav->refcnt > 0.
---
Let key_getsavbyspi take a reference of a returning sav
---
Use time_mono_to_wall (NFC)
---
Separate sending message routine (NFC)
---
Simplify; remove unnecessary zero-clears
key_freesaval is used only when a target sav is being destroyed.
---
Omit NULL checks for sav->lft_c
sav->lft_c can be NULL only when initializing or destroying sav.
---
Omit unnecessary NULL checks for sav->sah
---
Omit unnecessary check of sav->state
key_allocsa_policy picks a sav of either MATURE or DYING so we
don't need to check its state again.
---
Simplify; omit unnecessary saidx passing
- ipsec_nextisr returns a saidx but no caller uses it
- key_checkrequest is passed a saidx but it can be gotton by
another argument (isr)
---
Fix splx isn't called on some error paths
---
Fix header size calculation of esp where sav is NULL
---
Fix header size calculation of ah in the case sav is NULL
This fix was also needed for esp.
---
Pass sav directly to opencrypto callback
In a callback, use a passed sav as-is by default and look up a sav
only if the passed sav is dead.
---
Avoid examining freshness of sav on packet processing
If a sav list is sorted (by lft_c->sadb_lifetime_addtime) in advance,
we don't need to examine each sav and also don't need to delete one
on the fly and send up a message. Fortunately every sav lists are sorted
as we need.
Added key_validate_savlist validates that each sav list is surely sorted
(run only if DEBUG because it's not cheap).
---
Add test cases for SAs with different SPIs
---
Prepare to stop using isr->sav
isr is a shared resource and using isr->sav as a temporal storage
for each packet processing is racy. And also having a reference from
isr to sav makes the lifetime of sav non-deterministic; such a reference
is removed when a packet is processed and isr->sav is overwritten by
new one. Let's have a sav locally for each packet processing instead of
using shared isr->sav.
However this change doesn't stop using isr->sav yet because there are
some users of isr->sav. isr->sav will be removed after the users find
a way to not use isr->sav.
---
Fix wrong argument handling
---
fix printf format.
---
Don't validate sav lists of LARVAL or DEAD states
We don't sort the lists so the validation will always fail.
Fix PR kern/52405
---
Make sure to sort the list when changing the state by key_sa_chgstate
---
Rename key_allocsa_policy to key_lookup_sa_bysaidx
---
Separate test files
---
Calculate ah_max_authsize on initialization as well as esp_max_ivlen
---
Remove m_tag_find(PACKET_TAG_IPSEC_PENDING_TDB) because nobody sets the tag
---
Restore a comment removed in previous
The comment is valid for the below code.
---
Make tests more stable
sleep command seems to wait longer than expected on anita so
use polling to wait for a state change.
---
Add tests that explicitly delete SAs instead of waiting for expirations
---
Remove invalid M_AUTHIPDGM check on ESP isr->sav
M_AUTHIPDGM flag is set to a mbuf in ah_input_cb. An sav of ESP can
have AH authentication as sav->tdb_authalgxform. However, in that
case esp_input and esp_input_cb are used to do ESP decryption and
AH authentication and M_AUTHIPDGM never be set to a mbuf. So
checking M_AUTHIPDGM of a mbuf on isr->sav of ESP is meaningless.
---
Look up sav instead of relying on unstable sp->req->sav
This code is executed only in an error path so an additional lookup
doesn't matter.
---
Correct a comment
---
Don't release sav if calling crypto_dispatch again
---
Remove extra KEY_FREESAV from ipsec_process_done
It should be done by the caller.
---
Don't bother the case of crp->crp_buf == NULL in callbacks
---
Hold a reference to an SP during opencrypto processing
An SP has a list of isr (ipsecrequest) that represents a sequence
of IPsec encryption/authentication processing. One isr corresponds
to one opencrypto processing. The lifetime of an isr follows its SP.
We pass an isr to a callback function of opencrypto to continue
to a next encryption/authentication processing. However nobody
guaranteed that the isr wasn't freed, i.e., its SP wasn't destroyed.
In order to avoid such unexpected destruction of isr, hold a reference
to its SP during opencrypto processing.
---
Don't make SAs expired on tests that delete SAs explicitly
---
Fix a debug message
---
Dedup error paths (NFC)
---
Use pool to allocate tdb_crypto
For ESP and AH, we need to allocate an extra variable space in addition
to struct tdb_crypto. The fixed size of pool items may be larger than
an actual requisite size of a buffer, but still the performance
improvement by replacing malloc with pool wins.
---
Don't use unstable isr->sav for header size calculations
We may need to optimize to not look up sav here for users that
don't need to know an exact size of headers (e.g., TCP segmemt size
caclulation).
---
Don't use sp->req->sav when handling NAT-T ESP fragmentation
In order to do this we need to look up a sav however an additional
look-up degrades performance. A sav is later looked up in
ipsec4_process_packet so delay the fragmentation check until then
to avoid an extra look-up.
---
Don't use key_lookup_sp that depends on unstable sp->req->sav
It provided a fast look-up of SP. We will provide an alternative
method in the future (after basic MP-ification finishes).
---
Stop setting isr->sav on looking up sav in key_checkrequest
---
Remove ipsecrequest#sav
---
Stop setting mtag of PACKET_TAG_IPSEC_IN_DONE because there is no users anymore
---
Skip ipsec_spi_*_*_preferred_new_timeout when running on qemu
Probably due to PR 43997
---
Add localcount to rump kernels
---
Remove unused macro
---
Fix key_getcomb_setlifetime
The fix adjusts a soft limit to be 80% of a corresponding hard limit.
I'm not sure the fix is really correct though, at least the original
code is wrong. A passed comb is zero-cleared before calling
key_getcomb_setlifetime, so
comb->sadb_comb_soft_addtime = comb->sadb_comb_soft_addtime * 80 / 100;
is meaningless.
---
Provide and apply key_sp_refcnt (NFC)
It simplifies further changes.
---
Fix indentation
Pointed out by knakahara@
---
Use pslist(9) for sptree
---
Don't acquire global locks for IPsec if NET_MPSAFE
Note that the change is just to make testing easy and IPsec isn't MP-safe yet.
---
Let PF_KEY socks hold their own lock instead of softnet_lock
Operations on SAD and SPD are executed via PF_KEY socks. The operations
include deletions of SAs and SPs that will use synchronization mechanisms
such as pserialize_perform to wait for references to SAs and SPs to be
released. It is known that using such mechanisms with holding softnet_lock
causes a dead lock. We should avoid the situation.
---
Make IPsec SPD MP-safe
We use localcount(9), not psref(9), to make the sptree and secpolicy (SP)
entries MP-safe because SPs need to be referenced over opencrypto
processing that executes a callback in a different context.
SPs on sockets aren't managed by the sptree and can be destroyed in softint.
localcount_drain cannot be used in softint so we delay the destruction of
such SPs to a thread context. To do so, a list to manage such SPs is added
(key_socksplist) and key_timehandler_spd deletes dead SPs in the list.
For more details please read the locking notes in key.c.
Proposed on tech-kern@ and tech-net@
---
Fix updating ipsec_used
- key_update_used wasn't called in key_api_spddelete2 and key_api_spdflush
- key_update_used wasn't called if an SP had been added/deleted but
a reply to userland failed
---
Fix updating ipsec_used; turn on when SPs on sockets are added
---
Add missing IPsec policy checks to icmp6_rip6_input
icmp6_rip6_input is quite similar to rip6_input and the same checks exist
in rip6_input.
---
Add test cases for setsockopt(IP_IPSEC_POLICY)
---
Don't use KEY_NEWSP for dummy SP entries
By the change KEY_NEWSP is now not called from softint anymore
and we can use kmem_zalloc with KM_SLEEP for KEY_NEWSP.
---
Comment out unused functions
---
Add test cases that there are SPs but no relevant SAs
---
Don't allow sav->lft_c to be NULL
lft_c of an sav that was created by SADB_GETSPI could be NULL.
---
Clean up clunky eval strings
- Remove unnecessary \ at EOL
- This allows to omit ; too
- Remove unnecessary quotes for arguments of atf_set
- Don't expand $DEBUG in eval
- We expect it's expanded on execution
Suggested by kre@
---
Remove unnecessary KEY_FREESAV in an error path
sav should be freed (unreferenced) by the caller.
---
Use pslist(9) for sahtree
---
Use pslist(9) for sah->savtree
---
Rename local variable newsah to sah
It may not be new.
---
MP-ify SAD slightly
- Introduce key_sa_mtx and use it for some list operations
- Use pserialize for some list iterations
---
Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
---
Remove out-of-date log output
Pointed out by riastradh@
---
Use KDASSERT instead of KASSERT for mutex_ownable
Because mutex_ownable is too heavy to run in a fast path
even for DIAGNOSTIC + LOCKDEBUG.
Suggested by riastradh@
---
Assemble global lists and related locks into cache lines (NFCI)
Also rename variable names from *tree to *list because they are
just lists, not trees.
Suggested by riastradh@
---
Move locking notes
---
Update the locking notes
- Add locking order
- Add locking notes for misc lists such as reglist
- Mention pserialize, key_sp_ref and key_sp_unref on SP operations
Requested by riastradh@
---
Describe constraints of key_sp_ref and key_sp_unref
Requested by riastradh@
---
Hold key_sad.lock on SAVLIST_WRITER_INSERT_TAIL
---
Add __read_mostly to key_psz
Suggested by riastradh@
---
Tweak wording (pserialize critical section => pserialize read section)
Suggested by riastradh@
---
Add missing mutex_exit
---
Fix setkey -D -P outputs
The outputs were tweaked (by me), but I forgot updating libipsec
in my local ATF environment...
---
MP-ify SAD (key_sad.sahlist and sah entries)
localcount(9) is used to protect key_sad.sahlist and sah entries
as well as SPD (and will be used for SAD sav).
Please read the locking notes of SAD for more details.
---
Introduce key_sa_refcnt and replace sav->refcnt with it (NFC)
---
Destroy sav only in the loop for DEAD sav
---
Fix KASSERT(solocked(sb->sb_so)) failure in sbappendaddr that is called eventually from key_sendup_mbuf
If key_sendup_mbuf isn't passed a socket, the assertion fails.
Originally in this case sb->sb_so was softnet_lock and callers
held softnet_lock so the assertion was magically satisfied.
Now sb->sb_so is key_so_mtx and also softnet_lock isn't always
held by callers so the assertion can fail.
Fix it by holding key_so_mtx if key_sendup_mbuf isn't passed a socket.
Reported by knakahara@
Tested by knakahara@ and ozaki-r@
---
Fix locking notes of SAD
---
Fix deadlock between key_sendup_mbuf called from key_acquire and localcount_drain
If we call key_sendup_mbuf from key_acquire that is called on packet
processing, a deadlock can happen like this:
- At key_acquire, a reference to an SP (and an SA) is held
- key_sendup_mbuf will try to take key_so_mtx
- Some other thread may try to localcount_drain to the SP with
holding key_so_mtx in say key_api_spdflush
- In this case localcount_drain never return because key_sendup_mbuf
that has stuck on key_so_mtx never release a reference to the SP
Fix the deadlock by deferring key_sendup_mbuf to the timer
(key_timehandler).
---
Fix that prev isn't cleared on retry
---
Limit the number of mbufs queued for deferred key_sendup_mbuf
It's easy to be queued hundreds of mbufs on the list under heavy
network load.
---
MP-ify SAD (savlist)
localcount(9) is used to protect savlist of sah. The basic design is
similar to MP-ifications of SPD and SAD sahlist. Please read the
locking notes of SAD for more details.
---
Simplify ipsec_reinject_ipstack (NFC)
---
Add per-CPU rtcache to ipsec_reinject_ipstack
It reduces route lookups and also reduces rtcache lock contentions
when NET_MPSAFE is enabled.
---
Use pool_cache(9) instead of pool(9) for tdb_crypto objects
The change improves network throughput especially on multi-core systems.
---
Update
ipsec(4), opencrypto(9) and vlan(4) are now MP-safe.
---
Write known issues on scalability
---
Share a global dummy SP between PCBs
It's never be changed so it can be pre-allocated and shared safely between PCBs.
---
Fix race condition on the rawcb list shared by rtsock and keysock
keysock now protects itself by its own mutex, which means that
the rawcb list is protected by two different mutexes (keysock's one
and softnet_lock for rtsock), of course it's useless.
Fix the situation by having a discrete rawcb list for each.
---
Use a dedicated mutex for rt_rawcb instead of softnet_lock if NET_MPSAFE
---
fix localcount leak in sav. fixed by ozaki-r@n.o.
I commit on behalf of him.
---
remove unnecessary comment.
---
Fix deadlock between pserialize_perform and localcount_drain
A typical ussage of localcount_drain looks like this:
mutex_enter(&mtx);
item = remove_from_list();
pserialize_perform(psz);
localcount_drain(&item->localcount, &cv, &mtx);
mutex_exit(&mtx);
This sequence can cause a deadlock which happens for example on the following
situation:
- Thread A calls localcount_drain which calls xc_broadcast after releasing
a specified mutex
- Thread B enters the sequence and calls pserialize_perform with holding
the mutex while pserialize_perform also calls xc_broadcast
- Thread C (xc_thread) that calls an xcall callback of localcount_drain tries
to hold the mutex
xc_broadcast of thread B doesn't start until xc_broadcast of thread A
finishes, which is a feature of xcall(9). This means that pserialize_perform
never complete until xc_broadcast of thread A finishes. On the other hand,
thread C that is a callee of xc_broadcast of thread A sticks on the mutex.
Finally the threads block each other (A blocks B, B blocks C and C blocks A).
A possible fix is to serialize executions of the above sequence by another
mutex, but adding another mutex makes the code complex, so fix the deadlock
by another way; the fix is to release the mutex before pserialize_perform
and instead use a condvar to prevent pserialize_perform from being called
simultaneously.
Note that the deadlock has happened only if NET_MPSAFE is enabled.
---
Add missing ifdef NET_MPSAFE
---
Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
---
- sanitize key debugging so that we don't print extra newlines or unassociated
debugging messages.
- remove unused functions and make internal ones static
- print information in one line per message
---
humanize printing of ip addresses
---
cast reduction, NFC.
---
Fix typo in comment
---
Pull out ipsec_fill_saidx_bymbuf (NFC)
---
Don't abuse key_checkrequest just for looking up sav
It does more than expected for example key_acquire.
---
Fix SP is broken on transport mode
isr->saidx was modified accidentally in ipsec_nextisr.
Reported by christos@
Helped investigations by christos@ and knakahara@
---
Constify isr at many places (NFC)
---
Include socketvar.h for softnet_lock
---
Fix buffer length for ipsec_logsastr
|
| 1.198.2.6 | 18-Jan-2019 |
pgoyette | Synch with HEAD
|
| 1.198.2.5 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.198.2.4 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.198.2.3 | 07-Apr-2018 |
pgoyette | Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
|
| 1.198.2.2 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.198.2.1 | 15-Mar-2018 |
pgoyette | Synch with HEAD
|
| 1.208.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.208.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.218.2.1 | 21-Sep-2023 |
martin | Pull up following revision(s) (requested by bouyer in ticket #377):
sys/netinet/tcp_output.c: revision 1.219
Handle EHOSTDOWN the same way as EHOSTUNREACH and ENETDOWN for established
connections. Avoid premature end of tcp connection with "Host is down" error
in case of transient link-layer failure.
Discussed and patch proposed in
http://mail-index.netbsd.org/tech-net/2023/09/11/msg008610.html
and followups.
|
| 1.220.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.6 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.5 | 17-Feb-2021 |
christos | - pass the alignment instead of the mask (as Roy asked and to match the
other macro)
- use alignof to determine that alignment and CTASSERT what we expect
- remove unused macros
|
| 1.4 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.3 | 28-Apr-2008 |
martin | branches: 1.3.4; 1.3.102;
Remove clause 3 and 4 from TNF licenses
|
| 1.2 | 23-Apr-2008 |
thorpej | branches: 1.2.2;
Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.1 | 12-Apr-2008 |
thorpej | branches: 1.1.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.1.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.2.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.3.102.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.3.4.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.3.4.1 | 28-Apr-2008 |
mjf | file tcp_private.h was added on branch mjf-devfs2 on 2008-06-02 13:24:25 +0000
|
| 1.36 | 18-May-2018 |
maxv | IP6_EXTHDR_GET -> M_REGION_GET, no functional change.
|
| 1.35 | 03-May-2018 |
maxv | Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.34 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.33 | 13-Dec-2016 |
ozaki-r | branches: 1.33.14;
Remove unnecessary inclusions of nd6.h
|
| 1.32 | 24-Aug-2015 |
pooka | branches: 1.32.2;
sprinkle _KERNEL_OPT
|
| 1.31 | 31-Mar-2015 |
ozaki-r | Remove unnecessary opt_ipsec.h inclusions
|
| 1.30 | 10-Nov-2014 |
maxv | branches: 1.30.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.29 | 12-Nov-2013 |
kefren | branches: 1.29.4;
* implement TCP CUBIC congestion control algorithm
* move tcp_sack_newack bits inside reno and newreno_fast_retransmit_newack
* notify ECN peer about cwnd shrink in [new]reno_slow_retransmit
Based on the patch proposed on tech-net@ on Nov 7 with minor improvments:
* adapt wmax for no-fast convergence case
* correct cbrt calculation for big window sizes (>750KB)
|
| 1.28 | 30-Jan-2012 |
matt | branches: 1.28.6; 1.28.10;
Use proper ANSI prototypes for foo() -> foo(void)
Caught when compiling with -Wold-style-definition
|
| 1.27 | 17-Jul-2011 |
joerg | branches: 1.27.2; 1.27.6;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.26 | 14-Apr-2011 |
yamt | - comments
- whitespace
|
| 1.25 | 27-May-2009 |
pooka | branches: 1.25.4; 1.25.6;
POOL_INIT -> pool_init
|
| 1.24 | 28-Apr-2008 |
martin | branches: 1.24.14;
Remove clause 3 and 4 from TNF licenses
|
| 1.23 | 12-Mar-2007 |
ad | branches: 1.23.34; 1.23.36; 1.23.38;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.22 | 21-Oct-2006 |
yamt | branches: 1.22.4; 1.22.8;
add sack_dump(), a function to dump sack holes, if defined(DDB).
|
| 1.21 | 21-Oct-2006 |
yamt | - constify.
- make tcp_dooptions and tcpipqent_pool static.
|
| 1.20 | 20-Oct-2006 |
reinoud | Fix alignment problems causing regular panics in tpc_sack_option on
NetBSD/alpha and NetBSD/sparc. This fixes PR#34751.
The problem most likely started to show in gcc4 and is caused by the use of
a casting to an uint32_t pointer that is later copied from using memcpy.
Gcc detects the copying of 4 bytes from an uint32_t pointer and decides to
just replace it with an aligned copy causing the trap.
Fix provided by Izumi Tsutsui and ok'd by Martin.
|
| 1.19 | 07-Oct-2006 |
yamt | - make sackhole_pool static.
- unify duplicated hole allocation and accounting code.
(no functional changes.)
|
| 1.18 | 07-Oct-2006 |
yamt | revert tcp_sack.c rev.1.15 because it's unnecessary.
all callers of these functions are at splsoftnet already:
tcp_sack_option
tcp_input ok
tcp_del_sackholes
tcp_input ok
tcp_free_sackholes
tcp_close ok
tcp_timer_rexmt ok
tcp_timer_2msl ok
|
| 1.17 | 07-Oct-2006 |
yamt | tcp_sack_output: whitespace.
|
| 1.16 | 07-Oct-2006 |
yamt | tcp_del_sackholes: whitespace.
|
| 1.15 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.14 | 11-Dec-2005 |
christos | branches: 1.14.4; 1.14.20; 1.14.22;
merge ktrace-lwp.
|
| 1.13 | 08-May-2005 |
yamt | branches: 1.13.2;
tcp_sack_option: ignore sack beyond snd_max.
|
| 1.12 | 05-Apr-2005 |
kurahone | branches: 1.12.2;
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.11 | 18-Mar-2005 |
kurahone | branches: 1.11.2;
TCP/SACK changes from FreeBSD.
Ignore the SACK option if
* The packet is not an ACK.
* The ACK is outside of snd_una -> snd_max
|
| 1.10 | 16-Mar-2005 |
yamt | branches: 1.10.2;
simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]),
build it directly from the segment list when needed.
|
| 1.9 | 16-Mar-2005 |
yamt | - use full sized segments unless we actually have SACKs to send.
- avoid TSO duplicate D-SACK.
- send SACKs regardless of TF_ACKNOW.
- don't clear rcv_sack_num when transmitting.
discussed on tech-net@.
|
| 1.8 | 08-Mar-2005 |
yamt | tcp_sack_option: handle the case that the right-most sack'ed block is expanded.
a fix from Noritoshi Demizu (FreeBSD PR/78226) via Kentaro A. Kurahone.
|
| 1.7 | 07-Mar-2005 |
yamt | tcp_sack_option: fix the cases that some sack blocks go into a hole.
|
| 1.6 | 07-Mar-2005 |
yamt | tcp_sack_option: fix a typo(?), which can cause to ignore valid blocks.
|
| 1.5 | 07-Mar-2005 |
yamt | tcp_sack_option: the max number of sack blocks in a packet is 4, not 3.
|
| 1.4 | 06-Mar-2005 |
yamt | - unwrap short lines.
- remove unneeded parenthesis.
- whitespace.
|
| 1.3 | 06-Mar-2005 |
yamt | don't assume alignment of sack options.
|
| 1.2 | 06-Mar-2005 |
yamt | wrap long lines.
|
| 1.1 | 28-Feb-2005 |
jonathan | branches: 1.1.2;
Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.1.2.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.1.2.4 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.1.2.3 | 08-Mar-2005 |
skrll | Sync with HEAD.
|
| 1.1.2.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.1.2.1 | 28-Feb-2005 |
skrll | file tcp_sack.c was added on branch ktrace-lwp on 2005-03-04 16:53:29 +0000
|
| 1.10.2.4 | 11-Nov-2006 |
bouyer | Pull up following revision(s) (requested by reinoud in ticket #1561):
sys/netinet/tcp_sack.c: revision 1.20
Fix alignment problems causing regular panics in tpc_sack_option on
NetBSD/alpha and NetBSD/sparc. This fixes PR#34751.
The problem most likely started to show in gcc4 and is caused by the use of
a casting to an uint32_t pointer that is later copied from using memcpy.
Gcc detects the copying of 4 bytes from an uint32_t pointer and decides to
just replace it with an aligned copy causing the trap.
Fix provided by Izumi Tsutsui and ok'd by Martin.
|
| 1.10.2.3 | 11-May-2005 |
tron | branches: 1.10.2.3.2; 1.10.2.3.4;
Pull up revision 1.13 (requested by yamt in ticket #293):
tcp_sack_option: ignore sack beyond snd_max.
|
| 1.10.2.2 | 06-May-2005 |
tron | Pull up revision 1.12 (requested by kurahone in ticket #199):
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.10.2.1 | 06-May-2005 |
tron | Pull up revision 1.11 (requested by yamt in ticket #250):
TCP/SACK changes from FreeBSD.
Ignore the SACK option if
* The packet is not an ACK.
* The ACK is outside of snd_una -> snd_max
|
| 1.10.2.3.4.1 | 11-Nov-2006 |
bouyer | Pull up following revision(s) (requested by reinoud in ticket #1561):
sys/netinet/tcp_sack.c: revision 1.20
Fix alignment problems causing regular panics in tpc_sack_option on
NetBSD/alpha and NetBSD/sparc. This fixes PR#34751.
The problem most likely started to show in gcc4 and is caused by the use of
a casting to an uint32_t pointer that is later copied from using memcpy.
Gcc detects the copying of 4 bytes from an uint32_t pointer and decides to
just replace it with an aligned copy causing the trap.
Fix provided by Izumi Tsutsui and ok'd by Martin.
|
| 1.10.2.3.2.1 | 11-Nov-2006 |
bouyer | Pull up following revision(s) (requested by reinoud in ticket #1561):
sys/netinet/tcp_sack.c: revision 1.20
Fix alignment problems causing regular panics in tpc_sack_option on
NetBSD/alpha and NetBSD/sparc. This fixes PR#34751.
The problem most likely started to show in gcc4 and is caused by the use of
a casting to an uint32_t pointer that is later copied from using memcpy.
Gcc detects the copying of 4 bytes from an uint32_t pointer and decides to
just replace it with an aligned copy causing the trap.
Fix provided by Izumi Tsutsui and ok'd by Martin.
|
| 1.11.2.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.11.2.1 | 18-Mar-2005 |
yamt | file tcp_sack.c was added on branch yamt-km on 2005-03-19 08:36:38 +0000
|
| 1.12.2.2 | 29-Apr-2005 |
kent | sync with -current
|
| 1.12.2.1 | 05-Apr-2005 |
kent | file tcp_sack.c was added on branch kent-audio2 on 2005-04-29 11:29:34 +0000
|
| 1.13.2.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.13.2.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.14.22.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.14.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.14.4.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.22.8.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.22.4.1 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.23.38.2 | 20-Jun-2009 |
yamt | sync with head
|
| 1.23.38.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.23.36.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.23.34.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.24.14.1 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.25.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.25.4.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.27.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.27.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.27.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.28.10.1 | 18-May-2014 |
rmind | sync with head
|
| 1.28.6.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.28.6.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.29.4.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.30.2.3 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.30.2.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.30.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.32.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.33.14.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.33.14.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.17 | 25-Jul-2014 |
ryo | fix some case of reference to uninitialized tp->snd_fack.
This bug causes dropping FIN mistekenly.
pointed out in PR/48283 by YASUOKA Masahiko, thanks!
|
| 1.16 | 10-Dec-2005 |
elad | branches: 1.16.120; 1.16.136;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.15 | 05-Apr-2005 |
kurahone | branches: 1.15.2;
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.14 | 16-Feb-2005 |
briggs | branches: 1.14.4;
Initialize snd_high as part of tcp_sendseqinit().
From Kentaro A. Kurahone.
|
| 1.13 | 07-Aug-2003 |
agc | branches: 1.13.8; 1.13.10;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.12 | 04-Oct-1998 |
matt | branches: 1.12.46;
Adapt the NEWRENO changes from the UCSB diffs of BSDI 3.0's TCP
to NetBSD. Ignore the SACK & FACK stuff for now.
|
| 1.11 | 04-Sep-1998 |
mycroft | Make the randomized part of the ISS 24 bits.
|
| 1.10 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.9 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.8 | 13-Oct-1997 |
explorer | o Make usage of /dev/random dependant on
pseudo-device rnd # /dev/random and in-kernel generator
in config files.
o Add declaration to all architectures.
o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include
that this code is derived in part from Ted Tyso's linux code.
|
| 1.7 | 10-Oct-1997 |
explorer | Add hooks to use the kernel random system to generate TCP sequence numbers.
|
| 1.6 | 26-Mar-1995 |
jtc | branches: 1.6.14;
KERNEL -> _KERNEL
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.6.14.1 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.12.46.6 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.12.46.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.12.46.4 | 17-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.12.46.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.12.46.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.12.46.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.13.10.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.13.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.14.4.1 | 06-May-2005 |
tron | Pull up revision 1.15 (requested by kurahone in ticket #199):
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.15.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.16.136.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.16.120.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.298 | 26-Feb-2025 |
andvar | Fix typos in comments, mainly s/calcurate/calculate/.
|
| 1.297 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.296 | 04-Nov-2022 |
ozaki-r | branches: 1.296.8;
inpcb: rename functions to in6pcb_*
|
| 1.295 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.294 | 31-Oct-2022 |
ozaki-r | tcp: fix wrong logic in tcp_drop
Pointed out by mlelstv@
|
| 1.293 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.292 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.291 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.290 | 27-Jun-2022 |
knakahara | Remove confusable comment.
The comment was added in tcp_subr.c:r1.124 (2002/03/15).
tcp_drain() is called from softint context only, now.
|
| 1.289 | 31-Jul-2021 |
andvar | s/threshhold/threshold
|
| 1.288 | 09-Mar-2021 |
christos | branches: 1.288.4;
Move the offset addition in one place and mask the random generated value
to make sure that the isn is monotonic.
|
| 1.287 | 08-Mar-2021 |
christos | Remove the unused "addin" argument (it was always 0) and go back using
a random iss by default (instead of rfc1948)
|
| 1.286 | 08-Mar-2021 |
christos | Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
|
| 1.285 | 07-Mar-2021 |
christos | netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
|
| 1.284 | 12-Jun-2020 |
roy | branches: 1.284.2;
Remove in-kernel handling of Router Advertisements
This is much better handled by a user-land tool.
Proposed on tech-net here:
https://mail-index.netbsd.org/tech-net/2020/04/22/msg007766.html
Note that the ioctl SIOCGIFINFO_IN6 no longer sets flags. That now
needs to be done using the pre-existing SIOCSIFINFO_FLAGS ioctl.
Compat is fully provided where it makes sense, but trying to turn on
RA handling will obviously throw an error as it no longer exists.
Note that if you use IPv6 temporary addresses, this now needs to be
turned on in dhcpcd.conf(5) rather than in sysctl.conf(5).
|
| 1.283 | 06-Aug-2019 |
riastradh | Clamp tcp timer quantities to reasonable ranges.
Reported-by: syzbot+259675123340bf46a6de@syzkaller.appspotmail.com
|
| 1.282 | 27-Dec-2018 |
maxv | branches: 1.282.4;
Remove unused arguments.
|
| 1.281 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.280 | 23-May-2018 |
maxv | branches: 1.280.2;
Add XXX.
|
| 1.279 | 03-May-2018 |
maxv | Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.278 | 18-Apr-2018 |
maxv | Remove unused netipsec/xform.h includes.
|
| 1.277 | 18-Apr-2018 |
maxv | Remove misleading comments.
|
| 1.276 | 29-Mar-2018 |
maxv | Remove TCPREASS_DEBUG. It was introduced 20 years ago when the reassembler
was being developed, but it's irrelevant today. Makes the code clearer.
|
| 1.275 | 29-Mar-2018 |
maxv | Misc changes; no real functional change.
|
| 1.274 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.273 | 26-Feb-2018 |
maxv | branches: 1.273.2;
Dedup: merge ipsec4_hdrsiz and ipsec6_hdrsiz into ipsec_hdrsiz.
ok ozaki-r@
|
| 1.272 | 19-Jan-2018 |
ozaki-r | Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
|
| 1.271 | 29-Jul-2017 |
maxv | Forgot to commit this file yesterday.
|
| 1.270 | 03-Mar-2017 |
ozaki-r | branches: 1.270.6;
Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.269 | 02-Jan-2017 |
christos | branches: 1.269.2;
Fix TCP signature code:
1. pack options more tightly instead of being generous with no/op
2. put TCP_SIGNATURE option before SACK
3. fix computation of options length, by deferring it
XXX: Really we should move the options setting code in one place instead
of having two copies one for input and one for output.
XXX: tcp_optlen/tcp_hdrsiz need to be fixed; they were wrong before too.
|
| 1.268 | 08-Dec-2016 |
ozaki-r | Add rtcache_unref to release points of rtentry stemming from rtcache
In the MP-safe world, a rtentry stemming from a rtcache can be freed at any
points. So we need to protect rtentries somehow say by reference couting or
passive references. Regardless of the method, we need to call some release
function of a rtentry after using it.
The change adds a new function rtcache_unref to release a rtentry. At this
point, this function does nothing because for now we don't add a reference
to a rtentry when we get one from a rtcache. We will add something useful
in a further commit.
This change is a part of changes for MP-safe routing table. It is separated
to avoid one big change that makes difficult to debug by bisecting.
|
| 1.267 | 09-Nov-2016 |
ozaki-r | Cleanup/KNF tcp6_mtudisc
No functional change.
|
| 1.266 | 10-Jun-2016 |
ozaki-r | branches: 1.266.2;
Introduce m_set_rcvif and m_reset_rcvif
The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.
No functional change.
|
| 1.265 | 15-Feb-2016 |
rtr | Reduce code duplication.
Split creation of IPv4-Mapped IPv6 addresses into its own function
and use it.
No functional change intended. As posted to tech-net@
|
| 1.264 | 07-Sep-2015 |
ozaki-r | Refactor tcp_mtudisc
No functional change.
|
| 1.263 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.262 | 19-May-2015 |
kefren | Use RUN_ONCE to initialize iss secret. Suggested by riastradh@
|
| 1.261 | 16-May-2015 |
kefren | Don't overexpose tcp_iss_secret and don't bother compute it unless
RFC1948 compliance is activated
|
| 1.260 | 27-Apr-2015 |
ozaki-r | Introduce in6_selecthlim_rt to consolidate an idiom for rt->rt_ifp
It consolidates a scattered routine:
(rt = rtcache_validate(&in6p->in6p_route)) != NULL ? rt->rt_ifp : NULL
|
| 1.259 | 13-Apr-2015 |
riastradh | cprng_strong(kern_cprng, ...) never blocks, pass 0 for flags.
FASYNC was wrong anyway! It's FNONBLOCK.
|
| 1.258 | 14-Feb-2015 |
he | Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
|
| 1.257 | 10-Nov-2014 |
maxv | branches: 1.257.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.256 | 05-Sep-2014 |
matt | Don't use C++ keyword (template) as variable.
|
| 1.255 | 16-Mar-2014 |
dholland | branches: 1.255.4;
Change (mostly mechanically) every cdevsw/bdevsw I can find to use
designated initializers.
I have not built every extant kernel so I have probably broken at
least one build; however I've also found and fixed some wrong
cdevsw/bdevsw entries so even if so I think we come out ahead.
|
| 1.254 | 02-Jan-2014 |
pooka | need atomic.h, from uwe
|
| 1.253 | 02-Jan-2014 |
pooka | Allow kernels compiled with INET+INET6 to be booted as IPv4-only or IPv6-only.
|
| 1.252 | 23-Nov-2013 |
christos | convert from CIRCLEQ to TAILQ.
|
| 1.251 | 12-Nov-2013 |
kefren | * implement TCP CUBIC congestion control algorithm
* move tcp_sack_newack bits inside reno and newreno_fast_retransmit_newack
* notify ECN peer about cwnd shrink in [new]reno_slow_retransmit
Based on the patch proposed on tech-net@ on Nov 7 with minor improvments:
* adapt wmax for no-fast convergence case
* correct cbrt calculation for big window sizes (>750KB)
|
| 1.250 | 05-Jun-2013 |
christos | branches: 1.250.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.249 | 10-Apr-2013 |
christos | Limit the tcp initial window setting to 10, leaving it by default to 4
and simplifying the code in process. Per draft-ietf-initcwnd-08.txt.
|
| 1.248 | 08-Sep-2012 |
msaitoh | branches: 1.248.2;
Fix a bug that kmem_alloc() is called from the interrupt context.
|
| 1.247 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.246 | 31-Dec-2011 |
christos | branches: 1.246.2;
- fix offsetof usage, and redundant defines
- kill pointer casts to 0
|
| 1.245 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.244 | 17-Dec-2011 |
tls | Separate /dev/random pseudodevice implemenation from kernel entropy pool
implementation. Rewrite pseudodevice code to use cprng_strong(9).
The new pseudodevice is cloning, so each caller gets bits from a stream
generated with its own key. Users of /dev/urandom get their generators
keyed on a "best effort" basis -- the kernel will rekey generators
whenever the entropy pool hits the high water mark -- while users of
/dev/random get their generators rekeyed every time key-length bits
are output.
The underlying cprng_strong API can use AES-256 or AES-128, but we use
AES-128 because of concerns about related-key attacks on AES-256. This
improves performance (and reduces entropy pool depletion) significantly
for users of /dev/urandom but does cause users of /dev/random to rekey
twice as often.
Also fixes various bugs (including some missing locking and a reseed-counter
overflow in the CTR_DRBG code) found while testing this.
For long reads, this generator is approximately 20 times as fast as the
old generator (dd with bs=64K yields 53MB/sec on 2Ghz Core2 instead of
2.5MB/sec) and also uses a separate mutex per instance so concurrency
is greatly improved. For reads of typical key sizes for modern
cryptosystems (16-32 bytes) performance is about the same as the old
code: a little better for 32 bytes, a little worse for 16 bytes.
|
| 1.243 | 19-Nov-2011 |
tls | branches: 1.243.2;
First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>. This change includes
the following:
An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are
fixed. Some effort is made to accumulate entropy more quickly at
boot time.
A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.
The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.
An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c). This
is based on Greg Rose's implementation from Qualcomm.
A new random stream generator, nist_ctr_drbg, is provided. It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim. This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.
An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness. The arc4random/arc4randbytes API is deprecated for
in-kernel use. It is replaced by "cprng_strong". The current
cprng_fast implementation wraps the existing arc4random
implementation. The current cprng_strong implementation wraps the
new CTR_DRBG implementation. Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.
In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data(). Performance is expected to improve.
The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.
The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot. There is
approximately a 3/10000 chance of a false positive from these
tests. Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.
A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32). A new kernel will require a new rndctl.
The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.
Manual pages for the new kernel interfaces are forthcoming.
|
| 1.242 | 31-Oct-2011 |
yamt | branches: 1.242.2;
tcp_drain: grab softnet_lock where appropriate
|
| 1.241 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.240 | 03-May-2011 |
dyoung | *_drain() routines may be called with locks held, so instead of doing
any work in *_drain(), set a drain-needed flag. Do the work in the
fasttimo handler.
Contributed by Coyote Point Systems, Inc.
|
| 1.239 | 20-Apr-2011 |
gdt | Rewrite comments about TCP RTO calculations.
Long ago, the storage representations of srtt and rttvar were changed
from the 4.4BSD scheme, and the comments are out of sync with the
code. This commit rewrites most of the comments that explain the RTO
calculations, and points out some issues in the code.
Joint work with Bev Schwartz of BBN (original analysis and comments),
but I have rewritten and extended them, so errors are mine.
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073. Approved for Public
Release, Distribution Unlimited
|
| 1.238 | 16-Sep-2009 |
pooka | branches: 1.238.4; 1.238.6;
Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.237 | 27-May-2009 |
pooka | POOL_INIT -> pool_init
|
| 1.236 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.235 | 18-Mar-2009 |
cegger | bcmp -> memcmp
|
| 1.234 | 29-Jan-2009 |
pooka | branches: 1.234.2;
stinkset purge: POOL_INIT -> pool_init
also, make the syncache pool static in scope
|
| 1.233 | 13-Oct-2008 |
pooka | branches: 1.233.2;
POOL_INIT -> pool_init
|
| 1.232 | 10-Oct-2008 |
ad | tcp_close: rearrange sequence of events slightly to make this atomic.
It was possible for a half-destroyed tcpcb to be visble, as softnet_lock
was being dropped.
|
| 1.231 | 02-May-2008 |
ad | branches: 1.231.2; 1.231.6;
PR kern/38497 Out of memory allocating ksiginfo
Work around: don't acquire softnet_lock in protocol drain routines.
|
| 1.230 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.229 | 26-Apr-2008 |
yamt | branches: 1.229.2;
tcp_init: don't forget to allocate tcpstat_percpu.
|
| 1.228 | 24-Apr-2008 |
ad | Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.227 | 12-Apr-2008 |
thorpej | branches: 1.227.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.226 | 08-Apr-2008 |
thorpej | Change TCP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old tcpstat structure; old netstat
binaries will continue to work properly.
|
| 1.225 | 27-Mar-2008 |
cube | - Make sure we send a reasonable fragment size when IPSEC is configured.
Otherwise we end up sending a dubious "0" whenever we cannot find a
proper association for the packet.
- Reset sack_newdata along with snd_nxt to avoid improper integer
arithmetics that lead to sending data from an incorrect place in the
stream, making it appear as corrupted.
Patch by Michael Van Elst, based on an analysis by Michael for the IPSEC
stuff and I for the SACK issue.
|
| 1.224 | 29-Feb-2008 |
matt | Rework tcp congctl selection code so that the congctl entries can be const.
Don't access tcp_congctl stuff outside of tcp_congctl.c, use routines to
update t_congctl. This code is slightly now more complicated.
|
| 1.223 | 27-Feb-2008 |
matt | Convert stragglers to ansi definitions from old-style definitons.
Remember that func() is not ansi, func(void) is.
|
| 1.222 | 05-Feb-2008 |
yamt | branches: 1.222.2; 1.222.6;
- start tcp timestamp from 1 instead of 0.
- add a comment to explain why:
+ * We start with 1, because 0 doesn't work with linux, which
+ * considers timestamp 0 in a SYN packet as a bug and disables
+ * timestamps.
|
| 1.221 | 14-Jan-2008 |
dyoung | Use rtcache_validate() instead of rtcache_getrt(). Shorten staircase
in in_losing().
|
| 1.220 | 20-Dec-2007 |
martin | A few missing ifdefs to make non-INET6 kernels build again.
|
| 1.219 | 20-Dec-2007 |
dyoung | Poison struct route->ro_rt uses in the kernel by changing the name
to _ro_rt. Use rtcache_getrt() to access a route cache's struct
rtentry *.
Introduce struct ifnet->if_dl that always points at the interface
identifier/link-layer address. Make code that treated the first
ifaddr on struct ifnet->if_addrlist as the interface address use
if_dl, instead.
Remove stale debugging code from net/route.c. Move the rtflush()
code into rtcache_clear() and delete rtflush(). Delete rtalloc(),
because nothing uses it any more.
Make ND6_HINT an inline, lowercase subroutine, nd6_hint.
I've done my best to convert IP Filter, the ISO stack, and the
AppleTalk stack to rtcache_getrt(). They compile, but I have not
tested them. I have given the changes to PF, GRE, IPv4 and IPv6
stacks a lot of exercise.
|
| 1.218 | 02-Aug-2007 |
rmind | branches: 1.218.4; 1.218.10; 1.218.12; 1.218.16; 1.218.20;
TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.217 | 19-Jul-2007 |
dyoung | Take steps to hide the radix_node implementation of the forwarding table
from the forwarding table's users:
Introduce rt_walktree() for walking the routing table and
applying a function to each rtentry. Replace most
rn_walktree() calls with it.
Use rt_getkey()/rt_setkey() to get/set a route's destination.
Keep a pointer to the sockaddr key in the rtentry, so that
rtentry users do not have to grovel in the radix_node for
the key.
Add a RTM_GET method to rtrequest. Use that instead of
radix_node lookups in, e.g., carp(4).
Add sys/net/link_proto.c, which supplies sockaddr routines for
link-layer socket addresses (sockaddr_dl).
Cosmetic:
Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH,
et cetera. Use NULL instead of 0 for null pointers. Use
__arraycount(). Reduce gratuitous parenthesization.
Stop using variadic arguments for rip6_output(), it is
unnecessary.
Remove the unnecessary rtentry member rt_genmask and the
code to maintain it, since nothing actually used it.
Make rt_maskedcopy() easier to read by using meaningful variable
names.
Extract a subroutine intern_netmask() for looking up a netmask in
the masks table.
Start converting backslash-ridden IPv6 macros in
sys/netinet6/in6_var.h into inline subroutines that one
can read without special eyeglasses.
One functional change: when the kernel serves an RTM_GET, RTM_LOCK,
or RTM_CHANGE request, it applies the netmask (if supplied) to a
destination before searching for it in the forwarding table.
I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove
the unlawful radix_node knowledge.
Apart from the changes to carp(4), netiso, ATM, and strip(4), I
have run the changes on three nodes in my wireless routing testbed,
which involves IPv4 + IPv6 dynamic routing acrobatics, and it's
working beautifully so far.
|
| 1.216 | 09-Jul-2007 |
ad | branches: 1.216.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.215 | 20-Jun-2007 |
christos | - per socket keepalive settings
- settable connection establishment timeout
|
| 1.214 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.213 | 12-Mar-2007 |
ad | branches: 1.213.2;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.212 | 04-Mar-2007 |
christos | branches: 1.212.2;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.211 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.210 | 10-Feb-2007 |
degroote | branches: 1.210.2;
Commit my SoC work
Add ipv6 support for fast_ipsec
Note that currently, packet with extensions headers are not correctly
supported
Change the ipcomp logic
|
| 1.209 | 06-Dec-2006 |
yamt | add some more tcp mowners.
|
| 1.208 | 16-Nov-2006 |
christos | branches: 1.208.2; 1.208.4;
__unused removal on arguments; approved by core.
|
| 1.207 | 19-Oct-2006 |
yamt | implement RFC3465 appropriate byte counting.
from Kentaro A. Kurahone, with minor adjustments by me.
the ack prediction part of the original patch was omitted because
it's a separate change. reviewed by Rui Paulo.
|
| 1.206 | 17-Oct-2006 |
dogcow | now that we have -Wno-unused-parameter, back out all the tremendously ugly
code to gratuitously access said parameters.
|
| 1.205 | 13-Oct-2006 |
dogcow | more unused variable fallout.
|
| 1.204 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.203 | 10-Oct-2006 |
dogcow | change the MOWNER_INIT define to take two args; fix extant struct mowner
decls to use it. Makes options MBUFTRACE compile again and not whinge about
missing structure declarations. (Also makes initialization consistent.)
|
| 1.202 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.201 | 07-Oct-2006 |
yamt | revert tcp_close part of tcp_subr.c rev.1.200 because it's unnecessary.
all callers of tcp_close are at splsoftnet already:
tcp_close
tcp_input ok
tcp_disconnect
tcp_usrreq ok
tcp_usrclosed
tcp_usrreq ok
tcp_disconnect
tcp_timer_2msl ok
tcp_drop
tcp_usrreq
tcp_disconnect
tcp_timer_rexmt ok
tcp_timer_persist ok
tcp_timer_keep ok
tcp_input
syn_cache_get
tcp_input
|
| 1.200 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.199 | 05-Sep-2006 |
rpaulo | branches: 1.199.2; 1.199.4;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.198 | 15-Apr-2006 |
christos | Coverity CID 1149: Add KASSERT before deref.
|
| 1.197 | 15-Apr-2006 |
christos | Coverity CID 1148: Add KASSERT before deref.
|
| 1.196 | 11-Dec-2005 |
christos | branches: 1.196.4; 1.196.6; 1.196.8; 1.196.10; 1.196.12;
merge ktrace-lwp.
|
| 1.195 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.194 | 10-Aug-2005 |
yamt | device independent part of ipv6 rx checksum offloading.
|
| 1.193 | 20-Jul-2005 |
he | Make this build without INET6.
|
| 1.192 | 19-Jul-2005 |
christos | Implement PMTU checks from:
http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.
From OpenBSD.
|
| 1.191 | 29-May-2005 |
christos | branches: 1.191.2;
- add const
- remove bogus casts
- avoid nested variables
|
| 1.190 | 18-Apr-2005 |
yamt | fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.189 | 05-Apr-2005 |
kurahone | Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.188 | 29-Mar-2005 |
yamt | protect tcpipqent with splvm.
|
| 1.187 | 16-Mar-2005 |
yamt | branches: 1.187.2;
simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]),
build it directly from the segment list when needed.
|
| 1.186 | 16-Mar-2005 |
yamt | - use full sized segments unless we actually have SACKs to send.
- avoid TSO duplicate D-SACK.
- send SACKs regardless of TF_ACKNOW.
- don't clear rcv_sack_num when transmitting.
discussed on tech-net@.
|
| 1.185 | 09-Mar-2005 |
simonb | s/quence/quench/.
|
| 1.184 | 09-Mar-2005 |
simonb | Add an extra `i' to notifes/notifed.
|
| 1.183 | 28-Feb-2005 |
jonathan | Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.182 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.181 | 16-Feb-2005 |
briggs | Initialize t_partialacks in the tcpcb template.
From Kentaro A. Kurahone.
|
| 1.180 | 12-Feb-2005 |
heas | ntohs->htons for ip6 plen (payload length).
It is not technically necessary to set plen here, since ip6_output() starts
off by calculating it, but leaving it keeps it consistent with other code.
|
| 1.179 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.178 | 02-Feb-2005 |
perry | de-__P -- will ANSIfy .c files later.
|
| 1.177 | 03-Jan-2005 |
heas | branches: 1.177.2; 1.177.4;
In tcp_respond(), clear the m_pkthdr.csum_flags that was inherited from the
received packet so that the checksum is not performed twice. Also,
tcp_respond() does not fill-in the m_pkthdr.csum_data, so a h/w checksum may
have the wrong offset.
OK from Jason Thorpe.
|
| 1.176 | 19-Dec-2004 |
christos | yamt's changes seem to fix all the checksumming issues. Turn the loopback
checksums back off so we can make sure that everything works.
|
| 1.175 | 17-Dec-2004 |
christos | Turn checksumming on loopback back on until we fix the bugs in it.
Connect over tcp on the loopback is broken:
4729 amq 0.000007 CALL connect(4,0x804f2a0,0x1c)
4729 amq 75.007420 RET connect -1 errno 60 Connection timed out
|
| 1.174 | 15-Dec-2004 |
thorpej | Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.173 | 15-Sep-2004 |
yamt | fix ipqent pool corruption problems. make tcp reass code use
its own pool of ipqent rather than sharing it with ip reass code.
PR/24782.
|
| 1.172 | 18-May-2004 |
itojun | fix MD5 signature support to actually validate inbound signature, and
drop packet if fails.
|
| 1.171 | 01-May-2004 |
matt | Use EVCNT_ATTACH_STATIC{,2}
|
| 1.170 | 26-Apr-2004 |
itojun | zero-clear ip6?pseudo before use
|
| 1.169 | 26-Apr-2004 |
itojun | declare ip6_hdr_pseudo (for kernel only) and use it for TCP MD5 signature
|
| 1.168 | 26-Apr-2004 |
itojun | sync comment with reality
|
| 1.167 | 26-Apr-2004 |
itojun | make TCP MD5 signature work with KAME IPSEC (#define IPSEC).
support IPv6 if KAME IPSEC (RFC is not explicit about how we make data stream
for checksum with IPv6, but i'm pretty sure using normal pseudo-header is the
right thing).
XXX
current TCP MD5 signature code has giant flaw:
it does not validate signature on input (can't believe it! what is the point?)
|
| 1.166 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.165 | 25-Apr-2004 |
simonb | Initialise (most) pools from a link set instead of explicit calls
to pool_init. Untouched pools are ones that either in arch-specific
code, or aren't initialiased during initial system startup.
Convert struct session, ucred and lockf to pools.
|
| 1.164 | 22-Apr-2004 |
tls | Change the default state of two tunables; bring our TCP a little bit
closer to normal behaviour for the current century.
New Reno is now on by default (which is really the only reasonable
choice, since we don't do SACK); instead of an initial window of 1
for non-local nets, we now use Sally Floyd's magic 4K rule.
|
| 1.163 | 20-Apr-2004 |
itojun | - respond to RST by ACK, as suggested in NISCC recommendation
- rate-limit ACKs against RSTs and SYNs
|
| 1.162 | 17-Apr-2004 |
christos | adjust to the sbreserve prototype change.
|
| 1.161 | 05-Apr-2004 |
christos | PR/22551: Invoking tcpcb's get erroneously free'd resulting in to_ticks <= 0
assertion. Approved by he.
|
| 1.160 | 07-Jan-2004 |
matt | branches: 1.160.2;
When accepting a peer's MSS, never let it drop below 256 (SLIP + TCP will
be the lowest MSS we should ever enounter).
|
| 1.159 | 27-Oct-2003 |
thorpej | - Change callout_setfunc() to require that the callout handle is already
initialized. Update the txp(4) to compensate.
- Statically initialize the TCP timer callout handles in the tcpcb
template. We still use callout_setfunc(), but that call is now much
less expensive. Add a comment that the compiler is likely to unroll
the loop (so don't sweat that it's there).
|
| 1.158 | 25-Oct-2003 |
christos | initialize off
|
| 1.157 | 22-Oct-2003 |
thorpej | Oops, a little to aggressive in the previous patch; TCP_TIMER_INIT()
still needs to be in tcp_newtcpcb(), for now. Pointed out by enami.
|
| 1.156 | 22-Oct-2003 |
thorpej | Rather than zeroing a tcpcb structure and filling in all the fields
individually, create a tcpcb template pre-initialized (and pre-zero'd)
with the static and mostly-static tcpcb parameters. The template is
now copied into the new tcpcb, which zeros and initializes most of the
tcpcb in one pass. The template is kept up-to-date as TCP sysctl
variables are changed.
Combined with the previous sb_max change, TCP socket creation is now
25% faster.
|
| 1.155 | 21-Oct-2003 |
thorpej | Add event counters that measure FAST_MBSEARCH.
|
| 1.154 | 25-Sep-2003 |
mycroft | Fix glaring errors in recent changes.
|
| 1.153 | 08-Sep-2003 |
itojun | initialize ip_hl for ipsec policy lookup. PR kern/22715
|
| 1.152 | 06-Sep-2003 |
itojun | randomize IPv4/v6 fragment ID and IPv6 flowlabel. avoids predictability
of these fields. ip_id.c is from openbsd. ip6_id.c is adapted by kame.
|
| 1.151 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.150 | 22-Aug-2003 |
itojun | tp could be null in tcp_respond()
|
| 1.149 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.148 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.147 | 22-Aug-2003 |
jonathan | Replace the set_socket() method of passing an extra struct socket*
argument to ip6_output() with a new explicit struct in6pcb* argument.
(The underlying socket can be obtained via in6pcb->inp6_socket.)
In preparation for fast-ipsec. Reviewed by itojun.
|
| 1.146 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.145 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.144 | 20-Jul-2003 |
he | As a temporary workaround, apply the fix from PR#20390, thereby
cooperating with the callout code in working around the race
condition caused by the TCP code's use of the callout facility.
Instead of unconditionally releasing memory in tcp_close() and
SYN_CACHE_PUT(), check whether any of the related callout handlers
are about to be invoked (but have not yet done callout_ack()), and
if so, just mark the associated data structure (tcpcb or syn cache
entry) as "dead", and test for this (and release storage) in the
callout handler functions.
|
| 1.143 | 03-Jul-2003 |
ragge | Make it possible to set TCP_INIT_WIN and TCP_INIT_WIN_LOCAL in the config
file as options.
|
| 1.142 | 29-Jun-2003 |
fvdl | branches: 1.142.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.141 | 29-Jun-2003 |
ragge | Add code to remember where in the send queue of mbufs the last packet was
sent from. This change avoid a linear search through all mbufs when using
large TCP windows, and therefore permit high-speed connections on long
distances.
Tested on a 1 Gigabit connection between Lule� and San Francisco, a distance
of about 15000km. With TCP windows of just over 20 Mbytes it could keep up
with 950Mbit/s.
After discussions with Matt Thomas and Jason Thorpe.
|
| 1.140 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.139 | 01-Mar-2003 |
thorpej | Allow TCP connections to hosts on a local network to use a larger
slow start initial window. Default this larger initial window to
4 packets, allowing it to be adjusted with net.inet.tcp.init_win_local.
|
| 1.138 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.137 | 24-Nov-2002 |
scw | Quell an uninitialised variable warning.
|
| 1.136 | 22-Oct-2002 |
lukem | fix typo in previous: s/tip/top/
|
| 1.135 | 22-Oct-2002 |
simonb | Micro-optimisation: don't check if the high bit is set and then mask it
off - just mask it off anyways. Saves a branch 50% of the time.
|
| 1.134 | 25-Sep-2002 |
itojun | minor KNF
|
| 1.133 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.132 | 01-Jul-2002 |
itojun | check AF_INET6 socketes when IPv4 "too big" messages arrive.
PR 17448
|
| 1.131 | 09-Jun-2002 |
itojun | whitespace
|
| 1.130 | 29-May-2002 |
itojun | attach nd_ifinfo structure into if_afdata.
split IPv6 link MTU (advertised by RA) from real link MTU.
sync with kame
|
| 1.129 | 28-May-2002 |
itojun | use arc4random() on tcp iss generation
|
| 1.128 | 26-May-2002 |
itojun | path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.127 | 12-May-2002 |
matt | branches: 1.127.2; 1.127.4;
Eliminate commons.
|
| 1.126 | 07-May-2002 |
matt | Change struct ipqe to use TAILQ's instead of LIST's (primarily for TCP's
benefit currently). Rework tcp_reass code to optimize the 4 most likely causes
of out-of-order packets: first OoO pkt, next OoO pkt in seq, OoO pkt is part
of new chuck of OoO packets, and the OoO pkt fills the first hole. Add evcnts
to instrument tcp_reass (enabled by the options TCP_REASS_COUNTERS). This is
part 1/2 of tcp_reass changes.
|
| 1.125 | 27-Apr-2002 |
thorpej | * Instrument tcp_build_datapkt().
* Remove the code that allocates a cluster if the packet would
fit in one; it totally defeats doing references to M_EXT mbufs
in the socket buffer. This drastically reduces the number of
data copies in the tcp_output() path for applications which use
large writes. Kudos to Matt Thomas for pointing me in the right
direction.
|
| 1.124 | 15-Mar-2002 |
itojun | have tcp6_drain
|
| 1.123 | 08-Mar-2002 |
thorpej | Pool deals fairly well with physical memory shortage, but it doesn't
deal with shortages of the VM maps where the backing pages are mapped
(usually kmem_map). Try to deal with this:
* Group all information about the backend allocator for a pool in a
separate structure. The pool references this structure, rather than
the individual fields.
* Change the pool_init() API accordingly, and adjust all callers.
* Link all pools using the same backend allocator on a list.
* The backend allocator is responsible for waiting for physical memory
to become available, but will still fail if it cannot callocate KVA
space for the pages. If this happens, carefully drain all pools using
the same backend allocator, so that some KVA space can be freed.
* Change pool_reclaim() to indicate if it actually succeeded in freeing
some pages, and use that information to make draining easier and more
efficient.
* Get rid of PR_URGENT. There was only one use of it, and it could be
dealt with by the caller.
From art@openbsd.org.
|
| 1.122 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.121 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.120 | 04-Nov-2001 |
matt | Change a few variable/tables to const since they are read-only.
|
| 1.119 | 11-Sep-2001 |
thorpej | branches: 1.119.2;
Use callouts for SYN cache timers, rather than traversing time queues
in tcp_slowtimo().
|
| 1.118 | 10-Sep-2001 |
thorpej | Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.
|
| 1.117 | 10-Sep-2001 |
thorpej | Initialize TCP timer variables in a new function, tcp_timer_init().
|
| 1.116 | 10-Sep-2001 |
thorpej | Add explicit initialization of TCP timer state. A noop right now.
|
| 1.115 | 10-Sep-2001 |
thorpej | Use a callout for the delayed ACK timer, and delete tcp_fasttimo().
Expose the delayed ACK timer as net.inet.tcp.delack_ticks.
|
| 1.114 | 23-Jul-2001 |
itojun | branches: 1.114.2;
wrap IPv6 code by #ifdef INET6
|
| 1.113 | 23-Jul-2001 |
itojun | use in6_maxmtu, not in_maxmtu, for IPv6 mss computation
|
| 1.112 | 12-Jun-2001 |
wiz | branches: 1.112.2;
receive, not recieve
|
| 1.111 | 02-Jun-2001 |
thorpej | Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.110 | 24-May-2001 |
itojun | call icmp6_mtudisc_update(foo, 0) even if ICMPv6 messages are very short.
let icmp6 layer decide whether we take PMTUD routes or not.
|
| 1.109 | 21-Mar-2001 |
chs | make this compile without rnd.
|
| 1.108 | 20-Mar-2001 |
thorpej | Two changes, designed to make us even more resilient against TCP
ISS attacks (which we already fend off quite well).
1. First-cut implementation of RFC1948, Steve Bellovin's cryptographic
hash method of generating TCP ISS values. Note, this code is experimental
and disabled by default (experimental enough that I don't export the
variable via sysctl yet, either). There are a couple of issues I'd
like to discuss with Steve, so this code should only be used by people
who really know what they're doing.
2. Per a recent thread on Bugtraq, it's possible to determine a system's
uptime by snooping the RFC1323 TCP timestamp options sent by a host; in
4.4BSD, timestamps are created by incrementing the tcp_now variable
at 2 Hz; there's even a company out there that uses this to determine
web server uptime. According to Newsham's paper "The Problem With
Random Increments", while NetBSD's TCP ISS generation method is much
better than the "random increment" method used by FreeBSD and OpenBSD,
it is still theoretically possible to mount an attack against NetBSD's
method if the attacker knows how many times the tcp_iss_seq variable
has been incremented. By not leaking uptime information, we can make
that much harder to determine. So, we avoid the leak by giving each
TCP connection a timebase of 0.
|
| 1.107 | 11-Feb-2001 |
itojun | branches: 1.107.2;
pull latest kame pcbnotify code. synchronizes ICMPv6 path mtu discovery
behavior with other protocols (i.e. validation, use of hiwat/lowat).
|
| 1.106 | 24-Jan-2001 |
itojun | - record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.105 | 21-Dec-2000 |
itojun | fix call to in6_pcbnotify. s/EMSGSIZE/PRC_MSGSIZE/.
|
| 1.104 | 09-Dec-2000 |
itojun | update icmp6 too big validation. the change is necessary since pmtud is
mandatory for IPv6 (so we can't just validate by using connected pcb - we need
to allow traffic from unconnected pcb to do pmtud).
- if the traffic is validated by xx_ctlinput, allow up to "hiwat" pmtud
route entries.
- if the traffic was not validated by xx_ctlinput, allow up to "lowat" pmtud
route entries (there's upper limit, so bad guys cannot blow up our routing
table).
sync with kame
XXX need to think again about default hiwat/lowat value.
XXX victim selection to help starvation case
|
| 1.103 | 29-Oct-2000 |
itojun | cleanup tcp_drop
|
| 1.102 | 29-Oct-2000 |
itojun | process IPv4 tcp RST packet right. reported by thorpej.
|
| 1.101 | 19-Oct-2000 |
itojun | remove #ifdef TCP6. it is not likely for us to bring in sys/netinet6/tcp6*.c
(separate TCP/IPv6 stack) into netbsd-current.
|
| 1.100 | 19-Oct-2000 |
itojun | memcpy -> bcopy, for sync with kame tree
|
| 1.99 | 18-Oct-2000 |
itojun | verify ICMPv6 too big messages based on TCP pcbs, and/or IPsec SA.
TODO: udp6, and sendto consideration. as pmtud is mandatory for IPv6,
it is rather important for us to support those cases.
TODO: more testing
TODO: kame sync
|
| 1.98 | 18-Oct-2000 |
thorpej | Restructure the Path MTU Discovery code somewhat to avoid
entering rtentry's for hosts we're not actually communicating
with.
Do this by invoking the ctlinput for the protocol, which is
responsible for validating the ICMP message:
* TCP -- Lookup the connection based on the address/port
pairs in the ICMP message.
* AH/ESP -- Lookup the SA based on the SPI in the ICMP message.
If validation succeeds, ctlinput is responsible for calling
icmp_mtudisc(). icmp_mtudisc() then invokes callbacks registered
by protocols (such as TCP) which want to take some sort of special
action when a path's MTU changes. For TCP, this is where we now
refresh cached routes and re-enter slow-start.
As a side-effect, this fixes the problem where TCP would not be
notified when a path's MTU changed if AH/ESP were being used.
XXX Note, this is only a fix for the IPv4 case. For the IPv6
XXX case, we need to wait for the KAME folks.
Reviewed by sommerfeld@netbsd.org and itojun@netbsd.org.
|
| 1.97 | 18-Oct-2000 |
itojun | move tcp syn cache parameters from in_proto.c to tcp_subr.c.
it makes more sense and helps INET6-only (INET-less) build.
|
| 1.96 | 17-Oct-2000 |
itojun | be more friendly with INET-less build.
XXX we need to do more to do a working INET-less build
|
| 1.95 | 17-Oct-2000 |
thorpej | Add an IP_MTUDISC flag to the flags that can be passed to
ip_output(). This flag, if set, causes ip_output() to set
DF in the IP header if the MTU in the route is not locked.
This allows a bunch of redundant code, which I was never
really all that happy about adding in the first place, to
be eliminated.
Inspired by a similar change made by provos@openbsd.org when
he integrated NetBSD's Path MTU Discovery code into OpenBSD.
|
| 1.94 | 13-Oct-2000 |
itojun | validate mbuf chain length on *_ctlinput. remote node may be able to
transmit a truncated icmp6 packet and panic the system. sync with kame.
|
| 1.93 | 19-Sep-2000 |
itojun | for t_template, allocate mbuf cluster only if really necessary.
this avoids too aggressive memory usage on heavy load web server, for example.
From: Kevin Lahey <kml@dotrocket.com>
release and reallocate t_template, if t_template->m_len changes.
(this happens if we connect to IPv4 mapped destination and then IPv6
destination, on a single AF_INET6 socket)
KAME 1.26 -> 1.28
|
| 1.92 | 30-Jun-2000 |
itojun | remove old mbuf assumption (ip header and tcp header are on the same mbuf).
this is for m_pulldown use. (sync with kame)
|
| 1.91 | 30-Mar-2000 |
augustss | branches: 1.91.4;
Remove register declarations.
|
| 1.90 | 30-Mar-2000 |
simonb | Delete redundant decl of zeroin6_addr, it's in <netinet6/in6_var.h>.
|
| 1.89 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.88 | 29-Feb-2000 |
itojun | ensure tcp window size does not overflow (16bit unsigned after window scale).
FreeBSD PR: 16914
|
| 1.87 | 06-Feb-2000 |
itojun | don't chase mbuf pointer when it is NULL.
|
| 1.86 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.85 | 15-Dec-1999 |
itojun | do not overwrite traffic class field when we write IPv6 version field.
|
| 1.84 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.83 | 12-Dec-1999 |
ragge | Avoid GCC complaints (under some circumstances).
|
| 1.82 | 08-Dec-1999 |
itojun | do not drop from IP header to tcp option until sbappend(), to reduce
requirement to mbuf chain.
part of KAME sync, committed separately for its (possible) impact.
|
| 1.81 | 23-Sep-1999 |
enami | branches: 1.81.2; 1.81.8;
Make this compile without INET6.
|
| 1.80 | 23-Sep-1999 |
itojun | cleanup and correct TCP MSS consideration with IPsec headers.
MSS advertisement must always be:
max(if mtu) - ip hdr siz - tcp hdr siz
We violated this in the previous code so it was fixed.
tcp_mss_to_advertise() now takes af (af on wire) as its argument,
to compute right ip hdr siz.
tcp_segsize() will take care of IPsec header size.
One thing I'm not really sure is how to handle IPsec header size in
*rxsegsizep (inbound segment size estimation).
The current code subtracts possible *outbound* IPsec size from *rxsegsizep,
hoping that the peer is using the same IPsec policy as me.
It may not be applicable, could TCP gulu please comment...
|
| 1.79 | 27-Aug-1999 |
itojun | fix tcp mss consideration on ipsec operation.
now tcp-over-ipsec should not experience fragmentation due to
addition of ipsec header.
From: proff@suburbia.net (Julian Assange)
|
| 1.78 | 25-Aug-1999 |
itojun | When listening socket goes away, remove assockated syn cache entires.
Stale syn cache entries are useless because none of them will be used
if there is no listening socket, as tcp_input looks up listening socket by
in_pcblookup*() before looking into syn cache.
This fixes race condition due to dangling socket pointer from syn cache
entries to listening socket (this was introduced when ipsec is merged in).
This should preserve currently implemented behavior (but not 4.4BSD
behavior prior to syn cache).
Tested in KAME repository before commit, but we'd better run some
regression tests.
|
| 1.77 | 25-Aug-1999 |
itojun | ctlinput handling must look at ip6_src, not ip6_dst.
(this makes path mtu handling wrong)
|
| 1.76 | 09-Aug-1999 |
itojun | return with doing nothing from xx_ctlinput(), when sa->sa_family
is not the expected one.
I see PRC_REDIRECT_HOST with sa->sa_family == AF_UNIX coming to
{tcp,udp}_ctlinput() when I use dhclient, and I feel like adding
more sanity checks, without logging - if we log it it is too noisy.
|
| 1.75 | 31-Jul-1999 |
itojun | sync with recent KAME.
- loosen ipsec restriction on packet diredtion.
- revise icmp6 redirect handling on IsRouter bit.
- tcp/udp notification processing (link-local address case)
- cosmetic fixes (better code share across *BSD).
|
| 1.74 | 23-Jul-1999 |
itojun | do not include unnecessary include files.
|
| 1.73 | 22-Jul-1999 |
itojun | - implement IPv6 pmtud, which is necessary for TCP6.
- fix memory leak on SO_DEBUG over TCP.
|
| 1.72 | 14-Jul-1999 |
itojun | Use proper ip protocol # field and tcp hdr on sending RST against SYN,
when ip header and tcp header are not adjacent to each other
(i.e. when ip6 options are attached).
To test this, try
telnet @::1@::1 port
toward a port without responding server. Prior to the fix, the kernel will
generate broken RST packet.
|
| 1.71 | 14-Jul-1999 |
drochner | make sending of keepalive messages work again:
-remove bogus sanity check involving an uninitialized variable
-correct mbuf cluster allocation
-(non-critical) remove redundant check in cleanup after error
|
| 1.70 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.69 | 02-Jul-1999 |
fvdl | Fix for -Wunitialized warnings broke compiles without INET6, refix.
|
| 1.68 | 02-Jul-1999 |
itojun | avoid "variable not initialized" warnings on some of the platforms.
|
| 1.67 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.66 | 28-Feb-1999 |
explorer | branches: 1.66.4; 1.66.6;
Don't mix in data just to stir the rnd pool. Extracting data will do that,
any network packets received might, too, so this is already taken care of.
|
| 1.65 | 26-Jan-1999 |
thorpej | Fix a slight error in previous. Rearrange some code in tcp_respond() so
that a DIAGNOSTIC check against the destination address is actually
checking the destination address. "oops."
|
| 1.64 | 20-Jan-1999 |
thorpej | Fix a problem pointed out by Charles Hannum; DF wasn't being set in
SYN,ACK packets during Path MTU Discovery. Fix tcp_respond() to do the
appropriate route lookup and set DF as appropriate.
Also, fixup similar code in tcp_output() to relookup the route if it
is down.
|
| 1.63 | 18-Dec-1998 |
thorpej | Add a lock around the TCPCB's sequence queue, to prevent tcp_drain()
from corrupting the queue if called from a device's interrupt context.
Similar in nature to the problem reported in PR #5684.
|
| 1.62 | 08-Oct-1998 |
thorpej | Use the pool allocator for ipqent structures.
|
| 1.61 | 07-Oct-1998 |
thorpej | Use the pool allocator for the tcpcb's TCP/IP header template.
|
| 1.60 | 06-Oct-1998 |
matt | Add a sysctl for newreno (default to off).
|
| 1.59 | 19-Sep-1998 |
mycroft | Always send a 0 window with a RST. Suggested by Darren Reed.
|
| 1.58 | 04-Sep-1998 |
mycroft | Fix a couple of bogons related to tcp_new_iss():
* Don't add tcp_iss_seq when creating a new ISS from TIME-WAIT state.
* Do the clock increment even when using the rnd device.
|
| 1.57 | 02-Aug-1998 |
thorpej | Use the pool allocator for tcpcbs.
|
| 1.56 | 17-Jul-1998 |
thorpej | Document that we are more conservative after doing MTU discovery than the
suggestion in draft-floyd-incr-init-win-03. Rather than scaling cwnd back
by the ratio of new segment size to old segment size, we perform a slow start
using the Initial Window, computed with the new segment size.
|
| 1.55 | 17-Jul-1998 |
thorpej | Clarify that we're using the Loss Window when we receive a source quench.
|
| 1.54 | 12-May-1998 |
kml | Changed initialization of peermss to ensure that it didn't have
the TCP and IP options lengths removed from it -- the IP options can
change over the course of a connection...
|
| 1.53 | 07-May-1998 |
kml | Change comments on tcp_mss_to_advertise to match actual arguments
|
| 1.52 | 07-May-1998 |
thorpej | Rework the syn cache code somewhat:
- Don't use home-grown queue manipulation. Use <sys/queue.h> instead. The
data structures are a little larger, but we are otherwise wasting the
memory chunk anyway (we're already a 64-byte malloc bucket).
- Fix a bug in the cache-is-full case: if the oldest element removed from
the first non-empty bucket was the only element in the bucket, the
bucket wouldn't be removed from the bucket cache, causing queue corruption
later.
- Optimize the syn cache timers by using PRT timers rather than home-grown
decrement-and-propagate timers.
This code is now a fair bit smaller, and significantly easier to read
and understand.
|
| 1.51 | 06-May-1998 |
thorpej | Use macros from tcp_timer.h to manipulate TCP timers, so that their
implementation can be changed easily.
|
| 1.50 | 03-May-1998 |
thorpej | Once again, move a declaration for the benefit of TUBA (grumble).
|
| 1.49 | 29-Apr-1998 |
matt | New TCP reassembly code. The new code reduces the memory needed by
out-of-order packets and builds the infrastructure needed for sending
SACK blocks (to be added shortly).
|
| 1.48 | 29-Apr-1998 |
thorpej | Make use of the work-arounds for ancient broken TCP peers run-time
conditional (tcp_compat_42). The kernel config option TCP_COMPAT_42
will still enable this by default, or disable this by default if the
option is not included (i.e. current behavior). This will be made a
sysctl soon.
|
| 1.47 | 13-Apr-1998 |
kml | Fix to ensure that the correct MSS is advertised for loopback
TCP connections by using the MTU of the interface. Also added
a knob, mss_ifmtu, to force all connections to use the MTU of
the interface to calculate the advertised MSS.
|
| 1.46 | 31-Mar-1998 |
thorpej | Fix a potential-congestion case in the larger initial congestion window
code, as clarified in the TCPIMPL WG meeting at IETF #41: If the SYN
(active open) or SYN,ACK (passive open) was retransmitted, the initial
congestion window for the first slow start of that connection must be
one segment.
|
| 1.45 | 28-Mar-1998 |
thorpej | Remove a comment in tcp_mss_to_advertise() that no longer applies.
|
| 1.44 | 24-Mar-1998 |
kml | Ensure that we take the IP option length into account when we calculate
the effective maximum send size for TCP. ip_optlen() and tcp_optlen()
should probably be inlined for efficiency.
|
| 1.43 | 19-Mar-1998 |
kml | Fix a retransmission bug introduced by the Brakmo and Peterson
RTO estimation changes. Under some circumstances it would return a value
of 0, while the old Van Jacobson RTO code would return a minimum of 3.
This would result in 12 retransmissions, each 1 second apart.
This takes care of those instances, and ensures that t_rttmin is
used everywhere as a lower bound.
|
| 1.42 | 17-Mar-1998 |
kml | Ensure that the TCP segment size reflects the size of TCP options
in the packet. This fixes a bug that was resulting in extra packets
in retransmissions (the second packet would be 12 bytes long,
reflecting the RFC1323 timestamp option size).
|
| 1.41 | 19-Feb-1998 |
thorpej | Update copyright (sigh, should have done this long ago).
|
| 1.40 | 30-Jan-1998 |
mellon | Take PCB off delayed ack queue before freeing.
|
| 1.39 | 12-Jan-1998 |
scottr | Use option header file for TCP_COMPAT_42
|
| 1.38 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.37 | 31-Dec-1997 |
thorpej | Implement a queue for delayed ACK processing. This queue is used in
tcp_fasttimo() in lieu of scanning all open TCP connections.
|
| 1.36 | 11-Dec-1997 |
thorpej | Implement an infrastructure to allow larger initial congestion windows.
The sysctl'able variable "tcp_init_win", when set to 0, selects an
auto-tuning algorithm for selecting the initial window, based on transmit
segment size, per discussion in the IETF tcpimpl working group.
Default initial window is still 1 segment, but will soon become 2 segments,
per discussion in tcpimpl.
|
| 1.35 | 10-Dec-1997 |
thorpej | Implement tcp_drain().
|
| 1.34 | 11-Nov-1997 |
kml | Remove an extraneous call to rtfree() in the path mtu discovery code;
this was causing negative reference counts on routes...
|
| 1.33 | 08-Nov-1997 |
kml | TCP MSS fixes to provide cleaner slow-start and recovery.
|
| 1.32 | 18-Oct-1997 |
kml | branches: 1.32.2;
change sysctl net.inet.icmp.mtudisc to net.inet.ip.mtudisc
|
| 1.31 | 17-Oct-1997 |
kml | Path MTU Discovery support. This is turned off by default.
Use sysctl -w net.inet.icmp.mtudisc=1 to turn on.
Still to come: path removal after some period, black hole detection
|
| 1.30 | 13-Oct-1997 |
explorer | o Make usage of /dev/random dependant on
pseudo-device rnd # /dev/random and in-kernel generator
in config files.
o Add declaration to all architectures.
o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include
that this code is derived in part from Ted Tyso's linux code.
|
| 1.29 | 10-Oct-1997 |
explorer | Add hooks to use the kernel random system to generate TCP sequence numbers.
|
| 1.28 | 22-Sep-1997 |
thorpej | Fix several annoyances related to MSS handling in BSD TCP:
- Don't overload t_maxseg. Previous behavior was to set it to the min
of the peer's advertised MSS, our advertised MSS, and tcp_mssdflt
(for non-local networks). This breaks PMTU discovery running on
either host. Instead, remember the MSS we advertise, and use it
as appropriate (in silly window avoidance).
- Per last bullet, split tcp_mss() into several functions for handling
MSS (ours and peer's), and performing various tasks when a connection
becomes ESTABLISHED.
- Introduce a new function, tcp_segsize(), which computes the max size
for every segment transmitted in tcp_output(). This will eventually
be used to hook in PMTU discovery.
|
| 1.27 | 23-Jul-1997 |
thorpej | branches: 1.27.2;
Pull SYN_cache_branch down into the main line.
|
| 1.26 | 24-Jun-1997 |
thorpej | Eliminate use of dtom() from the network code, allowing more flexible
use of mbuf external storage and increasing performance (by eliminating
an m_pullup() for clusters in the IP reassembly code).
Changes from Koji Imada <koji@math.human.nagoya-u.ac.jp>, in PR #3628
and #3480, with ever-so-slight integration changes by me.
|
| 1.25 | 10-Dec-1996 |
mycroft | branches: 1.25.8;
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.24 | 15-Sep-1996 |
mycroft | Hash unconnected PCBs.
|
| 1.23 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.22 | 13-Feb-1996 |
christos | branches: 1.22.4;
netinet prototypes
|
| 1.21 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.20 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.19 | 12-Jun-1995 |
mycroft | branches: 1.19.2;
Fix bogon in previous.
|
| 1.18 | 12-Jun-1995 |
mycroft | Change in_pcbnotify*() to take an errno value. Make inetctlerrmap[] an
array on ints, not u_chars.
|
| 1.17 | 12-Jun-1995 |
mycroft | Oops. Make source quench work again.
|
| 1.16 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.15 | 11-Jun-1995 |
mycroft | As suggested by Brakmo and Peterson:
* Don't add the extra 1/8 of the mss when ramping up the congestion window.
* Scale the RTT values slightly to adjust for rounding errors.
* Set the lower bound of the RTO to RTT+2.
|
| 1.14 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.13 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.12 | 14-Oct-1994 |
mycroft | Don't return received data to the user until the initial handshake is complete.
Also use TCPS_HAVEESTABLISHED() in a few other places.
|
| 1.11 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.10 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.9 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.8 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.7 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.6 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.3 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.2 | 12-Apr-1993 |
mycroft | Ignore forged ICMP_UNREACH with dport==0 and sport==0.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.19.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.22.4.1 | 10-Dec-1996 |
mycroft | From trunk:
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.25.8.4 | 28-Jun-1997 |
thorpej | KNF.
|
| 1.25.8.3 | 26-Jun-1997 |
thorpej | u_short -> u_int16_t.
|
| 1.25.8.2 | 26-Jun-1997 |
thorpej | Update from trunk.
|
| 1.25.8.1 | 14-May-1997 |
mellon | More of David Borman's SYN cache patches for 4.4BSD-lite2:
tcp_respond:
- return a error code if the reponse could not be sent, zero
otherwise (was void).
- generate SYN/ACK packet if tp==0 and SYN bit is set. Do
not adjust window in this case.
- if SYN bit is set, use ti_off as passed in rather than
setting it locally.
tcp_ctlinput:
- call syn_cache_unreach if in_pcbnotify fails.
|
| 1.27.2.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.27.2.1 | 29-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.32.2.8 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.32.2.7 | 05-May-1998 |
mycroft | Pull up 1.42, per request of kml.
|
| 1.32.2.6 | 05-May-1998 |
mycroft | Pull up 1.43, per request of kml.
|
| 1.32.2.5 | 07-Feb-1998 |
mellon | Pull up 1.40 (mellon)
|
| 1.32.2.4 | 29-Jan-1998 |
mellon | Fix botched merge
|
| 1.32.2.3 | 29-Jan-1998 |
mellon | Pull up 1.35-1.38 (thorpej)
|
| 1.32.2.2 | 12-Nov-1997 |
thorpej | Pull up from trunk: nuke extra rtfree()
|
| 1.32.2.1 | 08-Nov-1997 |
thorpej | Pull up from trunk: TCP MSS fixes to provide cleaner slow-start and recovery.
(kml)
|
| 1.66.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.66.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.66.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.66.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.66.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.81.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.81.2.6 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.81.2.5 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.81.2.4 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.81.2.3 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.81.2.2 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.81.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.91.4.5 | 09-Sep-2003 |
msaitoh | Pull up rev. 1.153 (requested by itojun in ticket #78):
Initialize ip_hl for ipsec policy lookup. Fixes PR 22715.
|
| 1.91.4.4 | 06-Apr-2001 |
he | Pull up revision 1.106 (via patch, requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.91.4.3 | 17-Oct-2000 |
tv | Pullup 1.94 [itojun]:
validate mbuf chain length on *_ctlinput. remote node may be able to
transmit a truncated icmp6 packet and panic the system. sync with kame.
|
| 1.91.4.2 | 19-Sep-2000 |
itojun | pullup 1.92 -> 1.93 (requested by thorpej)
for t_template, allocate mbuf cluster only if really necessary.
this avoids too aggressive memory usage on heavy load web server, for example.
From: Kevin Lahey <kml@dotrocket.com>
release and reallocate t_template, if t_template->m_len changes.
(this happens if we connect to IPv4 mapped destination and then IPv6
destination, on a single AF_INET6 socket)
KAME 1.26 -> 1.28
|
| 1.91.4.1 | 23-Jul-2000 |
itojun | pullup from main trunc (approved by releng-1-5)
remove old mbuf assumption (ip header and tcp header are on the same mbuf).
this is for m_pulldown use. (sync with kame)
1.108 -> 1.109 syssrc/sys/netinet/tcp_input.c
1.56 -> 1.57 syssrc/sys/netinet/tcp_output.c
1.91 -> 1.92 syssrc/sys/netinet/tcp_subr.c
|
| 1.107.2.13 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.107.2.12 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.107.2.11 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.107.2.10 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.107.2.9 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.107.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.107.2.7 | 04-May-2002 |
thorpej | Update from trunk.
|
| 1.107.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.107.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.107.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.107.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.107.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.107.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.112.2.7 | 10-Oct-2002 |
jdolecek | sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work
|
| 1.112.2.6 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.112.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.112.2.4 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.112.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.112.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.112.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.114.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.119.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.127.4.5 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #1680)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.127.4.4 | 22-Oct-2003 |
jmc | Pullup rev 1.144 (requested by he in ticket #1530)
Introduce a new INVOKING status for callouts, and use it to close
a race condition in the TCP code. Fixes PR#20390.
|
| 1.127.4.3 | 09-Sep-2003 |
tron | Pull up revision 1.153 (requested by itojun in ticket #1452):
initialize ip_hl for ipsec policy lookup. PR kern/22715
|
| 1.127.4.2 | 05-Sep-2003 |
tron | Pull up revision 1.128 (requested by tls in ticket #1445):
path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.127.4.1 | 02-Jul-2002 |
lukem | Pull up revision 1.132 (requested by itojun in ticket #422):
check AF_INET6 socketes when IPv4 "too big" messages arrive.
PR 17448
|
| 1.127.2.4 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.127.2.3 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.127.2.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.127.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.142.2.11 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.142.2.10 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.142.2.9 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.142.2.8 | 17-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.142.2.7 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.142.2.6 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.142.2.5 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.142.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.142.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.142.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.142.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.160.2.5 | 19-Sep-2004 |
he | Apply patch (requested by yamt in ticket #861):
Fix this so it compiles again; we cannot use the link
set macros for pool initialization on this release branch.
|
| 1.160.2.4 | 18-Sep-2004 |
he | Pull up revision 1.173 (requested by yamt in ticket #861):
Fix ipqent pool corruption problems. Make the TCP reassembly
code use its own pool of ipqent rather than sharing it with
the IP reassembly code. Fixes PR#24782.
|
| 1.160.2.3 | 22-Apr-2004 |
tron | Pull up revision 1.164 (requested by tls in ticket #172):
Change the default state of two tunables; bring our TCP a little bit
closer to normal behaviour for the current century.
New Reno is now on by default (which is really the only reasonable
choice, since we don't do SACK); instead of an initial window of 1
for non-local nets, we now use Sally Floyd's magic 4K rule.
|
| 1.160.2.2 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #169)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.160.2.1 | 06-Apr-2004 |
jmc | Pullup rev 1.161 (requested by christos in ticket #69)
Invoking tcpcb's get erroneously free'd resulting in to_ticks <= 0 assertion.
PR#22551
|
| 1.177.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.177.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.177.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.187.2.3 | 06-May-2005 |
tron | Pull up revision 1.190 (requested by yamt in ticket #251):
fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.187.2.2 | 06-May-2005 |
tron | Pull up revision 1.189 (requested by kurahone in ticket #199):
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.187.2.1 | 04-Apr-2005 |
tron | Pull up revision 1.188 (requested by yamt in ticket #90):
protect tcpipqent with splvm.
|
| 1.191.2.7 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.191.2.6 | 11-Feb-2008 |
yamt | sync with head.
|
| 1.191.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.191.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.191.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.191.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.191.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.196.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.196.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.196.8.2 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.196.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.196.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.196.4.3 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.196.4.2 | 05-Feb-2006 |
rpaulo | Adapt to in6pcb removal.
|
| 1.196.4.1 | 05-Feb-2006 |
rpaulo | Bye netinet6/in6_pcb.h.
|
| 1.199.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.199.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.199.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.199.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.208.4.2 | 03-Jun-2008 |
skrll | Sync with netbsd-4.
|
| 1.208.4.1 | 04-Jun-2007 |
wrstuden | Update to today's netbsd-4.
|
| 1.208.2.2 | 30-Mar-2008 |
jdc | Pull up revisions:
src/sys/netinet/ip_input.c 1.263
src/sys/netinet/tcp_subr.c 1.225
(requested by cube in ticket #1109).
- Make sure we send a reasonable fragment size when IPSEC is configured.
Otherwise we end up sending a dubious "0" whenever we cannot find a
proper association for the packet.
- Reset sack_newdata along with snd_nxt to avoid improper integer
arithmetics that lead to sending data from an incorrect place in the
stream, making it appear as corrupted.
Patch by Michael Van Elst, based on an analysis by Michael for the IPSEC
stuff and I for the SACK issue.
|
| 1.208.2.1 | 24-May-2007 |
pavel | branches: 1.208.2.1.4;
Pull up following revision(s) (requested by degroote in ticket #667):
sys/netinet/tcp_input.c: revision 1.260
sys/netinet/tcp_output.c: revision 1.154
sys/netinet/tcp_subr.c: revision 1.210
sys/netinet6/icmp6.c: revision 1.129
sys/netinet6/in6_proto.c: revision 1.70
sys/netinet6/ip6_forward.c: revision 1.54
sys/netinet6/ip6_input.c: revision 1.94
sys/netinet6/ip6_output.c: revision 1.114
sys/netinet6/raw_ip6.c: revision 1.81
sys/netipsec/ipcomp_var.h: revision 1.4
sys/netipsec/ipsec.c: revision 1.26 via patch,1.31-1.32
sys/netipsec/ipsec6.h: revision 1.5
sys/netipsec/ipsec_input.c: revision 1.14
sys/netipsec/ipsec_netbsd.c: revision 1.18,1.26
sys/netipsec/ipsec_output.c: revision 1.21 via patch
sys/netipsec/key.c: revision 1.33,1.44
sys/netipsec/xform_ipcomp.c: revision 1.9
sys/netipsec/xform_ipip.c: revision 1.15
sys/opencrypto/deflate.c: revision 1.8
Commit my SoC work
Add ipv6 support for fast_ipsec
Note that currently, packet with extensions headers are not correctly
supported
Change the ipcomp logic
Add sysctl tree to modify the fast_ipsec options related to ipv6. Similar
to the sysctl kame interface.
Choose the good default policy, depending of the adress family of the
desired policy
Increase the refcount for the default ipv6 policy so nobody can reclaim it
Always compute the sp index even if we don't have any sp in spd. It will
let us to choose the right default policy (based on the adress family
requested).
While here, fix an error message
Use dynamic array instead of an static array to decompress. It lets us to
decompress any data, whatever is the radio decompressed data / compressed
data.
It fixes the last issues with fast_ipsec and ipcomp.
While here, bzero -> memset, bcopy -> memcpy, FREE -> free
Reviewed a long time ago by sam@
|
| 1.208.2.1.4.1 | 30-Mar-2008 |
jdc | Pull up revisions:
src/sys/netinet/ip_input.c 1.263
src/sys/netinet/tcp_subr.c 1.225
(requested by cube in ticket #1109).
- Make sure we send a reasonable fragment size when IPSEC is configured.
Otherwise we end up sending a dubious "0" whenever we cannot find a
proper association for the packet.
- Reset sack_newdata along with snd_nxt to avoid improper integer
arithmetics that lead to sending data from an incorrect place in the
stream, making it appear as corrupted.
Patch by Michael Van Elst, based on an analysis by Michael for the IPSEC
stuff and I for the SACK issue.
|
| 1.210.2.4 | 07-May-2007 |
yamt | sync with head.
|
| 1.210.2.3 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.210.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.210.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.212.2.5 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.212.2.4 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.212.2.3 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.212.2.2 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.212.2.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.213.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.216.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.218.20.2 | 02-Aug-2007 |
rmind | TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.218.20.1 | 02-Aug-2007 |
rmind | file tcp_subr.c was added on branch matt-mips64 on 2007-08-02 02:42:42 +0000
|
| 1.218.16.2 | 19-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.218.16.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.218.12.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.218.10.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.218.4.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.218.4.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.222.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.222.6.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.222.6.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.222.2.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.227.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.229.2.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.229.2.3 | 20-Jun-2009 |
yamt | sync with head
|
| 1.229.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.229.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.231.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.231.2.1 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.233.2.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.233.2.1 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.234.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.234.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.238.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.238.4.2 | 31-May-2011 |
rmind | sync with head
|
| 1.238.4.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.242.2.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.242.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.242.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.243.2.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.243.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.246.2.1 | 31-Oct-2012 |
riz | Pull up following revision(s) (requested by msaitoh in ticket #644):
sys/netinet/tcp_subr.c: revision 1.248
sys/kern/subr_cprng.c: revision 1.12
Fix a bug that kmem_alloc() is called from the interrupt context.
|
| 1.248.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.248.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.248.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.250.2.3 | 18-May-2014 |
rmind | sync with head
|
| 1.250.2.2 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.250.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.255.4.2 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by he in ticket #530):
sys/netinet/tcp_output.c: revision 1.180
sys/netinet/tcp_input.c: revision 1.336
sys/netinet/tcp_usrreq.c: revision 1.203
share/man/man4/tcp.4: revision 1.30
sys/netinet/tcp.h: revision 1.31
sys/netinet/tcp_subr.c: revision 1.258
sys/netinet/tcp_var.h: revision 1.176
sys/netinet/tcp_var.h: revision 1.177
sys/sys/param.h: bump revision
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.255.4.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.257.2.8 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.257.2.7 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.257.2.6 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.257.2.5 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.257.2.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.257.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.257.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.257.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.266.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.266.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.269.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.270.6.3 | 09-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1662):
sys/netinet/tcp_subr.c: revision 1.286
sys/netinet/tcp_timer.c: revision 1.96
sys/netinet/in_var.h: revision 1.102
sys/netinet/in_var.h: revision 1.99
Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
Add some randomness to the iss offset
Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
mv <sys/cprng.h> include to the kernel portion
|
| 1.270.6.2 | 07-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1661):
sys/netinet6/ip6_id.c: revision 1.19-1.21
sys/netinet6/ip6_var.h: revision 1.88
sys/netinet/ip_input.c: revision 1.400
sys/netinet/tcp_subr.c: revision 1.285
sys/netinet/ip6.h: revision 1.30
netinet: Enable random IP fragment ids by default (from riastradh)
netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
netinet6: Mark randomid unused.
Will make merging and bisection easier if anything goes wrong with
flow label or fragment id randomization changes.
(from riastradh)
netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
Replace randomid() by cprng_fast32()
|
| 1.270.6.1 | 03-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #514):
sys/net/route.c: 1.205
sys/net/rtsock.c: 1.237-1.238
sys/netinet/in.c: 1.215
sys/netinet/tcp_subr.c: 1.272
sys/netinet/tcp_timer.c: 1.93
sys/netinet/tcp_timer.h: 1.29
sys/netinet/tcp_var.h: 1.182
sys/netinet6/in6.c: 1.258
Remove extra pserialize_perform from in_purgeaddr
It's already performed in ifa_remove. Note so there (in in6_unlink_ifa too).
Release rt_so_mtx on updating a rtentry to avoid a deadlock with route_intr
The deadlock happened only if NET_MPSAFE on.
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
Fix a return value of rt_update_prepare
Callers expect it to be an errno.
Fix another deadlock
When waiting for a route update to finish, a waiter has to release its reference
to the route to avoid a deadlock. Because a updater tries to wait for references
to a target route (except for a reference by the updater itself) to be released.
|
| 1.273.2.6 | 18-Jan-2019 |
pgoyette | Synch with HEAD
|
| 1.273.2.5 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.273.2.4 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.273.2.3 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.273.2.2 | 22-Apr-2018 |
pgoyette | Sync with HEAD
|
| 1.273.2.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.280.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.280.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.282.4.3 | 09-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1229):
sys/netinet/tcp_subr.c: revision 1.286
sys/netinet/tcp_timer.c: revision 1.96
sys/netinet/in_var.h: revision 1.102
sys/netinet/in_var.h: revision 1.99
Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
Add some randomness to the iss offset
Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
mv <sys/cprng.h> include to the kernel portion
|
| 1.282.4.2 | 07-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1226):
sys/netinet6/ip6_id.c: revision 1.19-1.21
sys/netinet6/ip6_var.h: revision 1.88
sys/netinet/ip_input.c: revision 1.400
sys/netinet/tcp_subr.c: revision 1.285
sys/netinet/ip6.h: revision 1.30
netinet: Enable random IP fragment ids by default (from riastradh)
netinet: Enable RFC 1948 pseudorandom TCP ISS selection by default.
(from riastradh)
netinet6: Mark randomid unused.
Will make merging and bisection easier if anything goes wrong with
flow label or fragment id randomization changes.
(from riastradh)
netinet/netinet6: Add necessary includes to make these standalone.
(from riastradh)
Replace randomid() by cprng_fast32()
|
| 1.282.4.1 | 10-Sep-2019 |
martin | Pull up following revision(s) (requested by maxv in ticket #193):
sys/netinet/tcp_timer.h: revision 1.30
sys/netinet/tcp_input.c: revision 1.415
sys/netinet/tcp_usrreq.c: revision 1.225
sys/netinet/tcp_subr.c: revision 1.283
Clamp tcp timer quantities to reasonable ranges.
|
| 1.284.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.288.4.1 | 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.296.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.7 | 29-Jun-2024 |
riastradh | netinet: Use _NET_STAT* API instead of direct array access.
PR kern/58380
|
| 1.6 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to in6pcb_*
|
| 1.5 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.4 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.3 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.2 | 20-Sep-2022 |
ozaki-r | syncache: make some functions static
|
| 1.1 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.2 | 20-Sep-2022 |
ozaki-r | syncache: make some functions static
|
| 1.1 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.99 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.98 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.97 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.96 | 08-Mar-2021 |
christos | Add some randomness to the iss offset
|
| 1.95 | 03-May-2018 |
maxv | branches: 1.95.6; 1.95.14;
Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.94 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.93 | 19-Jan-2018 |
ozaki-r | branches: 1.93.2;
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
|
| 1.92 | 28-Jul-2017 |
maxv | Remove TCP_COMPAT_42. This feature is a workaround for a bug in the TCP
stack of BSD4.2. Having such features just does not make any sense, and
looking at the code, I'm not sure it actually works.
|
| 1.91 | 25-Jul-2016 |
knakahara | branches: 1.91.8;
fix: unlock in reverse order
|
| 1.90 | 26-Apr-2016 |
ozaki-r | branches: 1.90.2;
Sweep unnecessary route.h inclusions
|
| 1.89 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.88 | 10-Nov-2014 |
maxv | branches: 1.88.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.87 | 02-Jan-2014 |
pooka | branches: 1.87.4;
Allow kernels compiled with INET+INET6 to be booted as IPv4-only or IPv6-only.
|
| 1.86 | 31-Aug-2011 |
plunky | branches: 1.86.2; 1.86.12; 1.86.16;
NULL does not need a cast
|
| 1.85 | 20-Apr-2011 |
gdt | Rewrite comments about TCP RTO calculations.
Long ago, the storage representations of srtt and rttvar were changed
from the 4.4BSD scheme, and the comments are out of sync with the
code. This commit rewrites most of the comments that explain the RTO
calculations, and points out some issues in the code.
Joint work with Bev Schwartz of BBN (original analysis and comments),
but I have rewritten and extended them, so errors are mine.
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073. Approved for Public
Release, Distribution Unlimited
|
| 1.84 | 10-Nov-2008 |
uebayasi | branches: 1.84.8; 1.84.10;
Whitespace.
|
| 1.83 | 09-Nov-2008 |
bouyer | Fix kern/39769: race condition in TCP timers
When a TCP timer is disarmed (with callout_stop()) in the general case
callout_invoking() isn't checked, so the timer handler could be called run
when the current interrupt handler exits, athough the timer is disarmed.
This case cause bad things like TCPT_REXMT and TCPT_PERSIST being both pending,
causing a panic (see the PR for details).
Close the issue by aborting the handler if the timer is not callout_expired().
(the EXPIRED flag being cleared by callout_stop()).
|
| 1.82 | 10-Oct-2008 |
ad | branches: 1.82.2; 1.82.4;
tcp_delack: test for TF_DELACK.
|
| 1.81 | 28-Apr-2008 |
martin | branches: 1.81.2; 1.81.6;
Remove clause 3 and 4 from TNF licenses
|
| 1.80 | 24-Apr-2008 |
ad | branches: 1.80.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.79 | 12-Apr-2008 |
thorpej | branches: 1.79.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.78 | 08-Apr-2008 |
thorpej | Change TCP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old tcpstat structure; old netstat
binaries will continue to work properly.
|
| 1.77 | 20-Jun-2007 |
christos | branches: 1.77.28;
- per socket keepalive settings
- settable connection establishment timeout
|
| 1.76 | 09-Oct-2006 |
rpaulo | branches: 1.76.2; 1.76.8; 1.76.10;
Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.75 | 14-May-2006 |
elad | branches: 1.75.8; 1.75.10;
integrate kauth.
|
| 1.74 | 15-Apr-2006 |
christos | Coverity CID 1153: Add KASSERT before deref.
|
| 1.73 | 11-Dec-2005 |
christos | branches: 1.73.4; 1.73.6; 1.73.8; 1.73.10; 1.73.12;
merge ktrace-lwp.
|
| 1.72 | 19-Jul-2005 |
christos | Implement PMTU checks from:
http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.
From OpenBSD.
|
| 1.71 | 02-Mar-2005 |
mycroft | branches: 1.71.2; 1.71.4;
Copyright maintenance.
|
| 1.70 | 28-Feb-2005 |
jonathan | Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.69 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.68 | 27-Jan-2005 |
mycroft | Whoops. Exit fast recovery when handling a timeout.
|
| 1.67 | 26-Jan-2005 |
mycroft | Fix two problems in our TCP stack:
1) If an echoed RFC 1323 time stamp appears to be later than the current time,
ignore it and fall back to old-style RTT calculation. This prevents ending
up with a negative RTT and panicking later.
2) Fix NewReno. This involves a few changes:
a) Implement the send_high variable in RFC 2582. Our implementation is
subtly different; it is one *past* the last sequence number transmitted
rather than being equal to it. This simplifies some logic and makes
the code smaller. Additional logic was required to prevent sequence
number wraparound problems; this is not mentioned in RFC 2582.
b) Make sure we reset t_dupacks on new acks, but *not* on a partial ack.
All of the new ack code is pushed out into tcp_newreno(). (Later this
will probably be a pluggable function.) Thus t_dupacks keeps track of
whether we're in fast recovery all the time, with Reno or NewReno, which
keeps some logic simpler.
c) We do not need to update snd_recover when we're not in fast recovery.
See tech-net for an explanation of this.
d) In the gratuitous fast retransmit prevention case, do not send a packet.
RFC 2582 specifically says that we should "do nothing".
e) Do not inflate the congestion window on a partial ack. (This is done by
testing t_dupacks to see whether we're still in fast recovery.)
This brings the performance of NewReno back up to the same as Reno in a few
random test cases (e.g. transferring peer-to-peer over my wireless network).
I have not concocted a good test case for the behavior specific to NewReno.
|
| 1.66 | 02-Jan-2004 |
itojun | branches: 1.66.8; 1.66.10;
some corrections from markus@openbsd;
- callout_ack() was called with wrong argument
|
| 1.65 | 27-Oct-2003 |
itojun | make it compilable with TCP_DEBUG defined
|
| 1.64 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.63 | 20-Jul-2003 |
he | As a temporary workaround, apply the fix from PR#20390, thereby
cooperating with the callout code in working around the race
condition caused by the TCP code's use of the callout facility.
Instead of unconditionally releasing memory in tcp_close() and
SYN_CACHE_PUT(), check whether any of the related callout handlers
are about to be invoked (but have not yet done callout_ack()), and
if so, just mark the associated data structure (tcpcb or syn cache
entry) as "dead", and test for this (and release storage) in the
callout handler functions.
|
| 1.62 | 03-Feb-2003 |
thorpej | branches: 1.62.2;
Test callout_pending(), not callout_active(), and eliminate now-unnecessary
callout_deactivate() calls.
|
| 1.61 | 24-Nov-2002 |
scw | Quell an uninitialised variable warning.
|
| 1.60 | 22-Oct-2002 |
simonb | Guard use of "so" in tcp_timer_persist() and tcp_timer_2msl() with
#ifdef TCP_DEBUG.
|
| 1.59 | 09-Jun-2002 |
itojun | whitespace
|
| 1.58 | 26-May-2002 |
itojun | path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.57 | 13-Nov-2001 |
lukem | branches: 1.57.8; 1.57.10;
add RCSIDs
|
| 1.56 | 04-Nov-2001 |
matt | Change a few variable/tables to const since they are read-only.
|
| 1.55 | 11-Sep-2001 |
thorpej | branches: 1.55.2;
Use callouts for SYN cache timers, rather than traversing time queues
in tcp_slowtimo().
|
| 1.54 | 10-Sep-2001 |
thorpej | Update copyrights.
|
| 1.53 | 10-Sep-2001 |
thorpej | Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.
|
| 1.52 | 10-Sep-2001 |
thorpej | Initialize TCP timer variables in a new function, tcp_timer_init().
|
| 1.51 | 10-Sep-2001 |
thorpej | Split tcp_timers() into multiple functions, one for each timer,
and call it directly from tcp_slowtimo() (via a table) rather
than going through tcp_userreq().
This will allow us to call TCP timers directly from callouts,
in a future revision.
|
| 1.50 | 10-Sep-2001 |
thorpej | Change the way receive idle time and round trip time are measured.
Instead of incrementing t_idle and t_rtt in tcp_slowtimo(), we now
take a timstamp (via tcp_now) and use subtraction to compute the
delta when we actually need it (using unsigned arithmetic so that
tcp_now wrapping is handled correctly).
Based on similar changes in FreeBSD.
|
| 1.49 | 10-Sep-2001 |
thorpej | Use a callout for the delayed ACK timer, and delete tcp_fasttimo().
Expose the delayed ACK timer as net.inet.tcp.delack_ticks.
|
| 1.48 | 19-Oct-2000 |
itojun | branches: 1.48.2; 1.48.4; 1.48.6;
remove #ifdef TCP6. it is not likely for us to bring in sys/netinet6/tcp6*.c
(separate TCP/IPv6 stack) into netbsd-current.
|
| 1.47 | 17-Oct-2000 |
itojun | be more friendly with INET-less build.
XXX we need to do more to do a working INET-less build
|
| 1.46 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.45 | 14-Jul-1999 |
itojun | branches: 1.45.2;
Use proper ip protocol # field and tcp hdr on sending RST against SYN,
when ip header and tcp header are not adjacent to each other
(i.e. when ip6 options are attached).
To test this, try
telnet @::1@::1 port
toward a port without responding server. Prior to the fix, the kernel will
generate broken RST packet.
|
| 1.44 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.43 | 10-Sep-1998 |
mouse | branches: 1.43.8; 1.43.10;
Create tcp.keepidle, tcp.keepintvl, tcp.keepcnt, tcp.slowhz sysctls.
|
| 1.42 | 04-Sep-1998 |
mycroft | Fix a couple of bogons related to tcp_new_iss():
* Don't add tcp_iss_seq when creating a new ISS from TIME-WAIT state.
* Do the clock increment even when using the rnd device.
|
| 1.41 | 17-Jul-1998 |
thorpej | Comment where we use the Loss Window.
|
| 1.40 | 02-Jun-1998 |
thorpej | Loss window MUST be one segment, per draft-floyd-incr-init-win-03.
|
| 1.39 | 11-May-1998 |
thorpej | Make sure a timer is marked "disarmed" once it has expired.
|
| 1.38 | 11-May-1998 |
thorpej | Nuke TUBA per my note to tech-net; there's no reason to keep it around.
|
| 1.37 | 07-May-1998 |
thorpej | Define all TCP timers in terms of PRT timers.
|
| 1.36 | 06-May-1998 |
thorpej | Use macros from tcp_timer.h to manipulate TCP timers, so that their
implementation can be changed easily.
|
| 1.35 | 01-May-1998 |
kml | Remove bogus black hole discovery code
|
| 1.34 | 29-Apr-1998 |
thorpej | Make use of the work-arounds for ancient broken TCP peers run-time
conditional (tcp_compat_42). The kernel config option TCP_COMPAT_42
will still enable this by default, or disable this by default if the
option is not included (i.e. current behavior). This will be made a
sysctl soon.
|
| 1.33 | 29-Apr-1998 |
kml | Add support for deletion of routes added by path MTU discovery;
uses new generic route timeout code. Add sysctl for timeout period.
|
| 1.32 | 31-Mar-1998 |
thorpej | Fix a potential-congestion case in the larger initial congestion window
code, as clarified in the TCPIMPL WG meeting at IETF #41: If the SYN
(active open) or SYN,ACK (passive open) was retransmitted, the initial
congestion window for the first slow start of that connection must be
one segment.
|
| 1.31 | 19-Mar-1998 |
kml | Fix a retransmission bug introduced by the Brakmo and Peterson
RTO estimation changes. Under some circumstances it would return a value
of 0, while the old Van Jacobson RTO code would return a minimum of 3.
This would result in 12 retransmissions, each 1 second apart.
This takes care of those instances, and ensures that t_rttmin is
used everywhere as a lower bound.
|
| 1.30 | 19-Feb-1998 |
thorpej | Update copyright (sigh, should have done this long ago).
|
| 1.29 | 12-Jan-1998 |
scottr | Use option header file for TCP_COMPAT_42
|
| 1.28 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.27 | 31-Dec-1997 |
thorpej | Implement a queue for delayed ACK processing. This queue is used in
tcp_fasttimo() in lieu of scanning all open TCP connections.
|
| 1.26 | 17-Dec-1997 |
thorpej | From 4.4BSD-Lite2:
- When running the slow timers, skip PCBs in LISTEN state.
- When processing the persist timer, drop the connection if the connection
idle time exceeds the maximum backoff for retransmit. Part of
kern/2335 (pete@daemon.net).
|
| 1.25 | 11-Dec-1997 |
thorpej | Implement an infrastructure to allow larger initial congestion windows.
The sysctl'able variable "tcp_init_win", when set to 0, selects an
auto-tuning algorithm for selecting the initial window, based on transmit
segment size, per discussion in the IETF tcpimpl working group.
Default initial window is still 1 segment, but will soon become 2 segments,
per discussion in tcpimpl.
|
| 1.24 | 11-Dec-1997 |
thorpej | In tcp_fasttimo(), don't clear TF_DELACK; we need it to count delayed ACKs
in tcp_output(), and it will only be cleared in tcp_output() if the ACK was
transmitted sucessfully. Also, don't count delayed ACKs here, let tcp_output()
count them.
|
| 1.23 | 09-Dec-1997 |
thorpej | Costmetic change: use intotcpcb() in tcp_fasttimo().
|
| 1.22 | 08-Nov-1997 |
kml | TCP MSS fixes to provide cleaner slow-start and recovery.
|
| 1.21 | 13-Oct-1997 |
explorer | branches: 1.21.2;
o Make usage of /dev/random dependant on
pseudo-device rnd # /dev/random and in-kernel generator
in config files.
o Add declaration to all architectures.
o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include
that this code is derived in part from Ted Tyso's linux code.
|
| 1.20 | 10-Oct-1997 |
explorer | Add hooks to use the kernel random system to generate TCP sequence numbers.
|
| 1.19 | 28-Jul-1997 |
thorpej | branches: 1.19.2;
Garbage-collect some "extern"s.
|
| 1.18 | 23-Jul-1997 |
thorpej | Pull SYN_cache_branch down into the main line.
|
| 1.17 | 10-Dec-1996 |
mycroft | branches: 1.17.8;
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.16 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.15 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.14 | 13-Feb-1996 |
christos | branches: 1.14.4;
netinet prototypes
|
| 1.13 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.12 | 18-Jun-1995 |
cgd | convert pcb lists to CIRCLEQs, so that the end can be looked at more
easily, and so that the original (insque/remque) logic can be effectively
mimiced. (This fixes a bug in the previous set of list changes.)
also (since terminator is no longer null) reinstate uninitted list checks,
but mark them XXX.
|
| 1.11 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.10 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.9 | 14-Oct-1994 |
mycroft | Don't return received data to the user until the initial handshake is complete.
Also use TCPS_HAVEESTABLISHED() in a few other places.
|
| 1.8 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.7 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.6 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.5 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.4 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.3 | 22-May-1993 |
cgd | branches: 1.3.4;
add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.3.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.14.4.2 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.14.4.1 | 10-Dec-1996 |
mycroft | From trunk:
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.17.8.2 | 28-Jun-1997 |
thorpej | KNF.
|
| 1.17.8.1 | 14-May-1997 |
mellon | More of David Borman's SYN cache patches for Lite2:
tcp_slowtimo:
- call syn_cache_timer()
|
| 1.19.2.1 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.21.2.3 | 05-May-1998 |
mycroft | Pull up 1.31, per request of kml.
|
| 1.21.2.2 | 29-Jan-1998 |
mellon | Pull up 1.23-1.28 (thorpej)
|
| 1.21.2.1 | 08-Nov-1997 |
thorpej | Pull up from trunk: TCP MSS fixes to provide cleaner slow-start and recovery.
(kml)
|
| 1.43.10.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.43.10.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.43.10.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.43.8.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.43.8.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.45.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.48.6.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.48.4.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.48.4.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.48.4.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.48.2.5 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.48.2.4 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.48.2.3 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.48.2.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.48.2.1 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.55.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.57.10.3 | 15-Mar-2004 |
jmc | Pullup rev 1.66 (requested by he in ticket #1624)
callout_ack() was called with wrong argument
|
| 1.57.10.2 | 22-Oct-2003 |
jmc | Pullup rev 1.63 (requested by he in ticket #1530)
Introduce a new INVOKING status for callouts, and use it to close
a race condition in the TCP code. Fixes PR#20390.
|
| 1.57.10.1 | 05-Sep-2003 |
tron | Pull up revision 1.58 (requested by tls in ticket #1445):
path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.57.8.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.57.8.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.62.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.62.2.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.62.2.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.62.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.62.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.62.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.66.10.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.66.10.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.66.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.71.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.71.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.71.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.71.2.1 | 18-Nov-2008 |
snj | Pull up following revision(s) (requested by bouyer in ticket #1981):
sys/netinet/tcp_timer.c: revision 1.83 via patch
Fix kern/39769: race condition in TCP timers
When a TCP timer is disarmed (with callout_stop()) in the general case
callout_invoking() isn't checked, so the timer handler could be called run
when the current interrupt handler exits, athough the timer is disarmed.
This case cause bad things like TCPT_REXMT and TCPT_PERSIST being both
pending, causing a panic (see the PR for details).
Close the issue by aborting the handler if the timer is not
callout_expired(). (the EXPIRED flag being cleared by callout_stop()).
|
| 1.73.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.73.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.73.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.73.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.73.4.2 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.73.4.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.75.10.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.75.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.76.10.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.76.8.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.76.2.1 | 18-Nov-2008 |
snj | Pull up following revision(s) (requested by bouyer in ticket #1234):
sys/netinet/tcp_timer.c: revision 1.83 via patch
Fix kern/39769: race condition in TCP timers
When a TCP timer is disarmed (with callout_stop()) in the general case
callout_invoking() isn't checked, so the timer handler could be called run
when the current interrupt handler exits, athough the timer is disarmed.
This case cause bad things like TCPT_REXMT and TCPT_PERSIST being both
pending, causing a panic (see the PR for details).
Close the issue by aborting the handler if the timer is not
callout_expired(). (the EXPIRED flag being cleared by callout_stop()).
|
| 1.77.28.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.77.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.79.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.80.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.80.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.81.6.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.81.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.81.2.1 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.82.4.1 | 14-Nov-2008 |
snj | Pull up following revision(s) (requested by bouyer in ticket #56):
sys/netinet/tcp_timer.c: revision 1.83
Fix kern/39769: race condition in TCP timers
When a TCP timer is disarmed (with callout_stop()) in the general case
callout_invoking() isn't checked, so the timer handler could be called run
when the current interrupt handler exits, athough the timer is disarmed.
This case cause bad things like TCPT_REXMT and TCPT_PERSIST being both pending,
causing a panic (see the PR for details).
Close the issue by aborting the handler if the timer is not callout_expired().
(the EXPIRED flag being cleared by callout_stop()).
|
| 1.82.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.84.10.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.84.8.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.86.16.2 | 18-May-2014 |
rmind | sync with head
|
| 1.86.16.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.86.12.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.86.12.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.86.2.1 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.87.4.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.88.2.4 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.88.2.3 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.88.2.2 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.88.2.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.90.2.1 | 26-Jul-2016 |
pgoyette | Sync with HEAD
|
| 1.91.8.2 | 09-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1662):
sys/netinet/tcp_subr.c: revision 1.286
sys/netinet/tcp_timer.c: revision 1.96
sys/netinet/in_var.h: revision 1.102
sys/netinet/in_var.h: revision 1.99
Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
Add some randomness to the iss offset
Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
mv <sys/cprng.h> include to the kernel portion
|
| 1.91.8.1 | 03-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #514):
sys/net/route.c: 1.205
sys/net/rtsock.c: 1.237-1.238
sys/netinet/in.c: 1.215
sys/netinet/tcp_subr.c: 1.272
sys/netinet/tcp_timer.c: 1.93
sys/netinet/tcp_timer.h: 1.29
sys/netinet/tcp_var.h: 1.182
sys/netinet6/in6.c: 1.258
Remove extra pserialize_perform from in_purgeaddr
It's already performed in ifa_remove. Note so there (in in6_unlink_ifa too).
Release rt_so_mtx on updating a rtentry to avoid a deadlock with route_intr
The deadlock happened only if NET_MPSAFE on.
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
Fix a return value of rt_update_prepare
Callers expect it to be an errno.
Fix another deadlock
When waiting for a route update to finish, a waiter has to release its reference
to the route to avoid a deadlock. Because a updater tries to wait for references
to a target route (except for a reference by the updater itself) to be released.
|
| 1.93.2.2 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.93.2.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.95.14.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.95.6.1 | 09-Mar-2021 |
martin | Pull up following revision(s) (requested by christos in ticket #1229):
sys/netinet/tcp_subr.c: revision 1.286
sys/netinet/tcp_timer.c: revision 1.96
sys/netinet/in_var.h: revision 1.102
sys/netinet/in_var.h: revision 1.99
Don't increment the iss sequence on each connection because it exposes
information (Amit Klein)
Add some randomness to the iss offset
Use a random IPv4 ID because the shuffling algorithm used before could expose
information (Amit Klein)
mv <sys/cprng.h> include to the kernel portion
|
| 1.30 | 06-Aug-2019 |
riastradh | Clamp tcp timer quantities to reasonable ranges.
Reported-by: syzbot+259675123340bf46a6de@syzkaller.appspotmail.com
|
| 1.29 | 19-Jan-2018 |
ozaki-r | branches: 1.29.4; 1.29.8;
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
|
| 1.28 | 24-May-2011 |
gdt | branches: 1.28.48;
Note units and current value for TCP_DELACK_TICKS.
|
| 1.27 | 20-Apr-2011 |
gdt | Rewrite comments about TCP RTO calculations.
Long ago, the storage representations of srtt and rttvar were changed
from the 4.4BSD scheme, and the comments are out of sync with the
code. This commit rewrites most of the comments that explain the RTO
calculations, and points out some issues in the code.
Joint work with Bev Schwartz of BBN (original analysis and comments),
but I have rewritten and extended them, so errors are mine.
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073. Approved for Public
Release, Distribution Unlimited
|
| 1.26 | 28-Apr-2008 |
martin | branches: 1.26.22; 1.26.28;
Remove clause 3 and 4 from TNF licenses
|
| 1.25 | 20-Jun-2007 |
christos | branches: 1.25.28; 1.25.30; 1.25.32;
- per socket keepalive settings
- settable connection establishment timeout
|
| 1.24 | 26-Sep-2006 |
jeremy | branches: 1.24.8; 1.24.10;
Fixed a bug in the timeout range constraint macro that can cause a timeout
to break free of the constraint if the range minimum boundary is larger than
the maximum boundary.
Discovered by jmg@FreeBSD.org. (See FreeBSD's tcp_timer.h rev 1.31).
|
| 1.23 | 10-Dec-2005 |
elad | branches: 1.23.20; 1.23.22;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.22 | 02-Jun-2005 |
riz | branches: 1.22.2;
Fix some const fallout.
|
| 1.21 | 04-Mar-2005 |
mycroft | Re-add callout_active(), in a way compatible with the FreeBSD version, and use
it in the TCP stack to test which of the REXMT or PERSIST timer is in use.
This fixes a race condition that could cause "panic: tcp_output REXMT". See
tech-net for details.
|
| 1.20 | 07-Aug-2003 |
agc | branches: 1.20.8; 1.20.10;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.19 | 04-Feb-2003 |
thorpej | branches: 1.19.2;
Use callout_setfunc() and callout_schedule().
|
| 1.18 | 03-Feb-2003 |
thorpej | Test callout_pending(), not callout_active(), and eliminate now-unnecessary
callout_deactivate() calls.
|
| 1.17 | 04-Nov-2001 |
matt | Change a few variable/tables to const since they are read-only.
|
| 1.16 | 10-Sep-2001 |
thorpej | branches: 1.16.2;
Update copyrights.
|
| 1.15 | 10-Sep-2001 |
thorpej | Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.
|
| 1.14 | 10-Sep-2001 |
thorpej | Initialize TCP timer variables in a new function, tcp_timer_init().
|
| 1.13 | 10-Sep-2001 |
thorpej | Add explicit initialization of TCP timer state. A noop right now.
|
| 1.12 | 10-Sep-2001 |
thorpej | Split tcp_timers() into multiple functions, one for each timer,
and call it directly from tcp_slowtimo() (via a table) rather
than going through tcp_userreq().
This will allow us to call TCP timers directly from callouts,
in a future revision.
|
| 1.11 | 10-Sep-2001 |
thorpej | Use a callout for the delayed ACK timer, and delete tcp_fasttimo().
Expose the delayed ACK timer as net.inet.tcp.delack_ticks.
|
| 1.10 | 10-Sep-1998 |
mouse | branches: 1.10.24; 1.10.26; 1.10.28;
Create tcp.keepidle, tcp.keepintvl, tcp.keepcnt, tcp.slowhz sysctls.
|
| 1.9 | 07-May-1998 |
thorpej | Define all TCP timers in terms of PRT timers.
|
| 1.8 | 06-May-1998 |
thorpej | Use the monotonically increasing slow timer timestamp provided by
the protocol dispatch layer for TCP timers. This saves having to
modify a potentially large number of timer values (which were shorts,
and expanded to ... a lot of code on the Alpha).
|
| 1.7 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.6 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.10.28.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.10.26.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.10.26.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.10.24.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.10.24.1 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.16.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.19.2.6 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.19.2.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.19.2.4 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.19.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.19.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.19.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.20.10.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.20.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.22.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.22.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.22.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.23.22.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.23.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.24.10.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.24.8.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.25.32.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.25.30.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.25.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.26.28.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.26.22.2 | 31-May-2011 |
rmind | sync with head
|
| 1.26.22.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.28.48.1 | 03-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #514):
sys/net/route.c: 1.205
sys/net/rtsock.c: 1.237-1.238
sys/netinet/in.c: 1.215
sys/netinet/tcp_subr.c: 1.272
sys/netinet/tcp_timer.c: 1.93
sys/netinet/tcp_timer.h: 1.29
sys/netinet/tcp_var.h: 1.182
sys/netinet6/in6.c: 1.258
Remove extra pserialize_perform from in_purgeaddr
It's already performed in ifa_remove. Note so there (in in6_unlink_ifa too).
Release rt_so_mtx on updating a rtentry to avoid a deadlock with route_intr
The deadlock happened only if NET_MPSAFE on.
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
Fix a return value of rt_update_prepare
Callers expect it to be an errno.
Fix another deadlock
When waiting for a route update to finish, a waiter has to release its reference
to the route to avoid a deadlock. Because a updater tries to wait for references
to a target route (except for a reference by the updater itself) to be released.
|
| 1.29.8.1 | 10-Sep-2019 |
martin | Pull up following revision(s) (requested by maxv in ticket #193):
sys/netinet/tcp_timer.h: revision 1.30
sys/netinet/tcp_input.c: revision 1.415
sys/netinet/tcp_usrreq.c: revision 1.225
sys/netinet/tcp_subr.c: revision 1.283
Clamp tcp timer quantities to reasonable ranges.
|
| 1.29.4.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.238 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to in6pcb_*
|
| 1.237 | 04-Nov-2022 |
ozaki-r | inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.236 | 30-Oct-2022 |
ozaki-r | tcp: restore NULL check for inp in tcp_ctloutput
|
| 1.235 | 29-Oct-2022 |
ozaki-r | tcp: restore NULL checks for inp
|
| 1.234 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.233 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.232 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.231 | 28-Jun-2022 |
riastradh | tcp(4): Bail early on sendoob if not connected.
XXX Not sure if testing tp->t_template is the right way to discern
this -- I just reached for it because the downstream crash is a panic
on tp->t_template == NULL in tcp_output.
XXX In principle this could try connecting to the address, except
it's not passed down from the logic in uipc_socket.c to tcp_sendoob.
Reported-by: syzbot+a01f4cfec72790855ce2@syzkaller.appspotmail.com
|
| 1.230 | 04-Aug-2021 |
christos | Get the value of the right variable (from RVP)
|
| 1.229 | 08-Mar-2021 |
christos | Remove the unused "addin" argument (it was always 0) and go back using
a random iss by default (instead of rfc1948)
|
| 1.228 | 23-Nov-2020 |
chs | Restore correct functioning of SIOCATMARK by removing the previous
change that was done to fix poll(POLLPRI | POLLRDBAND) and instead
add a separate flag to track when poll() should indicate that a
MSG_OOB byte is available. Re-fixes PR 54435 properly.
|
| 1.227 | 17-Oct-2020 |
mlelstv | branches: 1.227.2;
Fix RTT values reported by TCP_INFO.
|
| 1.226 | 13-Apr-2020 |
maxv | hardclock_ticks -> getticks()
|
| 1.225 | 06-Aug-2019 |
riastradh | branches: 1.225.6;
Clamp tcp timer quantities to reasonable ranges.
Reported-by: syzbot+259675123340bf46a6de@syzkaller.appspotmail.com
|
| 1.224 | 05-Feb-2019 |
mrg | branches: 1.224.4;
adjust fallthru comments to appease gcc7.
|
| 1.223 | 28-Jan-2019 |
martin | Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.222 | 16-Dec-2018 |
christos | sbspace() does not return negative values anymore and that broke OOB data
sending. Instead of depending on negative values, account for the 1024
bytes sosend() adds so that it can use all the space here in a separate
function sbspace_oob(). Idea from mlelstv@
|
| 1.221 | 24-Nov-2018 |
maxv | KNF, no functional change.
|
| 1.220 | 24-Nov-2018 |
maxv | Fix kernel pointer leaks in sysctl_inpcblist.
|
| 1.219 | 03-May-2018 |
maxv | branches: 1.219.2;
Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.218 | 07-Apr-2018 |
maxv | Remove dead code.
|
| 1.217 | 29-Mar-2018 |
maxv | Remove #ifdef INET. Same as tcp_input.c. Makes the code easier to
understand.
Also make tcp6_mtudisc() static in tcp_subr.c.
|
| 1.216 | 15-Aug-2017 |
christos | branches: 1.216.2;
add some more getsockopt(2) params
|
| 1.215 | 28-Jul-2017 |
maxv | Remove TCP_COMPAT_42. This feature is a workaround for a bug in the TCP
stack of BSD4.2. Having such features just does not make any sense, and
looking at the code, I'm not sure it actually works.
|
| 1.214 | 24-Jan-2017 |
ozaki-r | branches: 1.214.6;
Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.213 | 18-Nov-2016 |
knakahara | branches: 1.213.2;
fix: "ifconfig destory" can stalls when "ifconfig" is done parallel.
This problem occurs only if NET_MPSAFE on.
ifconfig destroy side:
kernel entry point is ifioctl => if_clone_destroy.
pr_purgeif() acquires softnet_lock, and then ifa_remove() calls
pserialize_perform() holding softnet_lock.
ifconfig side:
kernel entry point is socreate.
pr_attach()(udp_attach_wrapper()) calls sosetlock(). In this call path,
sosetlock() try to acquire softnet_lock.
These can cause dead lock.
|
| 1.212 | 26-Apr-2016 |
ozaki-r | branches: 1.212.2;
Sweep unnecessary route.h inclusions
|
| 1.211 | 15-Feb-2016 |
rtr | Reduce code duplication.
Split creation of IPv4-Mapped IPv6 addresses into its own function
and use it.
No functional change intended. As posted to tech-net@
|
| 1.210 | 14-Feb-2016 |
rtr | remove duplicated #include of <netinet/in.h>
|
| 1.209 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.208 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.207 | 26-Apr-2015 |
rtr | remove pr_generic from struct pr_usrreqs and all implementations of
pr_generic in protocols.
bump to 7.99.13
approved by rmind@
|
| 1.206 | 24-Apr-2015 |
rtr | make accept, getsockname and getpeername syscalls use sockaddr_big and modify
pr_{accept,sockname,peername} nam parameter type from mbuf * to sockaddr *.
* retained use of mbuftypes[MT_SONAME] for now.
* bump to netbsd version 7.99.12 for parameter type change.
patch posted to tech-net@ 2015/04/19
|
| 1.205 | 03-Apr-2015 |
rtr | * change pr_bind to accept struct sockaddr * instead of struct mbuf *
* update protocol bind implementations to use/expect sockaddr *
instead of mbuf *
* introduce sockaddr_big struct for storage of addr data passed via
sys_bind; sockaddr_big is of sufficient size and alignment to
accommodate all addr data sizes received.
* modify sys_bind to allocate sockaddr_big instead of using an mbuf.
* bump kernel version to 7.99.9 for change to pr_bind() parameter type.
Patch posted to tech-net@
http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html
The choice to use a new structure sockaddr_big has been retained since
changing sockaddr_storage size would lead to unnecessary ABI change. The
use of the new structure does not preclude future work that increases
the size of sockaddr_storage and at that time sockaddr_big may be
trivially replaced.
Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
|
| 1.204 | 31-Mar-2015 |
ozaki-r | Remove unnecessary opt_ipsec.h inclusions
|
| 1.203 | 14-Feb-2015 |
he | Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
|
| 1.202 | 10-Nov-2014 |
maxv | branches: 1.202.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.201 | 18-Oct-2014 |
snj | src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.200 | 09-Aug-2014 |
rtr | branches: 1.200.2;
split PRU_CONNECT2 & PRU_PURGEIF function out of pr_generic() usrreq
switches and put into separate functions
- always KASSERT(solocked(so)) even if not implemented
(for PRU_CONNECT2 only)
- replace calls to pr_generic() with req = PRU_CONNECT2 with calls to
pr_connect2()
- replace calls to pr_generic() with req = PRU_PURGEIF with calls to
pr_purgeif()
put common code from unp_connect2() (used by unp_connect() into
unp_connect1() and call out to it when needed
patch only briefly reviewed by rmind@
|
| 1.199 | 08-Aug-2014 |
rtr | split PRU_RCVD function out of pr_generic() usrreq switches and put into
separate functions
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_RCVD with calls to
pr_rcvd()
|
| 1.198 | 05-Aug-2014 |
rtr | split PRU_SEND function out of pr_generic() usrreq switches and put into
separate functions
xxx_send(struct socket *, struct mbuf *, struct mbuf *,
struct mbuf *, struct lwp *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_SEND with calls to
pr_send()
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_send() PRUs
- l2cap_send() -> l2cap_send_pcb()
- sco_send() -> sco_send_pcb()
- rfcomm_send() -> rfcomm_send_pcb()
patch reviewed by rmind
|
| 1.197 | 05-Aug-2014 |
rtr | get_tcppcb() is nearly always called upon entry to usrreqs so
KASSERT(solocked(so)) inside it and remove the redundant KASSERT
everywhere we are using tcp_getpcb()
|
| 1.196 | 05-Aug-2014 |
rtr | revert the removal of struct lwp * parameter from bind, listen and connect
user requests.
this should resolve the issue relating to nfs client hangs presented
recently by wiz on current-users@
|
| 1.195 | 02-Aug-2014 |
rtr | restore splsoftnet() in various usrreqs that were removed during the PRU
splits. we will properly review removal after the PRU split work is
complete.
|
| 1.194 | 31-Jul-2014 |
rtr | split PRU_DISCONNECT, PRU_SHUTDOWN and PRU_ABORT function out of
pr_generic() usrreq switches and put into separate functions
xxx_disconnect(struct socket *)
xxx_shutdown(struct socket *)
xxx_abort(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req =
PRU_{DISCONNECT,SHUTDOWN,ABORT}
with calls to pr_{disconnect,shutdown,abort}() respectively
rename existing internal functions used to implement above functionality
to permit use of the names for xxx_{disconnect,shutdown,abort}().
- {l2cap,sco,rfcomm}_disconnect() ->
{l2cap,sco,rfcomm}_disconnect_pcb()
- {unp,rip,tcp}_disconnect() -> {unp,rip,tcp}_disconnect1()
- unp_shutdown() -> unp_shutdown1()
patch reviewed by rmind
|
| 1.193 | 30-Jul-2014 |
rtr | split PRU_CONNECT function out of pr_generic() usrreq switches and put
into seaparate functions
xxx_listen(struct socket *, struct mbuf *)
- always KASSERT(solocked(so)) and KASSERT(nam != NULL)
- replace calls to pr_generic() with req = PRU_CONNECT with
pr_connect()
- rename existin {l2cap,sco,rfcomm}_connect() to
{l2cap,sco,rfcomm}_connect_pcb() respectively to permit
naming consistency with other protocols functions.
- drop struct lwp * parameter from unp_connect() and at_pcbconnect()
and use curlwp instead where appropriate.
patch reviewed by rmind
|
| 1.192 | 30-Jul-2014 |
rtr | put boilerplate extraction of inpcb or in6pcb and tcpcb performed in tcp
usrreqs into a function that can be called instead of cut & pasting it
to every single usrreq function.
tcp_getpcb(struct socket *, struct inpcb **, struct in6pcb **, struct tcpcb **)
* examines the family of the provided socket and fills in either inpcb
or in6pcb and tcpcb.
* if the pcb is not present for the family of the socket EINVAL is
returned, if the family is not AF_INET{,6} EAFNOSUPPORT is returned.
signature provided by and patch reviewed by rmind
|
| 1.191 | 24-Jul-2014 |
rtr | cleanup after last commit
- add KASSERT(req != PRU_BIND) and KASSERT(req != PRU_LISTEN) inside
tcp_usrreq() as these reqs should no longer reach here.
- remove (now unreachable) PRU_LISTEN case in switch.
|
| 1.190 | 24-Jul-2014 |
rtr | split PRU_BIND and PRU_LISTEN function out of pr_generic() usrreq
switches and put into separate functions
xxx_bind(struct socket *, struct mbuf *)
xxx_listen(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_BIND with call to
pr_bind()
- replace calls to pr_generic() with req = PRU_LISTEN with call to
pr_listen()
- drop struct lwp * parameter from at_pcbsetaddr(), in_pcbbind() and
unp_bind() and always use curlwp.
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_{bind,listen}() PRUs
- l2cap_{bind,listen}() -> l2cap_{bind,listen}_pcb()
- sco_{bind,listen}() -> sco_{bind,listen}_pcb()
- rfcomm_{bind,listen}() -> rfcomm_{bind,listen}_pcb()
patch reviewed by rmind
welcome to netbsd 6.99.48
|
| 1.189 | 23-Jul-2014 |
rtr | split PRU_SENDOOB and PRU_RCVOOB function out of pr_generic() usrreq
switches and put into separate functions
xxx_sendoob(struct socket *, struct mbuf *, struct mbuf *)
xxx_recvoob(struct socket *, struct mbuf *, int)
- always KASSERT(solocked(so)) even if request is not implemented
- replace calls to pr_generic() with req = PRU_{SEND,RCV}OOB with
calls to pr_{send,recv}oob() respectively.
there is still some tweaking of m_freem(m) and m_freem(control) to come
for consistency. not performed with this commit for clarity.
reviewed by rmind
|
| 1.188 | 14-Jul-2014 |
rtr | pr_generic() for req = PRU_RCVOOB is always called with control == NULL
so don't bother with a conditional block that handles non-NULL, it
doesn't happen.
|
| 1.187 | 10-Jul-2014 |
rmind | tcp_accept: simplify a little.
|
| 1.186 | 09-Jul-2014 |
rtr | * split PRU_ACCEPT function out of pr_generic() usrreq switches and put
into a separate function xxx_accept(struct socket *, struct mbuf *)
note: future cleanup will take place to remove struct mbuf parameter
type and replace it with a more appropriate type.
patch reviewed by rmind
|
| 1.185 | 09-Jul-2014 |
rtr | * split PRU_PEERADDR and PRU_SOCKADDR function out of pr_generic()
usrreq switches and put into separate functions
xxx_{peer,sock}addr(struct socket *, struct mbuf *).
- KASSERT(solocked(so)) always in new functions even if request
is not implemented
- KASSERT(pcb != NULL) and KASSERT(nam) if the request is
implemented and not for tcp.
* for tcp roll #ifdef KPROF and #ifdef DEBUG code from tcp_usrreq() into
easier to cut & paste functions tcp_debug_capture() and
tcp_debug_trace()
- functions provided by rmind
- remaining use of PRU_{PEER,SOCK}ADDR #define to be removed in a
future commit.
* rename netbt functions to permit consistency of pru function names
(as has been done with other requests already split out).
- l2cap_{peer,sock}addr() -> l2cap_{peer,sock}_addr_pcb()
- rfcomm_{peer,sock}addr() -> rfcomm_{peer,sock}_addr_pcb()
- sco_{peer,sock}addr() -> sco_{peer,sock}_addr_pcb()
* split/refactor do_sys_getsockname(lwp, fd, which, nam) into
two functions do_sys_get{peer,sock}name(fd, nam).
- move PRU_PEERADDR handling into do_sys_getpeername() from
do_sys_getsockname()
- have svr4_stream directly call do_sys_get{sock,peer}name()
respectively instead of providing `which' & fix a DPRINTF string
that incorrectly wrote "getpeername" when it meant "getsockname"
- fix sys_getpeername() and sys_getsockname() to call
do_sys_get{sock,peer}name() without `which' and `lwp' & adjust
comments
- bump kernel version for removal of lwp & which parameters from
do_sys_getsockname()
note: future cleanup to remove struct mbuf * abuse in
xxx_{peer,sock}name()
still to come, not done in this commit since it is easier to do post
split.
patch reviewed by rmind
welcome to 6.99.47
|
| 1.184 | 07-Jul-2014 |
rtr | * sprinkle KASSERT(solocked(so)); in all pr_stat() functions.
* fix remaining inconsistent struct socket parameter names.
|
| 1.183 | 07-Jul-2014 |
rtr | backout change that made pr_stat return EOPNOTSUPP for protocols that
were not filling in struct stat.
decision made after further discussion with rmind and investigation of
how other operating systems behave. soo_stat() is doing just enough to
be able to call what gets returned valid and thus justifys a return of
success.
additional review will be done to determine of the pr_stat functions
that were already returning EOPNOTSUPP can be considered successful with
what soo_stat() is doing.
|
| 1.182 | 07-Jul-2014 |
rtr | * have pr_stat return EOPNOTSUPP consistently for all protocols that do
not fill in struct stat instead of returning success.
* in pr_stat remove all checks for non-NULL so->so_pcb except where the
pcb is actually used (i.e. cases where we don't return EOPNOTSUPP).
proposed on tech-net@
|
| 1.181 | 06-Jul-2014 |
rtr | * split PRU_SENSE functionality out of xxx_usrreq() switches and place into
separate xxx_stat(struct socket *, struct stat *) functions.
* replace calls using pr_generic with req == PRU_SENSE with pr_stat().
further change will follow that cleans up the pattern used to extract the
pcb and test for its presence.
reviewed by rmind
|
| 1.180 | 01-Jul-2014 |
rtr | fix parameter types in pr_ioctl, called xx_control() functions and remove
abuse of pointer to struct mbuf type.
param2 changed to u_long type and uses parameter name 'cmd' (ioctl command)
param3 changed to void * type and uses parameter name 'data'
param4 changed to struct ifnet * and uses parameter name 'ifp'
param5 has been removed (formerly struct lwp *) and uses of 'l' have been
replaced with curlwp from curproc(9).
callers have had (now unnecessary) casts to struct mbuf * removed, called
code has had (now unnecessary) casts to u_long, void * and struct ifnet *
respectively removed.
reviewed by rmind@
|
| 1.179 | 23-Jun-2014 |
rtr | where appropriate rename xxx_ioctl() struct mbuf * parameters from
`control' to `ifp' after split from xxx_usrreq().
sys_socket.c
fix wrapping of arguments to be consistent with other function calls
in the file after replacing pr_usrreq() call with pr_ioctl() which
required one less argument.
link_proto.c
fix indentation of parameters in link_ioctl() prototype to be
consistent with the rest of the file.
discussed with rmind@
|
| 1.178 | 22-Jun-2014 |
rtr | * split PRU_CONTROL functionality out of xxx_userreq() switches and place
into separate xxx_ioctl() functions.
* place KASSERT(req != PRU_CONTROL) inside xxx_userreq() as it is now
inappropriate for req = PRU_CONTROL in xxx_userreq().
* replace calls to pr_generic() with req = PRU_CONTROL with pr_ioctl().
* remove & fixup references to PRU_CONTROL xxx_userreq() function comments.
* fix various comments references for xxx_userreq() that mentioned
PRU_CONTROL as xxx_userreq() no longer handles the request.
a further change will follow to fix parameter and naming inconsistencies
retained from original code.
Reviewed by rmind@
|
| 1.177 | 22-May-2014 |
rmind | tcp_usrreq: fix the previous - the assert is still not true (but PRU_SENSE
case will handle it); eventually, pr_usrreq should not be called without
PCB attached.
|
| 1.176 | 21-May-2014 |
rmind | tcp_usrreq: fix the previous correctly - restore the assert logic,
but move it after the PRU_SENSE check.
|
| 1.175 | 21-May-2014 |
pgoyette | Restore original sense of the check, and allow both inp and in6p to be
NULL. This case is explicitly handled below.
|
| 1.174 | 20-May-2014 |
rmind | Adjust PR_WRAP_USRREQS() to include the attach/detach functions.
We still need the kernel-lock for some corner cases.
|
| 1.173 | 19-May-2014 |
rmind | - Split off PRU_ATTACH and PRU_DETACH logic into separate functions.
- Replace malloc with kmem and eliminate M_PCB while here.
- Sprinkle more asserts.
|
| 1.172 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.171 | 25-Feb-2014 |
pooka | branches: 1.171.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.170 | 02-Dec-2013 |
kefren | Update TCP CB with new values on rfc1323 and mssdflt sysctl updates
From yasuoka@iij.ad.jp in kern/44254
|
| 1.169 | 23-Nov-2013 |
christos | convert from CIRCLEQ to TAILQ.
|
| 1.168 | 04-Oct-2013 |
christos | PR/48098: Brian Marcotte: Avoid kernel assertion for embryonic sockets that
don't have credentials yet.
XXX: pullup-6
|
| 1.167 | 15-Sep-2013 |
martin | Remove unused variables
|
| 1.166 | 10-Apr-2013 |
christos | branches: 1.166.4;
Limit the tcp initial window setting to 10, leaving it by default to 4
and simplifying the code in process. Per draft-ietf-initcwnd-08.txt.
|
| 1.165 | 02-Jun-2012 |
dsl | branches: 1.165.2;
Add some pre-processor magic to verify that the type of the data item
passed to sysctl_createv() actually matches the declared type for
the item itself.
In the places where the caller specifies a function and a structure
address (typically the 'softc') an explicit (void *) cast is now needed.
Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c
sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting
AcpiGbl_EnableAmlDebugObject.
(mostly passing the address of a uint64_t when typed as CTLTYPE_INT).
I've test built quite a few kernels, but there may be some unfixed MD
fallout. Most likely passing &char[] to char *.
Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
|
| 1.164 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.163 | 17-Mar-2012 |
christos | PR/46077: M. Nunberg: Stat should not fial on connecting socket.
|
| 1.162 | 02-Feb-2012 |
tls | branches: 1.162.2;
Entropy-pool implementation move and cleanup.
1) Move core entropy-pool code and source/sink/sample management code
to sys/kern from sys/dev.
2) Remove use of NRND as test for presence of entropy-pool code throughout
source tree.
3) Remove use of RND_ENABLED in device drivers as microoptimization to
avoid expensive operations on disabled entropy sources; make the
rnd_add calls do this directly so all callers benefit.
4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might
have lead to slight entropy overestimation for some sources.
5) Add new source types for environmental sensors, power sensors, VM
system events, and skew between clocks, with a sample implementation
for each.
ok releng to go in before the branch due to the difficulty of later
pullup (widespread #ifdef removal and moved files). Tested with release
builds on amd64 and evbarm and live testing on amd64.
|
| 1.161 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.160 | 06-Jun-2011 |
dyoung | branches: 1.160.2; 1.160.6;
Don't allocate resources for vtw until/unless it is enabled. This will
further help those machines where memory is in short supply.
TBD: release resources after vtw is disabled and all entries have
expired.
|
| 1.159 | 03-May-2011 |
dyoung | branches: 1.159.2;
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.158 | 30-Dec-2009 |
elad | branches: 1.158.4; 1.158.6;
Get the uid from the socket's credentials.
|
| 1.157 | 16-Sep-2009 |
pooka | Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.156 | 09-Sep-2009 |
darran | Make tcp msl (max segment life) tunable via sysctl net.inet.tcp.msl.
Okayed by tls@.
|
| 1.155 | 07-Jun-2009 |
rmind | sysctl_inpcblist: fix a lock leak in error path (hi <matt>).
|
| 1.154 | 17-Apr-2009 |
elad | Pass the lwp argument to in6_pcbbind() for the INET6 PRU_LISTEN and
PRU_CONNECT cases of tcp_usrreq(). It seems they were forgotten a long
time ago.
Similar code in FreeBSD and OpenBSD passes the thread (credentials)/proc.
|
| 1.153 | 15-Apr-2009 |
elad | Remove a few KAUTH_GENERIC_ISSUSER in favor of more descriptive
alternatives.
Discussed on tech-kern:
http://mail-index.netbsd.org/tech-kern/2009/04/11/msg004798.html
Input from ad@, christos@, dyoung@, tsutsui@.
Okay ad@.
|
| 1.152 | 11-Mar-2009 |
mrg | like KERN_FILE2: *do* update "needed" when there is no count. we want
userland to know what sort of size to provide..
while here, slightly normalise the previous to init_sysctl.c.
|
| 1.151 | 18-Feb-2009 |
yamt | sysctl_net_inet_ip_ports: fix ipv6 sysctls.
|
| 1.150 | 06-Nov-2008 |
dyoung | branches: 1.150.4;
Cosmetic: change (type *)0 to NULL.
|
| 1.149 | 11-Oct-2008 |
pooka | branches: 1.149.2; 1.149.4; 1.149.6;
Move uidinfo to its own module in kern_uidinfo.c and include in rump.
No functional change to uidinfo.
|
| 1.148 | 20-Aug-2008 |
matt | Make the sysctl routines take out softnet_lock before dealing with
any data structures.
Change inet6ctlerrmap and zeroin6_addr to const.
|
| 1.147 | 06-Aug-2008 |
plunky | Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.146 | 04-May-2008 |
thorpej | branches: 1.146.2; 1.146.6;
Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.145 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.144 | 24-Apr-2008 |
ad | branches: 1.144.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.143 | 23-Apr-2008 |
thorpej | Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.142 | 12-Apr-2008 |
thorpej | branches: 1.142.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.141 | 08-Apr-2008 |
thorpej | Change TCP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old tcpstat structure; old netstat
binaries will continue to work properly.
|
| 1.140 | 16-Dec-2007 |
elad | branches: 1.140.6;
Really fix low port allocation, by always passing a valid lwp to
in_pcbbind().
Okay dyoung@.
Note that the network code is another candidate for major cleanup... also
note that this issue is likely to be present in netinet6 code, too.
|
| 1.139 | 27-Nov-2007 |
christos | branches: 1.139.2; 1.139.6;
require that the options argument is the right size, not that it is greater
or equal to the requested size. Suggested by Matt Thomas.
|
| 1.138 | 04-Nov-2007 |
rmind | branches: 1.138.2;
Pick the smallest possible TCP window scaling factor that will still allow
us to scale up to sb_max. This might fix the problems with some firewalls.
Taken from FreeBSD (silby).
OK by <dyoung>.
|
| 1.137 | 19-Sep-2007 |
dyoung | branches: 1.137.4;
1) Introduce a new socket option, (SOL_SOCKET, SO_NOHEADER), that
tells a socket that it should both add a protocol header to tx'd
datagrams and remove the header from rx'd datagrams:
int onoff = 1, s = socket(...);
setsockopt(s, SOL_SOCKET, SO_NOHEADER, &onoff);
2) Add an implementation of (SOL_SOCKET, SO_NOHEADER) for raw IPv4
sockets.
3) Reorganize the protocols' pr_ctloutput implementations a bit.
Consistently return ENOPROTOOPT when an option is unsupported,
and EINVAL if a supported option's arguments are incorrect.
Reorganize the flow of code so that it's more clear how/when
options are passed down the stack until they are handled.
Shorten some pr_ctloutput staircases for readability.
4) Extract common mbuf code into subroutines, add new sockaddr
methods, and introduce a new subroutine, fsocreate(), for reuse
later; use it first in sys_socket():
struct mbuf *m_getsombuf(struct socket *so)
Create an mbuf and make its owner the socket `so'.
struct mbuf *m_intopt(struct socket *so, int val)
Create an mbuf, make its owner the socket `so', put the
int `val' into it, and set its length to sizeof(int).
int fsocreate(..., int *fd)
Create a socket, a la socreate(9), put the socket into the
given LWP's descriptor table, return the descriptor at `fd'
on success.
void *sockaddr_addr(struct sockaddr *sa, socklen_t *slenp)
const void *sockaddr_const_addr(const struct sockaddr *sa, socklen_t *slenp)
Extract a pointer to the address part of a sockaddr. Write
the length of the address part at `slenp', if `slenp' is
not NULL.
socklen_t sockaddr_getlen(const struct sockaddr *sa)
Return the length of a sockaddr. This just evaluates to
sa->sa_len. I only add this for consistency with code that
appears in a portable userland library that I am going to
import.
const struct sockaddr *sockaddr_any(const struct sockaddr *sa)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
const void *sockaddr_anyaddr(const struct sockaddr *sa, socklen_t *slenp)
Return the "don't care" sockaddr in the same family as
`sa'. This is the address a client should sobind(9) if it
does not care the source address and, if applicable, the
port et cetera that it uses.
|
| 1.136 | 02-Aug-2007 |
rmind | branches: 1.136.2; 1.136.4; 1.136.6;
TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.135 | 28-Jun-2007 |
christos | branches: 1.135.2;
Handle mapped and scoped ipv6 addresses. From Anon Ymous.
|
| 1.134 | 26-Jun-2007 |
xtraeme | Protect inet6_ident_core() with #ifdef INET6, fixes building without
options INET6.
|
| 1.133 | 25-Jun-2007 |
christos | tcpdrop kernel bits (from anon ymous)
|
| 1.132 | 20-Jun-2007 |
christos | - per socket keepalive settings
- settable connection establishment timeout
|
| 1.131 | 04-Mar-2007 |
christos | branches: 1.131.2; 1.131.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.130 | 06-Dec-2006 |
yamt | branches: 1.130.2;
add some more tcp mowners.
|
| 1.129 | 10-Nov-2006 |
yamt | branches: 1.129.2; 1.129.4; 1.129.8;
tcp_ctloutput: when called for a socket which is not AF_INET or AF_INET6,
panic rather than returning possibly leaking an mbuf.
|
| 1.128 | 19-Oct-2006 |
rpaulo | Use a better way to create sysctl subtrees for ECN and Congctl.
Inspired on ABC subtree.
|
| 1.127 | 19-Oct-2006 |
yamt | implement RFC3465 appropriate byte counting.
from Kentaro A. Kurahone, with minor adjustments by me.
the ack prediction part of the original patch was omitted because
it's a separate change. reviewed by Rui Paulo.
|
| 1.126 | 16-Oct-2006 |
rpaulo | Export the tcp_do_rfc1948 variable to userland via sysctl.
The code to generate an ISS via an MD5 hash has been present in the
NetBSD kernel since 2001, but it wasn't even exported to userland at
that time. It was agreed on tech-net with the original author <thorpej>
that we should let the user decide if he wants to enable it or not.
Not enabled by default.
|
| 1.125 | 13-Oct-2006 |
elad | Introduce KAUTH_REQ_NETWORK_SOCKET_CANSEE. Since we're not gonna be having
credentials on sockets, at least not anytime soon, this is a way to check
if we can "look" at a socket. Later on when (and if) we do have socket
credentials, the interface usage remains the same because we pass the
socket.
This also fixes sysctl for inet/inet6 pcblist.
|
| 1.124 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.123 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.122 | 13-Sep-2006 |
elad | branches: 1.122.2;
Don't use KAUTH_RESULT_* where it's not applicable.
Prompted by yamt@.
|
| 1.121 | 08-Sep-2006 |
elad | First take at security model abstraction.
- Add a few scopes to the kernel: system, network, and machdep.
- Add a few more actions/sub-actions (requests), and start using them as
opposed to the KAUTH_GENERIC_ISSUSER place-holders.
- Introduce a basic set of listeners that implement our "traditional"
security model, called "bsd44". This is the default (and only) model we
have at the moment.
- Update all relevant documentation.
- Add some code and docs to help folks who want to actually use this stuff:
* There's a sample overlay model, sitting on-top of "bsd44", for
fast experimenting with tweaking just a subset of an existing model.
This is pretty cool because it's *really* straightforward to do stuff
you had to use ugly hacks for until now...
* And of course, documentation describing how to do the above for quick
reference, including code samples.
All of these changes were tested for regressions using a Python-based
testsuite that will be (I hope) available soon via pkgsrc. Information
about the tests, and how to write new ones, can be found on:
http://kauth.linbsd.org/kauthwiki
NOTE FOR DEVELOPERS: *PLEASE* don't add any code that does any of the
following:
- Uses a KAUTH_GENERIC_ISSUSER kauth(9) request,
- Checks 'securelevel' directly,
- Checks a uid/gid directly.
(or if you feel you have to, contact me first)
This is still work in progress; It's far from being done, but now it'll
be a lot easier.
Relevant mailing list threads:
http://mail-index.netbsd.org/tech-security/2006/01/25/0011.html
http://mail-index.netbsd.org/tech-security/2006/03/24/0001.html
http://mail-index.netbsd.org/tech-security/2006/04/18/0000.html
http://mail-index.netbsd.org/tech-security/2006/05/15/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/01/0000.html
http://mail-index.netbsd.org/tech-security/2006/08/25/0000.html
Many thanks to YAMAMOTO Takashi, Matt Thomas, and Christos Zoulas for help
stablizing kauth(9).
Full credit for the regression tests, making sure these changes didn't break
anything, goes to Matt Fleming and Jaime Fournier.
Happy birthday Randi! :)
|
| 1.120 | 05-Sep-2006 |
rpaulo | branches: 1.120.2;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.119 | 23-Jul-2006 |
ad | Use the LWP cached credentials where sane.
|
| 1.118 | 16-Jul-2006 |
elad | get rid of CURTAIN() macro. inline the last use of it, together
with a nice XXX comment (assigned to me of course) that we should
be doing this differently.
|
| 1.117 | 14-May-2006 |
elad | integrate kauth.
|
| 1.116 | 15-Apr-2006 |
christos | Move pf2 assignment after we've assigned pf.
|
| 1.115 | 14-Apr-2006 |
christos | Coverity CID 1154: Prevent NULL deref.
|
| 1.114 | 14-Apr-2006 |
christos | Coverity CID 738: Fix the query size vs. result returning setup.
|
| 1.113 | 11-Dec-2005 |
christos | branches: 1.113.4; 1.113.6; 1.113.8; 1.113.10; 1.113.12;
merge ktrace-lwp.
|
| 1.112 | 15-Nov-2005 |
dsl | Pass the current process structure to in_pcbconnect() so that it can
pass it to in_pcbbind() so that can allocate a low numbered port
if setsockopt() has been used to set IP_PORTRANGE to IP_PORTRANGE_LOW.
While there, fail in_pcbconnect() if the in_pcbbind() fails - rather
than sending the request out from a port of zero.
This has been largely broken since the socket option was added in 1998.
|
| 1.111 | 07-Sep-2005 |
elad | branches: 1.111.6;
Implement curtain for AF_INET{,6} PCB lists.
|
| 1.110 | 06-Sep-2005 |
rpaulo | Correct SYSCTL_DESCR for tcp.debx.
|
| 1.109 | 06-Sep-2005 |
rpaulo | Implement tcp.inet{,6}.tcp{,6}.(debug|debx) when TCP_DEBUG is set. They
can be used to ``transliterate protocol trace'' like trpt(8) does.
|
| 1.108 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.107 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.106 | 20-Jun-2005 |
atatat | branches: 1.106.2;
Change the rest of the sysctl subsystem to use const consistently.
The __UNCONST macro is now used only where necessary and the RW macros
are gone. Most of the changes here are consumers of the
sysctl_createv(9) interface that now takes a pair of const pointers
which used not to be.
|
| 1.105 | 09-Jun-2005 |
atatat | Properly fix the constipated lossage wrt -Wcast-qual and the sysctl
code. I know it's not the prettiest code, but it seems to work rather
well in spite of itself.
|
| 1.104 | 29-May-2005 |
christos | - add const
- remove bogus casts
- avoid nested variables
|
| 1.103 | 07-May-2005 |
christos | PR/30154: YAMAMOTO Takashi: tcp_close locking botch
chgsbsize() as mentioned in the PR can be called from an interrupt context
via tcp_close(). Avoid calling uid_find() in chgsbsize().
- Instead of storing so_uid in struct socketvar, store *so_uidinfo
- Add a simple lock to struct uidinfo.
|
| 1.102 | 05-Apr-2005 |
kurahone | Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.101 | 30-Mar-2005 |
yamt | s of sack is selective, not selection. pointed by Michael Eriksson.
|
| 1.100 | 11-Mar-2005 |
atatat | branches: 1.100.2;
Revert the change that made kern.file2 and net.*.*.pcblist into nodes
instead of structs. It had other deleterious side-effects that are
rather nasty. Another solution must be found.
|
| 1.99 | 10-Mar-2005 |
atatat | Make this build without INET6 xor INET (hah!) again.
|
| 1.98 | 10-Mar-2005 |
atatat | Change types of kern.file2 and net.*.*.pcblist to NODE
|
| 1.97 | 09-Mar-2005 |
atatat | Add the following nodes to the sysctl tree:
net.local.stream.pcblist
net.local.dgram.pcblist
net.inet.tcp.pcblist
net.inet.udp.pcblist
net.inet.raw.pcblist
net.inet6.tcp6.pcblist
net.inet6.udp6.pcblist
net.inet6.raw6.pcblist
which allow retrieval of the pcbs in use for those protocols. The
struct involved is 32/64 bit clean and incorporates parts of struct
inpcb, struct unpcb, a bit of struct tcpcb, and two socket addresses.
|
| 1.96 | 06-Mar-2005 |
yamt | update SYSCTL_DESCR; sack is implemented.
|
| 1.95 | 02-Mar-2005 |
mycroft | Copyright maintenance.
|
| 1.94 | 03-Feb-2005 |
perry | ANSIfy function declarations
|
| 1.93 | 15-Dec-2004 |
thorpej | branches: 1.93.2; 1.93.4;
Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.92 | 25-May-2004 |
atatat | Sysctl descriptions under net subtree (net.key not done)
|
| 1.91 | 18-May-2004 |
itojun | fix MD5 signature support to actually validate inbound signature, and
drop packet if fails.
|
| 1.90 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.89 | 20-Apr-2004 |
matt | export tcpstates for _KERNEL and remove tcp_usrreq.c's incorrect
declartion.
|
| 1.88 | 29-Mar-2004 |
atatat | Make these compile without INET. tcp_input probably needs a lot more
work...
|
| 1.87 | 24-Mar-2004 |
atatat | branches: 1.87.2;
Tango on sysctl_createv() and flags. The flags have all been renamed,
and sysctl_createv() now uses more arguments.
|
| 1.86 | 04-Dec-2003 |
atatat | Dynamic sysctl.
Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(),
vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all
nodes are registered with the tree, and nodes can be added (or
removed) easily, and I/O to and from the tree is handled generically.
Since the nodes are registered with the tree, the mapping from name to
number (and back again) can now be discovered, instead of having to be
hard coded. Adding new nodes to the tree is likewise much simpler --
the new infrastructure handles almost all the work for simple types,
and just about anything else can be done with a small helper function.
All existing nodes are where they were before (numerically speaking),
so all existing consumers of sysctl information should notice no
difference.
PS - I'm sorry, but there's a distinct lack of documentation at the
moment. I'm working on sysctl(3/8/9) right now, and I promise to
watch out for buses.
|
| 1.85 | 22-Oct-2003 |
thorpej | Rather than zeroing a tcpcb structure and filling in all the fields
individually, create a tcpcb template pre-initialized (and pre-zero'd)
with the static and mostly-static tcpcb parameters. The template is
now copied into the new tcpcb, which zeros and initializes most of the
tcpcb in one pass. The template is kept up-to-date as TCP sysctl
variables are changed.
Combined with the previous sb_max change, TCP socket creation is now
25% faster.
|
| 1.84 | 29-Sep-2003 |
tls | Increase default socket-buffer sizes from 16K to 32K. This increases
throughput significantly in a wide variety of test cases, including
local gigabit ethernet with both jumbo and standard frames,
transcontinental (U.S.) connections with e2e bandwidths ranging from
10Mbit/sec to 155Mbit/sec, and on a variety of test connections
between the NetBSD Project public servers and machines in Australia.
The impact of this change is less dramatic for high-delay connections
when Path MTU is in use but still measurable.
For optimal performance on local gigabit networks, a higher socket
buffer size (at least 64K) will still yield a substantial improvement
in performance, but 32K gets us most of the way there in my test
cases, with only a cost of _doubling_ memory use per socket rather
than _quadrupling_ it.
N.B. Windows NT, at least since Win2k SP2, uses a default socket buffer
size (or their analogue thereof) of 64K, which is a useful data
point.
|
| 1.83 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.82 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.81 | 29-Jun-2003 |
fvdl | branches: 1.81.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.80 | 29-Jun-2003 |
simonb | Fix a nit in a comment.
|
| 1.79 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.78 | 26-Jun-2003 |
christos | abuse the mib instead of abusing the new pointer. Idea from simon burge.
It allows the tcp_sysctl_ident to run by non-super-users. No backwards
compatibility provided.
|
| 1.77 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.76 | 19-Apr-2003 |
christos | PR/2352: Tor Egge: Add sysctl to get uid of connected socket.
|
| 1.75 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.74 | 22-Oct-2002 |
simonb | Guard use of "ostate" with #ifdef TCP_DEBUG in tcp_usrreq().
Don't put semicolons at the end of "#define token value".
|
| 1.73 | 03-Jul-2002 |
thorpej | Rename sbappend_stream() to sbappendstream(), per suggestion from
Jonathan Stone.
|
| 1.72 | 03-Jul-2002 |
thorpej | Make insertion of data into socket buffers O(C):
* Keep pointers to the first and last mbufs of the last record in the
socket buffer.
* Use the sb_lastrecord pointer in the sbappend*() family of functions
to avoid traversing the packet chain to find the last record.
* Add a new sbappend_stream() function for stream protocols which
guarantee that there will never be more than one record in the
socket buffer. This function uses the sb_mbtail pointer to perform
the data insertion. Make TCP use sbappend_stream().
On a profiling run, this makes sbappend of a TCP transmission using
a 1M socket buffer go from 50% of the time to .02% of the time.
Thanks to Bill Sommerfeld and YAMAMOTO Takashi for their debugging
assistance!
|
| 1.71 | 09-Jun-2002 |
itojun | whitespace
|
| 1.70 | 11-Mar-2002 |
martin | branches: 1.70.4;
KNFify my last change.
|
| 1.69 | 28-Feb-2002 |
martin | Enforce a lower bound of 32 for tcp_mssdflt.
This avoids kernel crashes when we don't handle nonsensial values
like 0 gracefully. Better check here once beforehand than having to
check for non meaningful values in time critical paths (like tcp_output).
Fixes PR 15709.
|
| 1.68 | 20-Nov-2001 |
lukem | - replace "defopt" with "defparam" for options which must take a value,
as config(8) will warn for value-less defparam options
- minor whitespace/formatting cleanup
- consolidate opt_tcp_recvspace.h and opt_tcp_sendspace.h into opt_tcp_space.h
|
| 1.67 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.66 | 29-Oct-2001 |
simonb | Don't need to include <uvm/uvm_extern.h> just to include <sys/sysctl.h>
anymore.
|
| 1.65 | 10-Sep-2001 |
thorpej | branches: 1.65.2;
Split tcp_timers() into multiple functions, one for each timer,
and call it directly from tcp_slowtimo() (via a table) rather
than going through tcp_userreq().
This will allow us to call TCP timers directly from callouts,
in a future revision.
|
| 1.64 | 25-Jul-2001 |
itojun | branches: 1.64.2;
allocate ipsec policy buffer attached to pcb in in*_pcballoc, before
giving anyone accesses to pcb (do not reveal an inconsistent ones).
sync with kame
|
| 1.63 | 08-Jul-2001 |
abs | branches: 1.63.2;
Rename TCPDEBUG to TCP_DEBUG, defopt TCP_DEBUG and TCP_NDEBUG, and
make all usage of tcp_trace dependent on TCP_DEBUG - resulting in
a 31K saving on an INET enabled i386 kernel.
|
| 1.62 | 03-Jul-2001 |
itojun | call in{,6}_pcbpurgeif0() before in{,6}_purgeif().
|
| 1.61 | 20-Mar-2001 |
thorpej | Two changes, designed to make us even more resilient against TCP
ISS attacks (which we already fend off quite well).
1. First-cut implementation of RFC1948, Steve Bellovin's cryptographic
hash method of generating TCP ISS values. Note, this code is experimental
and disabled by default (experimental enough that I don't export the
variable via sysctl yet, either). There are a couple of issues I'd
like to discuss with Steve, so this code should only be used by people
who really know what they're doing.
2. Per a recent thread on Bugtraq, it's possible to determine a system's
uptime by snooping the RFC1323 TCP timestamp options sent by a host; in
4.4BSD, timestamps are created by incrementing the tcp_now variable
at 2 Hz; there's even a company out there that uses this to determine
web server uptime. According to Newsham's paper "The Problem With
Random Increments", while NetBSD's TCP ISS generation method is much
better than the "random increment" method used by FreeBSD and OpenBSD,
it is still theoretically possible to mount an attack against NetBSD's
method if the attacker knows how many times the tcp_iss_seq variable
has been incremented. By not leaking uptime information, we can make
that much harder to determine. So, we avoid the leak by giving each
TCP connection a timebase of 0.
|
| 1.60 | 11-Feb-2001 |
itojun | branches: 1.60.2;
make sure we call tcp_output() only if we have template.
|
| 1.59 | 18-Jan-2001 |
jdolecek | constify
|
| 1.58 | 11-Dec-2000 |
itojun | make sure t_family has the correct protocol family, after connect(2)
and/or bind(2). sync with kame
|
| 1.57 | 17-Oct-2000 |
itojun | allow INET6-less build.
From: smd@ebone.net (Sean Doran)
|
| 1.56 | 17-Oct-2000 |
itojun | be more friendly with INET-less build.
XXX we need to do more to do a working INET-less build
|
| 1.55 | 06-Oct-2000 |
enami | Cosmetic changes to previous commit; indent break statement sanely.
|
| 1.54 | 06-Oct-2000 |
enami | Just call matching purgeif/pcbpurgeif routine for the protocol family.
Without this, if a v6 address is placed before a v4 address in if_addrlist,
a PRU_PURGEIF request for v6 tcp protocol purges also v4 addresses and,
as a result, if_detach fails to request PRU_PURGEIF for v4 protocols
other than tcp.
|
| 1.53 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.52 | 28-Jun-2000 |
mrg | <vm/vm.h> -> <uvm/uvm_extern.h>
|
| 1.51 | 05-Jun-2000 |
itojun | branches: 1.51.2;
pass struct proc * down to udp6_output and in6_pcbbind.
|
| 1.50 | 22-May-2000 |
itojun | branches: 1.50.2;
disallow negative numbers for ratelimit interval (tcp, icmp, icmp6).
|
| 1.49 | 30-Mar-2000 |
augustss | Remove register declarations.
|
| 1.48 | 15-Feb-2000 |
thorpej | Add support for rate-limiting RSTs sent in response to no socket for
an incoming packet. Default minimum interval is 10ms. The interval
is changeable via the "net.inet.tcp.rstratelimit" sysctl variable.
|
| 1.47 | 04-Feb-2000 |
itojun | avoid calling in6_control(SIOCDIFADDR_IN6) from interrupt context.
it is not supposed to work.
logging fix: add "\n" to some of log() in in6_prefix.c.
improve in6_ifdetach(). now almost all structure depend on ifnet
will be cleared up.
possible loose ends:
- cached route_in6 in static varaiables needs to be cleared as well
- there are ifaddr manipulation without reference counting,
which should be fixed
we still see panics after card removal, though... not sure what is left.
(sync with kame)
|
| 1.46 | 02-Feb-2000 |
thorpej | PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.45 | 01-Feb-2000 |
thorpej | First-draft if_detach() implementation, originally from Bill Studnemund,
although this version has been changed somewhat:
- reference counting on ifaddrs isn't as complete as Bill's original
work was. This is hard to get right, and we should attack one
protocol at a time.
- This doesn't do reference counting or dynamic allocation of ifnets yet.
- This version introduces a new PRU -- PRU_PURGEADDR, which is used to
purge an ifaddr from a protocol. The old method Bill used didn't work
on all protocols, and it only worked on some because it was Very Lucky.
This mostly works ... i.e. works for my USB Ethernet, except for a dangling
ifaddr reference left by the IPv6 code; have not yet tracked this down.
|
| 1.44 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.43 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.42 | 09-Jul-1999 |
thorpej | branches: 1.42.2; 1.42.8;
defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.41 | 02-Jul-1999 |
itojun | avoid "variable not initialized" warnings on some of the platforms.
|
| 1.40 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.39 | 10-Sep-1998 |
tv | branches: 1.39.8; 1.39.10;
egcs {brace} warning fix
|
| 1.38 | 10-Sep-1998 |
mouse | Create tcp.keepidle, tcp.keepintvl, tcp.keepcnt, tcp.slowhz sysctls.
|
| 1.37 | 06-May-1998 |
thorpej | Use macros from tcp_timer.h to manipulate TCP timers, so that their
implementation can be changed easily.
|
| 1.36 | 29-Apr-1998 |
matt | New TCP reassembly code. The new code reduces the memory needed by
out-of-order packets and builds the infrastructure needed for sending
SACK blocks (to be added shortly).
|
| 1.35 | 13-Apr-1998 |
kml | Fix to ensure that the correct MSS is advertised for loopback
TCP connections by using the MTU of the interface. Also added
a knob, mss_ifmtu, to force all connections to use the MTU of
the interface to calculate the advertised MSS.
|
| 1.34 | 19-Feb-1998 |
thorpej | Update copyright (sigh, should have done this long ago).
|
| 1.33 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.32 | 05-Jan-1998 |
thorpej | From 4.4BSD-Lite2 (noted by Frank van der Linden):
so_linger is used as an argument to tsleep(), so was stuffed with
clockticks for the TCP linger time. However, so_linger is set directly from
l_linger if the linger time is specified, and l_linger is seconds (although
this is not currently documented anywhere). Fix this to set the TCP
linger time in seconds, and multiply so_linger by hz when tsleep() is
called to actually perform the linger.
|
| 1.31 | 13-Dec-1997 |
thorpej | After further examination of traces of bulk transfers (with help from
Kevin Lahey), undo the "defer window update until next delayed ACK".
|
| 1.30 | 11-Dec-1997 |
thorpej | Implement an infrastructure to allow larger initial congestion windows.
The sysctl'able variable "tcp_init_win", when set to 0, selects an
auto-tuning algorithm for selecting the initial window, based on transmit
segment size, per discussion in the IETF tcpimpl working group.
Default initial window is still 1 segment, but will soon become 2 segments,
per discussion in tcpimpl.
|
| 1.29 | 11-Dec-1997 |
thorpej | In the PRU_RCVD entry point, if TF_DELACK is set, don't send the window
update now, since it will be sent within 200ms when the delayed ACK is
sent. Instrument how many hits we get on this optimization.
|
| 1.28 | 08-Nov-1997 |
kml | TCP MSS fixes to provide cleaner slow-start and recovery.
|
| 1.27 | 10-Oct-1997 |
explorer | branches: 1.27.2;
Add hooks to use the kernel random system to generate TCP sequence numbers.
|
| 1.26 | 28-Jul-1997 |
thorpej | branches: 1.26.2;
Generate dependencies for the TCP_SENDSPACE and TCP_RECVSPACE options.
|
| 1.25 | 28-Jul-1997 |
thorpej | Make the following tunable via sysctl, inspired by BSD/OS:
- tcp_sendspace
- tcp_recvspace
- tcp_mssdflt
- tcp_syn_cache_limit
- tcp_syn_bucket_limit
- tcp_syn_cache_timer
|
| 1.24 | 12-Jun-1997 |
kleink | Eliminate a superflouus `if' statement: when detaching the TCP protocol from
a socket, just calling tcp_disconnect() on the tcpcb will do the right thing.
From Thorsten Frueauf <frueauf@ira.uka.de> and W. Richard Stevens in PR/3738
resp. TCP/IP Illustrated, Vol. 2.
|
| 1.23 | 23-May-1996 |
mycroft | Make sure the control mbufs are freed in all cases.
|
| 1.22 | 23-May-1996 |
mycroft | Minor changes.
|
| 1.21 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.20 | 13-Feb-1996 |
christos | branches: 1.20.4;
netinet prototypes
|
| 1.19 | 31-Jan-1996 |
mycroft | Add a comment describing the previous.
|
| 1.18 | 31-Jan-1996 |
mycroft | If we close from FIN_WAIT_2 state, make sure we don't leave the socket
around forever if we don't get a final FIN. From Arne Juul, PR 1659.
|
| 1.17 | 30-Sep-1995 |
thorpej | branches: 1.17.2;
Implement tcp_sysctl(). Add a sysctl option to enable/disable RFC1323
extensions to TCP. From John Kohl <jtk@kolvir.blrc.ma.us>.
|
| 1.16 | 12-Aug-1995 |
mycroft | splnet --> splsoftnet
|
| 1.15 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.14 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.13 | 21-Mar-1995 |
glass | Default linger time was 120 clock ticks instead of the intended
2 minutes.
[Bug pointed out by Wright/Stevens in TCP/IP Illustrated Vol II]
|
| 1.12 | 14-Oct-1994 |
mycroft | Don't return received data to the user until the initial handshake is complete.
Also use TCPS_HAVEESTABLISHED() in a few other places.
|
| 1.11 | 13-Oct-1994 |
mycroft | Increase the default window size to 16k.
|
| 1.10 | 29-Jun-1994 |
cgd | branches: 1.10.2;
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.9 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.8 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.7 | 08-Jan-1994 |
mycroft | Remove some extra prototypes.
|
| 1.6 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 15-Jun-1993 |
cgd | branches: 1.4.4;
bump sendspace and recvspace up to 8k each; rod says
these should be safe values...
|
| 1.3 | 22-May-1993 |
cgd | add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.4.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.10.2.1 | 13-Oct-1994 |
mycroft | Update from trunk.
|
| 1.17.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.20.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.26.2.1 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.27.2.2 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.27.2.1 | 08-Nov-1997 |
thorpej | Pull up from trunk: TCP MSS fixes to provide cleaner slow-start and recovery.
(kml)
|
| 1.39.10.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.39.10.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.39.10.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.39.8.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.39.8.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.42.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.42.2.4 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.42.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.42.2.2 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.42.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.50.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.51.2.2 | 09-Oct-2000 |
enami | Pullup rev. 1.54 and 1.55 (approved by jhawk):
|
| 1.51.2.1 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.60.2.9 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.60.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.60.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.60.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.60.2.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.60.2.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.60.2.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.60.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.60.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.63.2.6 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.63.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.63.2.4 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.63.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.63.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.63.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.64.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.65.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.70.4.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.70.4.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.81.2.11 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.81.2.10 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.81.2.9 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.81.2.8 | 08-Mar-2005 |
skrll | Sync with HEAD.
|
| 1.81.2.7 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.81.2.6 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.81.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.81.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.81.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.81.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.81.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.87.2.1 | 28-May-2004 |
tron | Pull up revision 1.92 (requested by atatat in ticket #391):
Sysctl descriptions under net subtree (net.key not done)
|
| 1.93.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.93.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.93.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.100.2.2 | 06-May-2005 |
tron | Pull up revision 1.102 (requested by kurahone in ticket #199):
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.100.2.1 | 01-May-2005 |
tron | Pull up revision 1.101 (requested by yamt in ticket #221):
s of sack is selective, not selection. pointed by Michael Eriksson.
|
| 1.106.2.7 | 21-Jan-2008 |
yamt | sync with head
|
| 1.106.2.6 | 07-Dec-2007 |
yamt | sync with head
|
| 1.106.2.5 | 15-Nov-2007 |
yamt | sync with head.
|
| 1.106.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.106.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.106.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.106.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.111.6.1 | 22-Nov-2005 |
yamt | sync with head.
|
| 1.113.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.113.10.3 | 06-May-2006 |
christos | - Move kauth_cred_t declaration to <sys/types.h>
- Cleanup struct ucred; forward declarations that are unused.
- Don't include <sys/kauth.h> in any header, but include it in the c files
that need it.
Approved by core.
|
| 1.113.10.2 | 19-Apr-2006 |
elad | sync with head.
|
| 1.113.10.1 | 08-Mar-2006 |
elad | Adapt to kernel authorization KPI.
|
| 1.113.8.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.113.8.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.113.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.113.6.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.113.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.113.4.3 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.113.4.2 | 05-Feb-2006 |
rpaulo | in6pcb -> inpcb merge.
|
| 1.113.4.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.120.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.120.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.122.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.122.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.129.8.1 | 05-Sep-2009 |
bouyer | Pull up following revision(s) (requested by mlelstv in ticket #1358):
sys/netinet/tcp_usrreq.c: revision 1.148 via patch
Make the sysctl routines take raise to splnet() before dealing with
any data structures.
|
| 1.129.4.1 | 03-Jun-2008 |
skrll | Sync with netbsd-4.
|
| 1.129.2.3 | 05-Sep-2009 |
bouyer | Pull up following revision(s) (requested by mlelstv in ticket #1358):
sys/netinet/tcp_usrreq.c: revision 1.148 via patch
Make the sysctl routines take raise to splnet() before dealing with
any data structures.
|
| 1.129.2.2 | 29-Jan-2008 |
pavel | Pull up following revision(s) (requested by joerg in ticket #1057):
sys/netinet/tcp_usrreq.c: revision 1.134
Protect inet6_ident_core() with #ifdef INET6, fixes building without
options INET6.
|
| 1.129.2.1 | 21-Jan-2008 |
bouyer | Pull up following revision(s) (requested by ghen in ticket #1039):
sys/netinet/tcp_var.h: revision 1.148
distrib/sets/lists/comp/mi: revision 1.1035
distrib/sets/lists/man/mi: revision 1.1010
usr.sbin/tcpdrop/Makefile: revision 1.1
usr.sbin/tcpdrop/tcpdrop.c: revision 1.1 - 1.3
usr.sbin/tcpdrop/tcpdrop.8: revision 1.1
usr.sbin/Makefile: revision 1.228 via patch
sys/netinet/tcp_usrreq.c: revision 1.133
distrib/sets/lists/base/mi: revision 1.712
Import tcpdrop(8) from OpenBSD
|
| 1.130.2.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.131.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.131.2.3 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.131.2.2 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.131.2.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.135.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.136.6.2 | 02-Aug-2007 |
rmind | TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.136.6.1 | 02-Aug-2007 |
rmind | file tcp_usrreq.c was added on branch matt-mips64 on 2007-08-02 02:42:42 +0000
|
| 1.136.4.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.136.4.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.136.2.3 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.136.2.2 | 04-Nov-2007 |
jmcneill | Sync with HEAD.
|
| 1.136.2.1 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.137.4.1 | 13-Nov-2007 |
bouyer | Sync with HEAD
|
| 1.138.2.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.138.2.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.139.6.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.139.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.140.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.140.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.140.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.142.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.144.2.5 | 11-Mar-2010 |
yamt | sync with head
|
| 1.144.2.4 | 16-Sep-2009 |
yamt | sync with head
|
| 1.144.2.3 | 20-Jun-2009 |
yamt | sync with head
|
| 1.144.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.144.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.146.6.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.146.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.146.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.149.6.1 | 17-Jun-2009 |
bouyer | branches: 1.149.6.1.2;
Pull up following revision(s) (requested by rmind in ticket #812):
sys/netinet/tcp_usrreq.c: revision 1.155
sysctl_inpcblist: fix a lock leak in error path (hi <matt>).
|
| 1.149.6.1.2.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.149.4.2 | 26-Sep-2009 |
snj | Pull up following revision(s) (requested by darran in ticket #950):
sys/netinet/tcp_input.c: revision 1.299
sys/netinet/tcp_usrreq.c: revision 1.156
sys/netinet/tcp_var.h: revision 1.161
Make tcp msl (max segment life) tunable via sysctl net.inet.tcp.msl.
Okayed by tls@.
|
| 1.149.4.1 | 17-Jun-2009 |
bouyer | Pull up following revision(s) (requested by rmind in ticket #812):
sys/netinet/tcp_usrreq.c: revision 1.155
sysctl_inpcblist: fix a lock leak in error path (hi <matt>).
|
| 1.149.2.3 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.149.2.2 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.149.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.150.4.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.150.4.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.158.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.158.4.2 | 12-Jun-2011 |
rmind | sync with head
|
| 1.158.4.1 | 31-May-2011 |
rmind | sync with head
|
| 1.159.2.1 | 23-Jun-2011 |
cherry | Catchup with rmind-uvmplock merge.
|
| 1.160.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.160.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.160.2.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.160.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.160.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.162.2.3 | 14-Dec-2013 |
bouyer | Pull up following revision(s) (requested by kefren in ticket #992):
sys/netinet/tcp_usrreq.c: revision 1.170
Update TCP CB with new values on rfc1323 and mssdflt sysctl updates
=46rom yasuoka-cj7TXg5MjN14Eiagz67IpQ@public.gmane.org in kern/44254
|
| 1.162.2.2 | 20-Oct-2013 |
bouyer | Pull up following revision(s) (requested by spz in ticket #967):
sys/netinet/tcp_usrreq.c: revision 1.168
PR/48098: Brian Marcotte: Avoid kernel assertion for embryonic sockets that
don't have credentials yet.
XXX: pullup-6
|
| 1.162.2.1 | 17-Mar-2012 |
bouyer | branches: 1.162.2.1.4; 1.162.2.1.6;
Pull up following revision(s) (requested by jruoho in ticket #124):
sys/netinet/tcp_usrreq.c: revision 1.163
PR/46077: M. Nunberg: Stat should not fial on connecting socket.
|
| 1.162.2.1.6.1 | 20-Oct-2013 |
bouyer | Pull up following revision(s) (requested by spz in ticket #967):
sys/netinet/tcp_usrreq.c: revision 1.168
PR/48098: Brian Marcotte: Avoid kernel assertion for embryonic sockets that
don't have credentials yet.
XXX: pullup-6
|
| 1.162.2.1.4.1 | 20-Oct-2013 |
bouyer | Pull up following revision(s) (requested by spz in ticket #967):
sys/netinet/tcp_usrreq.c: revision 1.168
PR/48098: Brian Marcotte: Avoid kernel assertion for embryonic sockets that
don't have credentials yet.
XXX: pullup-6
|
| 1.165.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.165.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.165.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.166.4.5 | 18-May-2014 |
rmind | sync with head
|
| 1.166.4.4 | 17-Oct-2013 |
rmind | Eliminate some of the splsoftnet() calls, misc clean up.
|
| 1.166.4.3 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.166.4.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.166.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.171.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.200.2.4 | 29-Jan-2019 |
msaitoh | Pull up following revision(s) (requested by martin in ticket #1676):
sys/net/link_proto.c 1.37
sys/netatalk/ddp_usrreq.c 1.72
sys/netbt/hci_socket.c 1.46
sys/netbt/l2cap_socket.c 1.36
sys/netbt/rfcomm_socket.c 1.38
sys/netbt/sco_socket.c 1.38
sys/netinet/tcp_usrreq.c 1.223 via patch
sys/netinet6/raw_ip6.c 1.173
sys/netinet6/udp6_usrreq.c 1.146
sys/netmpls/mpls_proto.c 1.32
sys/netnatm/natm.c patch
Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.200.2.3 | 11-Sep-2017 |
snj | Pull up following revision(s) (requested by jdolecek in ticket #1498):
sys/netinet/tcp_usrreq.c: revision 1.216
add some more getsockopt(2) params
|
| 1.200.2.2 | 21-Feb-2015 |
martin | branches: 1.200.2.2.2; 1.200.2.2.6;
Pull up following revision(s) (requested by he in ticket #530):
sys/netinet/tcp_output.c: revision 1.180
sys/netinet/tcp_input.c: revision 1.336
sys/netinet/tcp_usrreq.c: revision 1.203
share/man/man4/tcp.4: revision 1.30
sys/netinet/tcp.h: revision 1.31
sys/netinet/tcp_subr.c: revision 1.258
sys/netinet/tcp_var.h: revision 1.176
sys/netinet/tcp_var.h: revision 1.177
sys/sys/param.h: bump revision
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.200.2.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.200.2.2.6.1 | 29-Jan-2019 |
msaitoh | Pull up following revision(s) (requested by martin in ticket #1676):
sys/net/link_proto.c 1.37
sys/netatalk/ddp_usrreq.c 1.72
sys/netbt/hci_socket.c 1.46
sys/netbt/l2cap_socket.c 1.36
sys/netbt/rfcomm_socket.c 1.38
sys/netbt/sco_socket.c 1.38
sys/netinet/tcp_usrreq.c 1.223 via patch
sys/netinet6/raw_ip6.c 1.173
sys/netinet6/udp6_usrreq.c 1.146
sys/netmpls/mpls_proto.c 1.32
sys/netnatm/natm.c patch
Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.200.2.2.2.1 | 29-Jan-2019 |
msaitoh | Pull up following revision(s) (requested by martin in ticket #1676):
sys/net/link_proto.c 1.37
sys/netatalk/ddp_usrreq.c 1.72
sys/netbt/hci_socket.c 1.46
sys/netbt/l2cap_socket.c 1.36
sys/netbt/rfcomm_socket.c 1.38
sys/netbt/sco_socket.c 1.38
sys/netinet/tcp_usrreq.c 1.223 via patch
sys/netinet6/raw_ip6.c 1.173
sys/netinet6/udp6_usrreq.c 1.146
sys/netmpls/mpls_proto.c 1.32
sys/netnatm/natm.c patch
Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.202.2.7 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.202.2.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.202.2.5 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.202.2.4 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.202.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.202.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.202.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.212.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.212.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.213.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.214.6.2 | 29-Jan-2019 |
msaitoh | Pull up following revision(s) (requested by martin in ticket #1175):
sys/net/link_proto.c 1.37
sys/netatalk/ddp_usrreq.c 1.72
sys/netbt/hci_socket.c 1.46
sys/netbt/l2cap_socket.c 1.36
sys/netbt/rfcomm_socket.c 1.38
sys/netbt/sco_socket.c 1.38
sys/netinet/sctp_usrreq.c 1.14
sys/netinet/tcp_usrreq.c 1.223
sys/netinet6/raw_ip6.c 1.173
sys/netinet6/sctp6_usrreq.c 1.17
sys/netinet6/udp6_usrreq.c 1.146
sys/netmpls/mpls_proto.c 1.32
sys/netnatm/natm.c patch
Fix memory leaks pointed out by Ilja Van Sprundel: all
sendoob() functions are expted to free both passed
mbuf chains.
|
| 1.214.6.1 | 25-Aug-2017 |
snj | Pull up following revision(s) (requested by jdolecek in ticket #216):
sys/netinet/tcp_usrreq.c: revision 1.216
add some more getsockopt(2) params
|
| 1.216.2.5 | 26-Dec-2018 |
pgoyette | Sync with HEAD, resolve a few conflicts
|
| 1.216.2.4 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.216.2.3 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.216.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.216.2.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.219.2.3 | 21-Apr-2020 |
martin | Sync with HEAD
|
| 1.219.2.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.219.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.224.4.1 | 10-Sep-2019 |
martin | Pull up following revision(s) (requested by maxv in ticket #193):
sys/netinet/tcp_timer.h: revision 1.30
sys/netinet/tcp_input.c: revision 1.415
sys/netinet/tcp_usrreq.c: revision 1.225
sys/netinet/tcp_subr.c: revision 1.283
Clamp tcp timer quantities to reasonable ranges.
|
| 1.225.6.1 | 20-Apr-2020 |
bouyer | Sync with HEAD
|
| 1.227.2.2 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.227.2.1 | 14-Dec-2020 |
thorpej | Sync w/ HEAD.
|
| 1.199 | 03-Dec-2024 |
andvar | s/packlets/packets/ in comment.
|
| 1.198 | 28-Oct-2022 |
ozaki-r | branches: 1.198.8;
inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.197 | 20-Sep-2022 |
ozaki-r | tcp: separate syn cache stuffs into tcp_syncache.[ch] files
No functional change.
|
| 1.196 | 31-Jul-2021 |
andvar | s/threshhold/threshold
|
| 1.195 | 08-Mar-2021 |
christos | branches: 1.195.4;
Remove the unused "addin" argument (it was always 0) and go back using
a random iss by default (instead of rfc1948)
|
| 1.194 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.193 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.192 | 05-Mar-2020 |
riastradh | branches: 1.192.4;
Revert "Include opt_diagnostic.h for DIAGNOSTIC."
This did not do what I thought it did. opt_diagnostic.h is only for
the unused _DIAGNOSTIC, which seems like an abortive attempt to
incrementally convert DIAGNOSTIC to an opt_*.h option rather than a
command-line option.
|
| 1.191 | 05-Mar-2020 |
riastradh | Include opt_diagnostic.h for DIAGNOSTIC.
...at least, in header files, which may not have already included
libkern.h.
|
| 1.190 | 27-Dec-2018 |
maxv | Remove unused arguments.
|
| 1.189 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.188 | 03-Sep-2018 |
riastradh | Rename min/max -> uimin/uimax for better honesty.
These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.
HOWEVER! Some subsystems have
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.
To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.
I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:
cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))
It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.
Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)
|
| 1.187 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.186 | 29-Apr-2018 |
maxv | branches: 1.186.2;
Move struct tcpiphdr from tcpip.h to tcp_var.h, to match UDP (udpiphdr in
udp_var.h).
tcpip.h is now empty, and can be removed.
|
| 1.185 | 28-Mar-2018 |
maxv | Remove two unused args from syn_cache_get().
|
| 1.184 | 12-Feb-2018 |
maxv | branches: 1.184.2;
Remove unused argument from tcp_signature_getsav.
|
| 1.183 | 12-Feb-2018 |
maxv | Remove the 'm' argument from syn_cache_respond(); all it does with it is
freeing it, so free in the caller instead.
|
| 1.182 | 19-Jan-2018 |
ozaki-r | Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
|
| 1.181 | 15-Nov-2017 |
ozaki-r | Make syn_cache_timer static
|
| 1.180 | 31-Jul-2017 |
maxv | Fix TCPCTL_NAMES, and remove TCPCTL_VARIABLES.
|
| 1.179 | 28-Jul-2017 |
maxv | Remove TCP_COMPAT_42. This feature is a workaround for a bug in the TCP
stack of BSD4.2. Having such features just does not make any sense, and
looking at the code, I'm not sure it actually works.
|
| 1.178 | 07-Jul-2017 |
ozaki-r | Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
|
| 1.177 | 14-Feb-2015 |
he | branches: 1.177.10;
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.176 | 14-Feb-2015 |
he | Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
|
| 1.175 | 31-Jul-2014 |
rtr | branches: 1.175.2; 1.175.4;
split PRU_DISCONNECT, PRU_SHUTDOWN and PRU_ABORT function out of
pr_generic() usrreq switches and put into separate functions
xxx_disconnect(struct socket *)
xxx_shutdown(struct socket *)
xxx_abort(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req =
PRU_{DISCONNECT,SHUTDOWN,ABORT}
with calls to pr_{disconnect,shutdown,abort}() respectively
rename existing internal functions used to implement above functionality
to permit use of the names for xxx_{disconnect,shutdown,abort}().
- {l2cap,sco,rfcomm}_disconnect() ->
{l2cap,sco,rfcomm}_disconnect_pcb()
- {unp,rip,tcp}_disconnect() -> {unp,rip,tcp}_disconnect1()
- unp_shutdown() -> unp_shutdown1()
patch reviewed by rmind
|
| 1.174 | 19-May-2014 |
rmind | - Split off PRU_ATTACH and PRU_DETACH logic into separate functions.
- Replace malloc with kmem and eliminate M_PCB while here.
- Sprinkle more asserts.
|
| 1.173 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.172 | 02-Jan-2014 |
pooka | branches: 1.172.2;
Allow kernels compiled with INET+INET6 to be booted as IPv4-only or IPv6-only.
|
| 1.171 | 12-Nov-2013 |
kefren | * implement TCP CUBIC congestion control algorithm
* move tcp_sack_newack bits inside reno and newreno_fast_retransmit_newack
* notify ECN peer about cwnd shrink in [new]reno_slow_retransmit
Based on the patch proposed on tech-net@ on Nov 7 with minor improvments:
* adapt wmax for no-fast convergence case
* correct cbrt calculation for big window sizes (>750KB)
|
| 1.170 | 10-Apr-2013 |
christos | branches: 1.170.4;
Limit the tcp initial window setting to 10, leaving it by default to 4
and simplifying the code in process. Per draft-ietf-initcwnd-08.txt.
|
| 1.169 | 02-Feb-2012 |
tls | branches: 1.169.6;
Entropy-pool implementation move and cleanup.
1) Move core entropy-pool code and source/sink/sample management code
to sys/kern from sys/dev.
2) Remove use of NRND as test for presence of entropy-pool code throughout
source tree.
3) Remove use of RND_ENABLED in device drivers as microoptimization to
avoid expensive operations on disabled entropy sources; make the
rnd_add calls do this directly so all callers benefit.
4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might
have lead to slight entropy overestimation for some sources.
5) Add new source types for environmental sensors, power sensors, VM
system events, and skew between clocks, with a sample implementation
for each.
ok releng to go in before the branch due to the difficulty of later
pullup (widespread #ifdef removal and moved files). Tested with release
builds on amd64 and evbarm and live testing on amd64.
|
| 1.168 | 31-Oct-2011 |
yamt | branches: 1.168.2; 1.168.6;
tcp_reass_unlock: assertion
|
| 1.167 | 25-May-2011 |
gdt | Add comment urging a separation of TCP_RTT_SHIFT into separate defines
describing the EWMA calculation and the storage representation.
(No code change.)
|
| 1.166 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.165 | 03-May-2011 |
dyoung | *_drain() routines may be called with locks held, so instead of doing
any work in *_drain(), set a drain-needed flag. Do the work in the
fasttimo handler.
Contributed by Coyote Point Systems, Inc.
|
| 1.164 | 20-Apr-2011 |
gdt | Rewrite comments about TCP RTO calculations.
Long ago, the storage representations of srtt and rttvar were changed
from the 4.4BSD scheme, and the comments are out of sync with the
code. This commit rewrites most of the comments that explain the RTO
calculations, and points out some issues in the code.
Joint work with Bev Schwartz of BBN (original analysis and comments),
but I have rewritten and extended them, so errors are mine.
This material is based upon work supported by the Defense Advanced
Research Projects Agency and Space and Naval Warfare Systems Center,
Pacific, under Contract No. N66001-09-C-2073. Approved for Public
Release, Distribution Unlimited
|
| 1.163 | 14-Apr-2011 |
yamt | comments
|
| 1.162 | 16-Sep-2009 |
pooka | branches: 1.162.4; 1.162.6;
Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.161 | 09-Sep-2009 |
darran | Make tcp msl (max segment life) tunable via sysctl net.inet.tcp.msl.
Okayed by tls@.
|
| 1.160 | 27-May-2009 |
pooka | POOL_INIT -> pool_init
|
| 1.159 | 29-Jan-2009 |
pooka | branches: 1.159.2;
stinkset purge: POOL_INIT -> pool_init
also, make the syncache pool static in scope
|
| 1.158 | 06-Aug-2008 |
plunky | branches: 1.158.2; 1.158.4; 1.158.10;
Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.157 | 28-Apr-2008 |
martin | branches: 1.157.2; 1.157.6;
Remove clause 3 and 4 from TNF licenses
|
| 1.156 | 24-Apr-2008 |
ad | branches: 1.156.2;
Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.155 | 12-Apr-2008 |
thorpej | branches: 1.155.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.154 | 08-Apr-2008 |
thorpej | Change TCP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old tcpstat structure; old netstat
binaries will continue to work properly.
|
| 1.153 | 29-Feb-2008 |
matt | Rework tcp congctl selection code so that the congctl entries can be const.
Don't access tcp_congctl stuff outside of tcp_congctl.c, use routines to
update t_congctl. This code is slightly now more complicated.
|
| 1.152 | 27-Feb-2008 |
matt | Convert stragglers to ansi definitions from old-style definitons.
Remember that func() is not ansi, func(void) is.
|
| 1.151 | 25-Dec-2007 |
perry | branches: 1.151.2; 1.151.6;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.150 | 02-Aug-2007 |
rmind | branches: 1.150.4; 1.150.10; 1.150.12; 1.150.16; 1.150.20;
TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.149 | 09-Jul-2007 |
ad | branches: 1.149.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.148 | 25-Jun-2007 |
christos | tcpdrop kernel bits (from anon ymous)
|
| 1.147 | 20-Jun-2007 |
christos | - per socket keepalive settings
- settable connection establishment timeout
|
| 1.146 | 02-May-2007 |
dyoung | Eliminate address family-specific route caches (struct route, struct
route_in6, struct route_iso), replacing all caches with a struct
route.
The principle benefit of this change is that all of the protocol
families can benefit from route cache-invalidation, which is
necessary for correct routing. Route-cache invalidation fixes an
ancient PR, kern/3508, at long last; it fixes various other PRs,
also.
Discussions with and ideas from Joerg Sonnenberger influenced this
work tremendously. Of course, all design oversights and bugs are
mine.
DETAILS
1 I added to each address family a pool of sockaddrs. I have
introduced routines for allocating, copying, and duplicating,
and freeing sockaddrs:
struct sockaddr *sockaddr_alloc(sa_family_t af, int flags);
struct sockaddr *sockaddr_copy(struct sockaddr *dst,
const struct sockaddr *src);
struct sockaddr *sockaddr_dup(const struct sockaddr *src, int flags);
void sockaddr_free(struct sockaddr *sa);
sockaddr_alloc() returns either a sockaddr from the pool belonging
to the specified family, or NULL if the pool is exhausted. The
returned sockaddr has the right size for that family; sa_family
and sa_len fields are initialized to the family and sockaddr
length---e.g., sa_family = AF_INET and sa_len = sizeof(struct
sockaddr_in). sockaddr_free() puts the given sockaddr back into
its family's pool.
sockaddr_dup() and sockaddr_copy() work analogously to strdup()
and strcpy(), respectively. sockaddr_copy() KASSERTs that the
family of the destination and source sockaddrs are alike.
The 'flags' argumet for sockaddr_alloc() and sockaddr_dup() is
passed directly to pool_get(9).
2 I added routines for initializing sockaddrs in each address
family, sockaddr_in_init(), sockaddr_in6_init(), sockaddr_iso_init(),
etc. They are fairly self-explanatory.
3 structs route_in6 and route_iso are no more. All protocol families
use struct route. I have changed the route cache, 'struct route',
so that it does not contain storage space for a sockaddr. Instead,
struct route points to a sockaddr coming from the pool the sockaddr
belongs to. I added a new method to struct route, rtcache_setdst(),
for setting the cache destination:
int rtcache_setdst(struct route *, const struct sockaddr *);
rtcache_setdst() returns 0 on success, or ENOMEM if no memory is
available to create the sockaddr storage.
It is now possible for rtcache_getdst() to return NULL if, say,
rtcache_setdst() failed. I check the return value for NULL
everywhere in the kernel.
4 Each routing domain (struct domain) has a list of live route
caches, dom_rtcache. rtflushall(sa_family_t af) looks up the
domain indicated by 'af', walks the domain's list of route caches
and invalidates each one.
|
| 1.145 | 04-Mar-2007 |
christos | branches: 1.145.2; 1.145.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.144 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.143 | 06-Dec-2006 |
yamt | branches: 1.143.2;
add some more tcp mowners.
|
| 1.142 | 06-Dec-2006 |
yamt | - make tcp_reass static.
- constify.
|
| 1.141 | 21-Oct-2006 |
yamt | branches: 1.141.2; 1.141.4;
- constify.
- make tcp_dooptions and tcpipqent_pool static.
|
| 1.140 | 19-Oct-2006 |
yamt | implement RFC3465 appropriate byte counting.
from Kentaro A. Kurahone, with minor adjustments by me.
the ack prediction part of the original patch was omitted because
it's a separate change. reviewed by Rui Paulo.
|
| 1.139 | 16-Oct-2006 |
rpaulo | Export the tcp_do_rfc1948 variable to userland via sysctl.
The code to generate an ISS via an MD5 hash has been present in the
NetBSD kernel since 2001, but it wasn't even exported to userland at
that time. It was agreed on tech-net with the original author <thorpej>
that we should let the user decide if he wants to enable it or not.
Not enabled by default.
|
| 1.138 | 09-Oct-2006 |
rpaulo | Modular (I tried ;-) TCP congestion control API. Whenever certain conditions
happen in the TCP stack, this interface calls the specified callback to
handle the situation according to the currently selected congestion
control algorithm.
A new sysctl node was created: net.inet.tcp.congctl.{available,selected}
with obvious meanings.
The old net.inet.tcp.newreno MIB was removed.
The API is discussed in tcp_congctl(9).
In the near future, it will be possible to selected a congestion control
algorithm on a per-socket basis.
Discussed on tech-net and reviewed by <yamt>.
|
| 1.137 | 05-Sep-2006 |
rpaulo | branches: 1.137.2; 1.137.4;
Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.
Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.
|
| 1.136 | 22-Jul-2006 |
rpaulo | revert stuff that shouldn't have gone in.
|
| 1.135 | 22-Jul-2006 |
rpaulo | TCP RFC is 793, not 783.
|
| 1.134 | 16-Feb-2006 |
perry | branches: 1.134.2;
Change "inline" back to "__inline" in .h files -- C99 is still too
new, and some apps compile things in C89 mode. C89 keywords stay.
As per core@.
|
| 1.133 | 24-Dec-2005 |
perry | branches: 1.133.2; 1.133.4; 1.133.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.132 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.131 | 10-Dec-2005 |
elad | Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.130 | 06-Sep-2005 |
rpaulo | Implement tcp.inet{,6}.tcp{,6}.(debug|debx) when TCP_DEBUG is set. They
can be used to ``transliterate protocol trace'' like trpt(8) does.
|
| 1.129 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.128 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.127 | 19-Jul-2005 |
christos | Implement PMTU checks from:
http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.
From OpenBSD.
|
| 1.126 | 29-May-2005 |
christos | branches: 1.126.2;
- add const
- remove bogus casts
- avoid nested variables
|
| 1.125 | 05-Apr-2005 |
kurahone | Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.124 | 29-Mar-2005 |
yamt | protect tcpipqent with splvm.
|
| 1.123 | 16-Mar-2005 |
yamt | branches: 1.123.2;
simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]),
build it directly from the segment list when needed.
|
| 1.122 | 16-Mar-2005 |
yamt | - use full sized segments unless we actually have SACKs to send.
- avoid TSO duplicate D-SACK.
- send SACKs regardless of TF_ACKNOW.
- don't clear rcv_sack_num when transmitting.
discussed on tech-net@.
|
| 1.121 | 09-Mar-2005 |
atatat | gc the tcp_sysctl() prototype since it's completely vestigial
|
| 1.120 | 02-Mar-2005 |
mycroft | Copyright maintenance.
|
| 1.119 | 28-Feb-2005 |
jonathan | Commit TCP SACK patches from Kentaro A. Karahone's patch at:
http://www.sigusr1.org/~kurahone/tcp-sack-netbsd-02152005.diff.gz
Fixes in that patch for pre-existing TCP pcb initializations were already
committed to NetBSD-current, so are not included in this commit.
The SACK patch has been observed to correctly negotiate and respond,
to SACKs in wide-area traffic.
There are two indepenently-observed, as-yet-unresolved anomalies:
First, seeing unexplained delays between in fast retransmission
(potentially explainable by an 0.2sec RTT between adjacent
ethernet/wifi NICs); and second, peculiar and unepxlained TCP
retransmits observed over an ath0 card.
After discussion with several interested developers, I'm committing
this now, as-is, for more eyes to use and look over. Current hypothesis
is that the anomalies above may in fact be due to link/level (hardware,
driver, HAL, firmware) abberations in the test setup, affecting both
Kentaro's wired-Ethernet NIC and in my two (different) WiFi NICs.
|
| 1.118 | 06-Feb-2005 |
pk | Update tcp_trace() prototype to match implementation.
|
| 1.117 | 27-Jan-2005 |
mycroft | Introduce a new state variable, t_partialacks. It has 3 states:
* t_partialacks<0 means we are not in fast recovery.
* t_partialacks==0 means we are in fast recovery, but we have not received
any partial acks yet.
* t_partialacks>0 means we are in fast recovery, and we have received
partial acks.
This is used to implement 2 changes in RFC 3782:
* We keep the notion that we are in fast recovery separate from t_dupacks, so
it is not reset due to out-of-order acks. (This affects both the Reno and
NewReno cases.)
* We only reset the retransmit timer on the first partial ack -- preventing us
from possibly taking one RTO per segment once fast recovery is initiated.
As before, it is hard to measure any difference between Reno and NewReno in the
real-world cases that I've tested.
|
| 1.116 | 26-Jan-2005 |
mycroft | Fix two problems in our TCP stack:
1) If an echoed RFC 1323 time stamp appears to be later than the current time,
ignore it and fall back to old-style RTT calculation. This prevents ending
up with a negative RTT and panicking later.
2) Fix NewReno. This involves a few changes:
a) Implement the send_high variable in RFC 2582. Our implementation is
subtly different; it is one *past* the last sequence number transmitted
rather than being equal to it. This simplifies some logic and makes
the code smaller. Additional logic was required to prevent sequence
number wraparound problems; this is not mentioned in RFC 2582.
b) Make sure we reset t_dupacks on new acks, but *not* on a partial ack.
All of the new ack code is pushed out into tcp_newreno(). (Later this
will probably be a pluggable function.) Thus t_dupacks keeps track of
whether we're in fast recovery all the time, with Reno or NewReno, which
keeps some logic simpler.
c) We do not need to update snd_recover when we're not in fast recovery.
See tech-net for an explanation of this.
d) In the gratuitous fast retransmit prevention case, do not send a packet.
RFC 2582 specifically says that we should "do nothing".
e) Do not inflate the congestion window on a partial ack. (This is done by
testing t_dupacks to see whether we're still in fast recovery.)
This brings the performance of NewReno back up to the same as Reno in a few
random test cases (e.g. transferring peer-to-peer over my wireless network).
I have not concocted a good test case for the behavior specific to NewReno.
|
| 1.115 | 21-Dec-2004 |
yamt | branches: 1.115.2; 1.115.4;
factor out receive side tcp/udp checksum handling code so that they
can be used by eg. packet filters.
reviewed by Christos Zoulas on tech-net@.
(slightly tweaked since then to make tcp and udp similar.)
|
| 1.114 | 15-Dec-2004 |
thorpej | Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.113 | 15-Sep-2004 |
yamt | fix ipqent pool corruption problems. make tcp reass code use
its own pool of ipqent rather than sharing it with ip reass code.
PR/24782.
|
| 1.112 | 18-May-2004 |
itojun | fix MD5 signature support to actually validate inbound signature, and
drop packet if fails.
|
| 1.111 | 26-Apr-2004 |
itojun | make TCP MD5 signature work with KAME IPSEC (#define IPSEC).
support IPv6 if KAME IPSEC (RFC is not explicit about how we make data stream
for checksum with IPv6, but i'm pretty sure using normal pseudo-header is the
right thing).
XXX
current TCP MD5 signature code has giant flaw:
it does not validate signature on input (can't believe it! what is the point?)
|
| 1.110 | 25-Apr-2004 |
jonathan | Initial commit of a port of the FreeBSD implementation of RFC 2385
(MD5 signatures for TCP, as used with BGP). Credit for original
FreeBSD code goes to Bruce M. Simpson, with FreeBSD sponsorship
credited to sentex.net. Shortening of the setsockopt() name
attributed to Vincent Jardin.
This commit is a minimal, working version of the FreeBSD code, as
MFC'ed to FreeBSD-4. It has received minimal testing with a ttcp
modified to set the TCP-MD5 option; BMS's additions to tcpdump-current
(tcpdump -M) confirm that the MD5 signatures are correct. Committed
as-is for further testing between a NetBSD BGP speaker (e.g., quagga)
and industry-standard BGP speakers (e.g., Cisco, Juniper).
NOTE: This version has two potential flaws. First, I do see any code
that verifies recieved TCP-MD5 signatures. Second, the TCP-MD5
options are internally padded and assumed to be 32-bit aligned. A more
space-efficient scheme is to pack all TCP options densely (and
possibly unaligned) into the TCP header ; then do one final padding to
a 4-byte boundary. Pre-existing comments note that accounting for
TCP-option space when we add SACK is yet to be done. For now, I'm
punting on that; we can solve it properly, in a way that will handle
SACK blocks, as a separate exercise.
In case a pullup to NetBSD-2 is requested, this adds sys/netipsec/xform_tcp.c
,and modifies:
sys/net/pfkeyv2.h,v 1.15
sys/netinet/files.netinet,v 1.5
sys/netinet/ip.h,v 1.25
sys/netinet/tcp.h,v 1.15
sys/netinet/tcp_input.c,v 1.200
sys/netinet/tcp_output.c,v 1.109
sys/netinet/tcp_subr.c,v 1.165
sys/netinet/tcp_usrreq.c,v 1.89
sys/netinet/tcp_var.h,v 1.109
sys/netipsec/files.netipsec,v 1.3
sys/netipsec/ipsec.c,v 1.11
sys/netipsec/ipsec.h,v 1.7
sys/netipsec/key.c,v 1.11
share/man/man4/tcp.4,v 1.16
lib/libipsec/pfkey.c,v 1.20
lib/libipsec/pfkey_dump.c,v 1.17
lib/libipsec/policy_token.l,v 1.8
sbin/setkey/parse.y,v 1.14
sbin/setkey/setkey.8,v 1.27
sbin/setkey/token.l,v 1.15
Note that the preceding two revisions to tcp.4 will be
required to cleanly apply this diff.
|
| 1.109 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.108 | 20-Apr-2004 |
itojun | - respond to RST by ACK, as suggested in NISCC recommendation
- rate-limit ACKs against RSTs and SYNs
|
| 1.107 | 18-Apr-2004 |
matt | De __P()
|
| 1.106 | 22-Oct-2003 |
thorpej | branches: 1.106.2;
Rather than zeroing a tcpcb structure and filling in all the fields
individually, create a tcpcb template pre-initialized (and pre-zero'd)
with the static and mostly-static tcpcb parameters. The template is
now copied into the new tcpcb, which zeros and initializes most of the
tcpcb in one pass. The template is kept up-to-date as TCP sysctl
variables are changed.
Combined with the previous sb_max change, TCP socket creation is now
25% faster.
|
| 1.105 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.104 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.103 | 20-Jul-2003 |
he | As a temporary workaround, apply the fix from PR#20390, thereby
cooperating with the callout code in working around the race
condition caused by the TCP code's use of the callout facility.
Instead of unconditionally releasing memory in tcp_close() and
SYN_CACHE_PUT(), check whether any of the related callout handlers
are about to be invoked (but have not yet done callout_ack()), and
if so, just mark the associated data structure (tcpcb or syn cache
entry) as "dead", and test for this (and release storage) in the
callout handler functions.
|
| 1.102 | 29-Jun-2003 |
fvdl | branches: 1.102.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.101 | 29-Jun-2003 |
ragge | Add code to remember where in the send queue of mbufs the last packet was
sent from. This change avoid a linear search through all mbufs when using
large TCP windows, and therefore permit high-speed connections on long
distances.
Tested on a 1 Gigabit connection between Lule� and San Francisco, a distance
of about 15000km. With TCP windows of just over 20 Mbytes it could keep up
with 950Mbit/s.
After discussions with Matt Thomas and Jason Thorpe.
|
| 1.100 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.99 | 26-Jun-2003 |
christos | abuse the mib instead of abusing the new pointer. Idea from simon burge.
It allows the tcp_sysctl_ident to run by non-super-users. No backwards
compatibility provided.
|
| 1.98 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.97 | 19-Apr-2003 |
christos | PR/2352: Tor Egge: Add sysctl to get uid of connected socket.
|
| 1.96 | 01-Mar-2003 |
thorpej | Allow TCP connections to hosts on a local network to use a larger
slow start initial window. Default this larger initial window to
4 packets, allowing it to be adjusted with net.inet.tcp.init_win_local.
|
| 1.95 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.94 | 02-Nov-2002 |
perry | /*CONTCOND*/ while (0)'ed macros
|
| 1.93 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.92 | 09-Jun-2002 |
itojun | whitespace
|
| 1.91 | 26-May-2002 |
itojun | path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.90 | 12-May-2002 |
matt | branches: 1.90.2; 1.90.4;
Eliminate commons.
|
| 1.89 | 15-Mar-2002 |
itojun | have tcp6_drain
|
| 1.88 | 24-Jan-2002 |
itojun | place NRL copyright notice itself, not a reference to it.
|
| 1.87 | 11-Sep-2001 |
thorpej | Use callouts for SYN cache timers, rather than traversing time queues
in tcp_slowtimo().
|
| 1.86 | 10-Sep-2001 |
thorpej | Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.
|
| 1.85 | 10-Sep-2001 |
thorpej | Split tcp_timers() into multiple functions, one for each timer,
and call it directly from tcp_slowtimo() (via a table) rather
than going through tcp_userreq().
This will allow us to call TCP timers directly from callouts,
in a future revision.
|
| 1.84 | 10-Sep-2001 |
thorpej | Change the way receive idle time and round trip time are measured.
Instead of incrementing t_idle and t_rtt in tcp_slowtimo(), we now
take a timstamp (via tcp_now) and use subtraction to compute the
delta when we actually need it (using unsigned arithmetic so that
tcp_now wrapping is handled correctly).
Based on similar changes in FreeBSD.
|
| 1.83 | 10-Sep-2001 |
thorpej | Use a callout for the delayed ACK timer, and delete tcp_fasttimo().
Expose the delayed ACK timer as net.inet.tcp.delack_ticks.
|
| 1.82 | 31-Jul-2001 |
thorpej | branches: 1.82.2;
Count the number of times we "self-quench" (ip_output() returns
ENOBUFS), and don't inline tcp_segsize() if profiling.
|
| 1.81 | 30-May-2001 |
mrg | branches: 1.81.2;
use _KERNEL_OPT
|
| 1.80 | 26-May-2001 |
matt | Make t_flags a u_int instead of u_short. It's followed by a mbuf pointer
so there's padding around it already. And it increases the amount of bits
available for TF_* flags.
|
| 1.79 | 13-Apr-2001 |
thorpej | Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.
|
| 1.78 | 20-Mar-2001 |
thorpej | Two changes, designed to make us even more resilient against TCP
ISS attacks (which we already fend off quite well).
1. First-cut implementation of RFC1948, Steve Bellovin's cryptographic
hash method of generating TCP ISS values. Note, this code is experimental
and disabled by default (experimental enough that I don't export the
variable via sysctl yet, either). There are a couple of issues I'd
like to discuss with Steve, so this code should only be used by people
who really know what they're doing.
2. Per a recent thread on Bugtraq, it's possible to determine a system's
uptime by snooping the RFC1323 TCP timestamp options sent by a host; in
4.4BSD, timestamps are created by incrementing the tcp_now variable
at 2 Hz; there's even a company out there that uses this to determine
web server uptime. According to Newsham's paper "The Problem With
Random Increments", while NetBSD's TCP ISS generation method is much
better than the "random increment" method used by FreeBSD and OpenBSD,
it is still theoretically possible to mount an attack against NetBSD's
method if the attacker knows how many times the tcp_iss_seq variable
has been incremented. By not leaking uptime information, we can make
that much harder to determine. So, we avoid the leak by giving each
TCP connection a timebase of 0.
|
| 1.77 | 19-Oct-2000 |
itojun | branches: 1.77.2;
remove #ifdef TCP6. it is not likely for us to bring in sys/netinet6/tcp6*.c
(separate TCP/IPv6 stack) into netbsd-current.
|
| 1.76 | 18-Oct-2000 |
thorpej | Restructure the Path MTU Discovery code somewhat to avoid
entering rtentry's for hosts we're not actually communicating
with.
Do this by invoking the ctlinput for the protocol, which is
responsible for validating the ICMP message:
* TCP -- Lookup the connection based on the address/port
pairs in the ICMP message.
* AH/ESP -- Lookup the SA based on the SPI in the ICMP message.
If validation succeeds, ctlinput is responsible for calling
icmp_mtudisc(). icmp_mtudisc() then invokes callbacks registered
by protocols (such as TCP) which want to take some sort of special
action when a path's MTU changes. For TCP, this is where we now
refresh cached routes and re-enter slow-start.
As a side-effect, this fixes the problem where TCP would not be
notified when a path's MTU changed if AH/ESP were being used.
XXX Note, this is only a fix for the IPv4 case. For the IPv6
XXX case, we need to wait for the KAME folks.
Reviewed by sommerfeld@netbsd.org and itojun@netbsd.org.
|
| 1.75 | 15-Aug-2000 |
itojun | net.inet.tcp.rstratelimit is deprecated. make it invalid and return
ENOPROTOOPT.
|
| 1.74 | 28-Jul-2000 |
itojun | nuke the following sysctl variables. "ppsratelimit" should work better.
need to recompile sbin/sysctl after updating /usr/include.
net.inet.tcp.rstratelimit
net.inet.icmp.errratelimit
net.inet6.icmp6.errratelimit
|
| 1.73 | 27-Jul-2000 |
itojun | implement net.inet.tcp.rstppslimit to limit TCP RSTs by packet-per-second
basis. default: 100pps
set default value for net.inet.tcp.rstratelimit to 0 (disabled),
NOTE: it does not work right for smaller-than-1/hz interval. maybe we should
nuke it, or make it impossible to set smaller-than-1/hz value.
|
| 1.72 | 15-Feb-2000 |
thorpej | branches: 1.72.4;
Add support for rate-limiting RSTs sent in response to no socket for
an incoming packet. Default minimum interval is 10ms. The interval
is changeable via the "net.inet.tcp.rstratelimit" sysctl variable.
|
| 1.71 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.70 | 08-Dec-1999 |
itojun | do not drop from IP header to tcp option until sbappend(), to reduce
requirement to mbuf chain.
part of KAME sync, committed separately for its (possible) impact.
|
| 1.69 | 19-Nov-1999 |
bouyer | Update protocoles and interfaces stats counters to 64bit.
RTM_IFINFO is now 0xf, 0xe is RTM_OIFINFO which returns the old (if_msghdr14)
struct with 32bit counters (binary compat, conditioned on COMPAT_14).
Same for sysctl: node 3 is renamed NET_RT_OIFLIST, NET_RT_IFLIST is now node 4.
Change rt_msg1() to add an mbuf to the mbuf chain instead of just panic()
when the message is larger than MHLEN.
|
| 1.68 | 23-Sep-1999 |
itojun | branches: 1.68.2; 1.68.8;
cleanup and correct TCP MSS consideration with IPsec headers.
MSS advertisement must always be:
max(if mtu) - ip hdr siz - tcp hdr siz
We violated this in the previous code so it was fixed.
tcp_mss_to_advertise() now takes af (af on wire) as its argument,
to compute right ip hdr siz.
tcp_segsize() will take care of IPsec header size.
One thing I'm not really sure is how to handle IPsec header size in
*rxsegsizep (inbound segment size estimation).
The current code subtracts possible *outbound* IPsec size from *rxsegsizep,
hoping that the peer is using the same IPsec policy as me.
It may not be applicable, could TCP gulu please comment...
|
| 1.67 | 25-Aug-1999 |
itojun | When listening socket goes away, remove assockated syn cache entires.
Stale syn cache entries are useless because none of them will be used
if there is no listening socket, as tcp_input looks up listening socket by
in_pcblookup*() before looking into syn cache.
This fixes race condition due to dangling socket pointer from syn cache
entries to listening socket (this was introduced when ipsec is merged in).
This should preserve currently implemented behavior (but not 4.4BSD
behavior prior to syn cache).
Tested in KAME repository before commit, but we'd better run some
regression tests.
|
| 1.66 | 12-Aug-1999 |
itojun | fix sototcpcb(). this sometimes caused panic on OOB data reception.
the macro may need to be expanded into dedicated function, rather than a macro,
to capture unsupported values.
|
| 1.65 | 31-Jul-1999 |
itojun | sync with recent KAME.
- loosen ipsec restriction on packet diredtion.
- revise icmp6 redirect handling on IsRouter bit.
- tcp/udp notification processing (link-local address case)
- cosmetic fixes (better code share across *BSD).
|
| 1.64 | 22-Jul-1999 |
itojun | - implement IPv6 pmtud, which is necessary for TCP6.
- fix memory leak on SO_DEBUG over TCP.
|
| 1.63 | 14-Jul-1999 |
itojun | Use proper ip protocol # field and tcp hdr on sending RST against SYN,
when ip header and tcp header are not adjacent to each other
(i.e. when ip6 options are attached).
To test this, try
telnet @::1@::1 port
toward a port without responding server. Prior to the fix, the kernel will
generate broken RST packet.
|
| 1.62 | 09-Jul-1999 |
thorpej | defopt INET6, and put it in opt_inet.h (most places already include this
file, which is why the file list is so short).
|
| 1.61 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.60 | 23-May-1999 |
ad | Add new sysctl (net.inet.tcp.log_refused) that when set, causes refused TCP
connections to be logged.
|
| 1.59 | 29-Apr-1999 |
thorpej | Implement retransmit logic for the SYN cache engine. Fixes a rare condition
where one side can think a connection exists, where the other side thinks
the connection was never established.
The original problem was first reported by Ty Sarna in PR #5909. The
original fix I made to the code didn't cover all cases. The problem this
fix addresses was reported by Christoph Badura via private e-mail.
Many thanks to Bill Sommerfeld for helping me to test this code, and
for finding a subtle bug.
|
| 1.58 | 24-Jan-1999 |
thorpej | branches: 1.58.2;
Oops, forgot to update copyright notice in previous.
|
| 1.57 | 24-Jan-1999 |
thorpej | * Completely rewrite syn_cache_respond().
- Don't use tcp_respond(), instead create the tcp/ip header from scratch,
and send it ourself.
- Reuse the mbuf that carried the SYN, or allocate one if that is not
available.
- Cache the route we look up to do the Path MTU Discovery check, and
transfer the reference to that route to the inpcb when the connection
completes.
* Macro'ize a small, but often repeated code fragment.
|
| 1.56 | 18-Dec-1998 |
thorpej | Add a lock around the TCPCB's sequence queue, to prevent tcp_drain()
from corrupting the queue if called from a device's interrupt context.
Similar in nature to the problem reported in PR #5684.
|
| 1.55 | 06-Oct-1998 |
matt | Add a sysctl for newreno (default to off).
|
| 1.54 | 04-Oct-1998 |
matt | Adapt the NEWRENO changes from the UCSB diffs of BSDI 3.0's TCP
to NetBSD. Ignore the SACK & FACK stuff for now.
|
| 1.53 | 10-Sep-1998 |
mouse | Create tcp.keepidle, tcp.keepintvl, tcp.keepcnt, tcp.slowhz sysctls.
|
| 1.52 | 09-Sep-1998 |
thorpej | Use an algorithm similar to that in tcp_notify() to determine if
syn_cache_unreach() should remove the entry, or just continue on.
Algorithm is to only remove the entry if we've had more than one unreach
error and have retransmitted 3 or more times. This prevents the following
scenario, as noted in PR #5909 (PR from Ty Sarna, scenario from
Charles Hannum):
* Host A sends a SYN.
* Host A retransmits the SYN.
* Host B gets the first SYN and sends a SYN-ACK.
* Host B gets the second SYN and sends a SYN-ACK.
* One of the SYN-ACK bounces with an
ICMP unreachable, causing the `SYN cache' entry to be
removed with no notification.
* Host A receives the other SYN-ACK, sends an ACK, and goes to
ESTABLISHED state.
Should fix PR #5909.
|
| 1.51 | 21-Jul-1998 |
mycroft | Implement a better fix for the `gratuitous FIN' problem, as
mentioned on tcp-impl but with a bit more commentary.
|
| 1.50 | 11-May-1998 |
thorpej | Nuke TUBA per my note to tech-net; there's no reason to keep it around.
|
| 1.49 | 07-May-1998 |
thorpej | Rework the syn cache code somewhat:
- Don't use home-grown queue manipulation. Use <sys/queue.h> instead. The
data structures are a little larger, but we are otherwise wasting the
memory chunk anyway (we're already a 64-byte malloc bucket).
- Fix a bug in the cache-is-full case: if the oldest element removed from
the first non-empty bucket was the only element in the bucket, the
bucket wouldn't be removed from the bucket cache, causing queue corruption
later.
- Optimize the syn cache timers by using PRT timers rather than home-grown
decrement-and-propagate timers.
This code is now a fair bit smaller, and significantly easier to read
and understand.
|
| 1.48 | 06-May-1998 |
thorpej | Use the monotonically increasing slow timer timestamp provided by
the protocol dispatch layer for TCP timers. This saves having to
modify a potentially large number of timer values (which were shorts,
and expanded to ... a lot of code on the Alpha).
|
| 1.47 | 02-May-1998 |
thorpej | Reintroduce the immediate ACK-on-PUSH behavior removed in revision 1.47,
but make the decision to do this dependent on the sysctl variable
net.inet.tcp.ack_on_push, which is disabled by default.
|
| 1.46 | 01-May-1998 |
thorpej | Garbage-collect.
|
| 1.45 | 30-Apr-1998 |
thorpej | In the CWM code, don't use the Floyd initial window computation as
the burst size allowed, but rather a fixed number of packets, as
described in the Internet Draft. Default allowed burst is 4 packets,
per the Draft.
Make the use of CWM and the allowed burst size tunable via sysctl.
|
| 1.44 | 30-Apr-1998 |
thorpej | Make tcp_compat_42 a sysctl option.
|
| 1.43 | 29-Apr-1998 |
matt | New TCP reassembly code. The new code reduces the memory needed by
out-of-order packets and builds the infrastructure needed for sending
SACK blocks (to be added shortly).
|
| 1.42 | 29-Apr-1998 |
thorpej | Make use of the work-arounds for ancient broken TCP peers run-time
conditional (tcp_compat_42). The kernel config option TCP_COMPAT_42
will still enable this by default, or disable this by default if the
option is not included (i.e. current behavior). This will be made a
sysctl soon.
|
| 1.41 | 13-Apr-1998 |
kml | Fix to ensure that the correct MSS is advertised for loopback
TCP connections by using the MTU of the interface. Also added
a knob, mss_ifmtu, to force all connections to use the MTU of
the interface to calculate the advertised MSS.
|
| 1.40 | 07-Apr-1998 |
thorpej | Remember any source routes that may have accompanied a SYN.
|
| 1.39 | 03-Apr-1998 |
thorpej | Now that we have a flags word in the syn cache entry, use a flag to indicate
"peer will do timestamps" rather than a bitfield, and give the now-unsed
bit to the hash, making it now 32 bits.
|
| 1.38 | 03-Apr-1998 |
thorpej | Clean up some comments wrt. the syn cache code.
|
| 1.37 | 31-Mar-1998 |
thorpej | Fix a potential-congestion case in the larger initial congestion window
code, as clarified in the TCPIMPL WG meeting at IETF #41: If the SYN
(active open) or SYN,ACK (passive open) was retransmitted, the initial
congestion window for the first slow start of that connection must be
one segment.
|
| 1.36 | 17-Mar-1998 |
kml | Ensure that the TCP segment size reflects the size of TCP options
in the packet. This fixes a bug that was resulting in extra packets
in retransmissions (the second packet would be 12 bytes long,
reflecting the RFC1323 timestamp option size).
|
| 1.35 | 19-Feb-1998 |
thorpej | Update copyright (sigh, should have done this long ago).
|
| 1.34 | 10-Feb-1998 |
perry | add/cleanup multiple inclusion protection.
|
| 1.33 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.32 | 31-Dec-1997 |
thorpej | Implement a queue for delayed ACK processing. This queue is used in
tcp_fasttimo() in lieu of scanning all open TCP connections.
|
| 1.31 | 17-Dec-1997 |
thorpej | Keep stats on connections dropped due to excessive persist timeout.
|
| 1.30 | 13-Dec-1997 |
thorpej | After further examination of traces of bulk transfers (with help from
Kevin Lahey), undo the "defer window update until next delayed ACK".
|
| 1.29 | 11-Dec-1997 |
thorpej | Implement an infrastructure to allow larger initial congestion windows.
The sysctl'able variable "tcp_init_win", when set to 0, selects an
auto-tuning algorithm for selecting the initial window, based on transmit
segment size, per discussion in the IETF tcpimpl working group.
Default initial window is still 1 segment, but will soon become 2 segments,
per discussion in tcpimpl.
|
| 1.28 | 11-Dec-1997 |
thorpej | In the PRU_RCVD entry point, if TF_DELACK is set, don't send the window
update now, since it will be sent within 200ms when the delayed ACK is
sent. Instrument how many hits we get on this optimization.
|
| 1.27 | 10-Dec-1997 |
thorpej | Implement tcp_drain().
|
| 1.26 | 08-Nov-1997 |
kml | TCP MSS fixes to provide cleaner slow-start and recovery.
|
| 1.25 | 17-Oct-1997 |
kml | branches: 1.25.2;
Path MTU Discovery support. This is turned off by default.
Use sysctl -w net.inet.icmp.mtudisc=1 to turn on.
Still to come: path removal after some period, black hole detection
|
| 1.24 | 10-Oct-1997 |
explorer | Add hooks to use the kernel random system to generate TCP sequence numbers.
|
| 1.23 | 22-Sep-1997 |
thorpej | Fix several annoyances related to MSS handling in BSD TCP:
- Don't overload t_maxseg. Previous behavior was to set it to the min
of the peer's advertised MSS, our advertised MSS, and tcp_mssdflt
(for non-local networks). This breaks PMTU discovery running on
either host. Instead, remember the MSS we advertise, and use it
as appropriate (in silly window avoidance).
- Per last bullet, split tcp_mss() into several functions for handling
MSS (ours and peer's), and performing various tasks when a connection
becomes ESTABLISHED.
- Introduce a new function, tcp_segsize(), which computes the max size
for every segment transmitted in tcp_output(). This will eventually
be used to hook in PMTU discovery.
|
| 1.22 | 29-Aug-1997 |
gwr | Tweaks to allow operation with an interface address of 0.0.0.0
(needed for NFS mountroot using BOOTP to get boot parameters)
|
| 1.21 | 28-Jul-1997 |
thorpej | branches: 1.21.2;
Make the following tunable via sysctl, inspired by BSD/OS:
- tcp_sendspace
- tcp_recvspace
- tcp_mssdflt
- tcp_syn_cache_limit
- tcp_syn_bucket_limit
- tcp_syn_cache_timer
|
| 1.20 | 23-Jul-1997 |
thorpej | Pull SYN_cache_branch down into the main line.
|
| 1.19 | 10-Dec-1996 |
mycroft | branches: 1.19.8;
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.18 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.17 | 13-Feb-1996 |
christos | branches: 1.17.4;
netinet prototypes
|
| 1.16 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.15 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.14 | 30-Sep-1995 |
thorpej | branches: 1.14.2;
Implement tcp_sysctl(). Add a sysctl option to enable/disable RFC1323
extensions to TCP. From John Kohl <jtk@kolvir.blrc.ma.us>.
|
| 1.13 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.12 | 11-Jun-1995 |
mycroft | As suggested by Brakmo and Peterson:
* Don't add the extra 1/8 of the mss when ramping up the congestion window.
* Scale the RTT values slightly to adjust for rounding errors.
* Set the lower bound of the RTO to RTT+2.
|
| 1.11 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.10 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.9 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.8 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.7 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.6 | 10-Jan-1994 |
mycroft | Don't prototype this until it's safe.
|
| 1.5 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.4 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.14.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.17.4.2 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.17.4.1 | 10-Dec-1996 |
mycroft | From trunk:
Fix RTT scaling problems introduced with Brakmo and Peterson changes.
|
| 1.19.8.6 | 16-Jul-1997 |
thorpej | Declare struct tcp_opt_info here; it's needed by tuba_tcpinput().
|
| 1.19.8.5 | 29-Jun-1997 |
thorpej | Instrument syn cache hash collisions.
|
| 1.19.8.4 | 28-Jun-1997 |
thorpej | KNF.
|
| 1.19.8.3 | 28-Jun-1997 |
thorpej | Use explicit type sizes in struct cyn_cache, and add a comment about
this structure being larger than intended on the Alpha.
|
| 1.19.8.2 | 26-Jun-1997 |
thorpej | tcp_mss() needs to take a u_int, not a u_int16_t.
|
| 1.19.8.1 | 14-May-1997 |
mellon | More of David Borman's SYN cache patches for Lite2:
- Define syn_cache entry and syn_cache_head structures.
- Add syn_cache statistics to tcpstat structure.
- Declare externs for syn cache variables.
- Update prototypes: tcp_dooptions, tcp_mss, tcp_respond.
- Add prototypes for syn_cache_* functions.
|
| 1.21.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.21.2.2 | 29-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.21.2.1 | 01-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.25.2.4 | 09-May-1998 |
mycroft | Pull up patch from kml.
|
| 1.25.2.3 | 05-May-1998 |
mycroft | Pull up 1.36, per request of kml.
|
| 1.25.2.2 | 29-Jan-1998 |
mellon | Pull up 1.27-1.33 (thorpej)
|
| 1.25.2.1 | 08-Nov-1997 |
thorpej | Pull up from trunk: TCP MSS fixes to provide cleaner slow-start and recovery.
(kml)
|
| 1.58.2.1 | 29-Apr-1999 |
perry | branches: 1.58.2.1.2; 1.58.2.1.4;
pullup 1.58->1.59 (thorpej)
|
| 1.58.2.1.4.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.58.2.1.4.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.58.2.1.4.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.58.2.1.2.3 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.58.2.1.2.2 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.58.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.68.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.68.2.3 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.68.2.2 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.68.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.72.4.3 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #143)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.72.4.2 | 24-Jan-2002 |
he | Pull up revision 1.88 (requested by itojun):
Clean up the NRL copyright.
|
| 1.72.4.1 | 16-Aug-2000 |
itojun | pullup (approved by releng-1-5)
switch from net.inet*.*.*ratelimit to net.inet*.*.ppslimit.
(tags are rough estimate - we had some try-and-error in main trunc)
sys/netinet/icmp6.h 1.9 -> 1.11
sys/netinet/icmp_var.h 1.15 -> 1.17
sys/netinet/in_proto.c 1.39 -> 1.42
sys/netinet/ip_icmp.c 1.50 -> 1.51, 1.52 -> 1.54
sys/netinet/tcp_input.c 1.111 -> 1.112, 1.115 -> 1.117
sys/netinet/tcp_usrreq.c 1.52 -> 1.53
sys/netinet/tcp_var.h 1.72 -> 1.75
sys/netinet6/icmp6.c 1.34 -> 1.35, 1.36 -> 1.38
sys/netinet6/in6_proto.c 1.17 -> 1.19
|
| 1.77.2.9 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.77.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.77.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.77.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.77.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.77.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.77.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.77.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.77.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.81.2.5 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.81.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.81.2.3 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.81.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.81.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.82.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.90.4.3 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #1680)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.90.4.2 | 22-Oct-2003 |
jmc | Pullup rev 1.03 (requested by he in ticket #1530)
Introduce a new INVOKING status for callouts, and use it to close
a race condition in the TCP code. Fixes PR#20390.
|
| 1.90.4.1 | 05-Sep-2003 |
tron | Pull up revision 1.91 (requested by tls in ticket #1445):
path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)
|
| 1.90.2.3 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.90.2.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.90.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.102.2.12 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.102.2.11 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.102.2.10 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.102.2.9 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.102.2.8 | 07-Feb-2005 |
skrll | Sunc with HEAD.
|
| 1.102.2.7 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.102.2.6 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.102.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.102.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.102.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.102.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.102.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.106.2.2 | 18-Sep-2004 |
he | Pull up revision 1.113 (requested by yamt in ticket #861):
Fix ipqent pool corruption problems. Make the TCP reassembly
code use its own pool of ipqent rather than sharing it with
the IP reassembly code. Fixes PR#24782.
|
| 1.106.2.1 | 20-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #169)
If a segment is received with RST set and the segment is completely to the
left of the receive window, ignore it. Add some additional comments to
the code that deals with received segemnts that are completely to the right
of the receive window. If an invalid SYN is received, force an ACK and
drop it; if the other side really sent the SYN; it'll respond with a reset.
Respond to RST by ACK, as suggested in NISCC recommendation.
Rate-limit ACKs against RSTs and SYNs.
If SYN is coming and RCV.NXT == SEG.SEQ, then ACK with value - 1.
|
| 1.115.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.115.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.115.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.123.2.2 | 06-May-2005 |
tron | Pull up revision 1.125 (requested by kurahone in ticket #199):
Added sysctl tunable limits for the number of maximum SACK holes
per connection and per system.
Idea taken from FreeBSD.
|
| 1.123.2.1 | 04-Apr-2005 |
tron | Pull up revision 1.124 (requested by yamt in ticket #90):
protect tcpipqent with splvm.
|
| 1.126.2.6 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.126.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.126.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.126.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.126.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.126.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.133.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.133.4.3 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.133.4.2 | 14-Mar-2006 |
rpaulo | Remove in6pcb in parameter list.
|
| 1.133.4.1 | 14-Mar-2006 |
rpaulo | Remove back pointer to in6pcb.
|
| 1.133.2.1 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.134.2.2 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.134.2.1 | 11-Aug-2006 |
yamt | sync with head
|
| 1.137.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.137.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.137.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.137.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.141.4.1 | 03-Jun-2008 |
skrll | Sync with netbsd-4.
|
| 1.141.2.1 | 21-Jan-2008 |
bouyer | Pull up following revision(s) (requested by ghen in ticket #1039):
sys/netinet/tcp_var.h: revision 1.148
distrib/sets/lists/comp/mi: revision 1.1035
distrib/sets/lists/man/mi: revision 1.1010
usr.sbin/tcpdrop/Makefile: revision 1.1
usr.sbin/tcpdrop/tcpdrop.c: revision 1.1 - 1.3
usr.sbin/tcpdrop/tcpdrop.8: revision 1.1
usr.sbin/Makefile: revision 1.228 via patch
sys/netinet/tcp_usrreq.c: revision 1.133
distrib/sets/lists/base/mi: revision 1.712
Import tcpdrop(8) from OpenBSD
|
| 1.143.2.3 | 07-May-2007 |
yamt | sync with head.
|
| 1.143.2.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.143.2.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.145.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.145.2.4 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.145.2.3 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.145.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.145.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.149.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.150.20.2 | 02-Aug-2007 |
rmind | TCP socket buffers automatic sizing - ported from FreeBSD.
http://mail-index.netbsd.org/tech-net/2007/02/04/0006.html
! Disabled by default, marked as experimental. Testers are very needed.
! Someone should thoroughly test this, and improve if possible.
Discussed on <tech-net>:
http://mail-index.netbsd.org/tech-net/2007/07/12/0002.html
Thanks Greg Troxel for comments.
OK by the long silence on <tech-net>.
|
| 1.150.20.1 | 02-Aug-2007 |
rmind | file tcp_var.h was added on branch matt-mips64 on 2007-08-02 02:42:43 +0000
|
| 1.150.16.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.150.12.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.150.10.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.150.4.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.150.4.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.151.6.3 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.151.6.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.151.6.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.151.2.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.155.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.156.2.5 | 11-Mar-2010 |
yamt | sync with head
|
| 1.156.2.4 | 16-Sep-2009 |
yamt | sync with head
|
| 1.156.2.3 | 20-Jun-2009 |
yamt | sync with head
|
| 1.156.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.156.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.157.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.157.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.158.10.1 | 21-Apr-2010 |
matt | sync to netbsd-5
|
| 1.158.4.1 | 26-Sep-2009 |
snj | Pull up following revision(s) (requested by darran in ticket #950):
sys/netinet/tcp_input.c: revision 1.299
sys/netinet/tcp_usrreq.c: revision 1.156
sys/netinet/tcp_var.h: revision 1.161
Make tcp msl (max segment life) tunable via sysctl net.inet.tcp.msl.
Okayed by tls@.
|
| 1.158.2.1 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.159.2.1 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.162.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.162.4.2 | 31-May-2011 |
rmind | sync with head
|
| 1.162.4.1 | 21-Apr-2011 |
rmind | sync with head
|
| 1.168.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.168.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.168.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.169.6.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.169.6.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.169.6.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.170.4.3 | 18-May-2014 |
rmind | sync with head
|
| 1.170.4.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.170.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.172.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.175.4.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.175.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.175.2.1 | 21-Feb-2015 |
martin | Pull up following revision(s) (requested by he in ticket #530):
sys/netinet/tcp_output.c: revision 1.180
sys/netinet/tcp_input.c: revision 1.336
sys/netinet/tcp_usrreq.c: revision 1.203
share/man/man4/tcp.4: revision 1.30
sys/netinet/tcp.h: revision 1.31
sys/netinet/tcp_subr.c: revision 1.258
sys/netinet/tcp_var.h: revision 1.176
sys/netinet/tcp_var.h: revision 1.177
sys/sys/param.h: bump revision
Port over the TCP_INFO socket option from FreeBSD, originally from
the Linux 2.6 TCP API. This permits the caller to query certain information
about a TCP connection, and is used by pkgsrc's net/iperf3 test program
if available.
This extends struct tcbcb with three fields to count retransmits,
out-of-sequence receives and zero window announcements, and will
therefore warrant a kernel revision bump (done separately).
Change the new counter variables in struct tcpcb to uint32_t, as
per christos' comments.
|
| 1.177.10.2 | 03-Feb-2018 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #514):
sys/net/route.c: 1.205
sys/net/rtsock.c: 1.237-1.238
sys/netinet/in.c: 1.215
sys/netinet/tcp_subr.c: 1.272
sys/netinet/tcp_timer.c: 1.93
sys/netinet/tcp_timer.h: 1.29
sys/netinet/tcp_var.h: 1.182
sys/netinet6/in6.c: 1.258
Remove extra pserialize_perform from in_purgeaddr
It's already performed in ifa_remove. Note so there (in in6_unlink_ifa too).
Release rt_so_mtx on updating a rtentry to avoid a deadlock with route_intr
The deadlock happened only if NET_MPSAFE on.
Run tcp_slowtimo in workqueue if NET_MPSAFE
If NET_MPSAFE is enabled, we have to avoid taking softnet_lock in softint as
much as possible to prevent any softint handlers including callout handlers
such as tcp_slowtimo from sticking on softnet_lock because it results in
undesired delays of executing subsequent softint handlers.
NFCI for !NET_MPSAFE
Fix a return value of rt_update_prepare
Callers expect it to be an errno.
Fix another deadlock
When waiting for a route update to finish, a waiter has to release its reference
to the route to avoid a deadlock. Because a updater tries to wait for references
to a target route (except for a reference by the updater itself) to be released.
|
| 1.177.10.1 | 21-Oct-2017 |
snj | Pull up following revision(s) (requested by ozaki-r in ticket #300):
crypto/dist/ipsec-tools/src/setkey/parse.y: 1.19
crypto/dist/ipsec-tools/src/setkey/token.l: 1.20
distrib/sets/lists/tests/mi: 1.754, 1.757, 1.759
doc/TODO.smpnet: 1.12-1.13
sys/net/pfkeyv2.h: 1.32
sys/net/raw_cb.c: 1.23-1.24, 1.28
sys/net/raw_cb.h: 1.28
sys/net/raw_usrreq.c: 1.57-1.58
sys/net/rtsock.c: 1.228-1.229
sys/netinet/in_proto.c: 1.125
sys/netinet/ip_input.c: 1.359-1.361
sys/netinet/tcp_input.c: 1.359-1.360
sys/netinet/tcp_output.c: 1.197
sys/netinet/tcp_var.h: 1.178
sys/netinet6/icmp6.c: 1.213
sys/netinet6/in6_proto.c: 1.119
sys/netinet6/ip6_forward.c: 1.88
sys/netinet6/ip6_input.c: 1.181-1.182
sys/netinet6/ip6_output.c: 1.193
sys/netinet6/ip6protosw.h: 1.26
sys/netipsec/ipsec.c: 1.100-1.122
sys/netipsec/ipsec.h: 1.51-1.61
sys/netipsec/ipsec6.h: 1.18-1.20
sys/netipsec/ipsec_input.c: 1.44-1.51
sys/netipsec/ipsec_netbsd.c: 1.41-1.45
sys/netipsec/ipsec_output.c: 1.49-1.64
sys/netipsec/ipsec_private.h: 1.5
sys/netipsec/key.c: 1.164-1.234
sys/netipsec/key.h: 1.20-1.32
sys/netipsec/key_debug.c: 1.18-1.21
sys/netipsec/key_debug.h: 1.9
sys/netipsec/keydb.h: 1.16-1.20
sys/netipsec/keysock.c: 1.59-1.62
sys/netipsec/keysock.h: 1.10
sys/netipsec/xform.h: 1.9-1.12
sys/netipsec/xform_ah.c: 1.55-1.74
sys/netipsec/xform_esp.c: 1.56-1.72
sys/netipsec/xform_ipcomp.c: 1.39-1.53
sys/netipsec/xform_ipip.c: 1.50-1.54
sys/netipsec/xform_tcp.c: 1.12-1.16
sys/rump/librump/rumpkern/Makefile.rumpkern: 1.170
sys/rump/librump/rumpnet/net_stub.c: 1.27
sys/sys/protosw.h: 1.67-1.68
tests/net/carp/t_basic.sh: 1.7
tests/net/if_gif/t_gif.sh: 1.11
tests/net/if_l2tp/t_l2tp.sh: 1.3
tests/net/ipsec/Makefile: 1.7-1.9
tests/net/ipsec/algorithms.sh: 1.5
tests/net/ipsec/common.sh: 1.4-1.6
tests/net/ipsec/t_ipsec_ah_keys.sh: 1.2
tests/net/ipsec/t_ipsec_esp_keys.sh: 1.2
tests/net/ipsec/t_ipsec_gif.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_l2tp.sh: 1.6-1.7
tests/net/ipsec/t_ipsec_misc.sh: 1.8-1.18
tests/net/ipsec/t_ipsec_sockopt.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tcp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_transport.sh: 1.5-1.6
tests/net/ipsec/t_ipsec_tunnel.sh: 1.9
tests/net/ipsec/t_ipsec_tunnel_ipcomp.sh: 1.1-1.2
tests/net/ipsec/t_ipsec_tunnel_odd.sh: 1.3
tests/net/mcast/t_mcast.sh: 1.6
tests/net/net/t_ipaddress.sh: 1.11
tests/net/net_common.sh: 1.20
tests/net/npf/t_npf.sh: 1.3
tests/net/route/t_flags.sh: 1.20
tests/net/route/t_flags6.sh: 1.16
usr.bin/netstat/fast_ipsec.c: 1.22
Do m_pullup before mtod
It may fix panicks of some tests on anita/sparc and anita/GuruPlug.
---
KNF
---
Enable DEBUG for babylon5
---
Apply C99-style struct initialization to xformsw
---
Tweak outputs of netstat -s for IPsec
- Get rid of "Fast"
- Use ipsec and ipsec6 for titles to clarify protocol
- Indent outputs of sub protocols
Original outputs were organized like this:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
(Fast) IPsec:
IPsec ah:
IPsec esp:
IPsec ipip:
IPsec ipcomp:
New outputs are organized like this:
ipsec:
ah:
esp:
ipip:
ipcomp:
ipsec6:
ah:
esp:
ipip:
ipcomp:
---
Add test cases for IPComp
---
Simplify IPSEC_OSTAT macro (NFC)
---
KNF; replace leading whitespaces with hard tabs
---
Introduce and use SADB_SASTATE_USABLE_P
---
KNF
---
Add update command for testing
Updating an SA (SADB_UPDATE) requires that a process issuing
SADB_UPDATE is the same as a process issued SADB_ADD (or SADB_GETSPI).
This means that update command must be used with add command in a
configuration of setkey. This usage is normally meaningless but
useful for testing (and debugging) purposes.
---
Add test cases for updating SA/SP
The tests require newly-added udpate command of setkey.
---
PR/52346: Frank Kardel: Fix checksumming for NAT-T
See XXX for improvements.
---
Remove codes for PACKET_TAG_IPSEC_IN_CRYPTO_DONE
It seems that PACKET_TAG_IPSEC_IN_CRYPTO_DONE is for network adapters
that have IPsec accelerators; a driver sets the mtag to a packet
when its device has already encrypted the packet.
Unfortunately no driver implements such offload features for long
years and seems unlikely to implement them soon. (Note that neither
FreeBSD nor Linux doesn't have such drivers.) Let's remove related
(unused) codes and simplify the IPsec code.
---
Fix usages of sadb_msg_errno
---
Avoid updating sav directly
On SADB_UPDATE a target sav was updated directly, which was unsafe.
Instead allocate another sav, copy variables of the old sav to
the new one and replace the old one with the new one.
---
Simplify; we can assume sav->tdb_xform cannot be NULL while it's valid
---
Rename key_alloc* functions (NFC)
We shouldn't use the term "alloc" for functions that just look up
data and actually don't allocate memory.
---
Use explicit_memset to surely zero-clear key_auth and key_enc
---
Make sure to clear keys on error paths of key_setsaval
---
Add missing KEY_FREESAV
---
Make sure a sav is inserted to a sah list after its initialization completes
---
Remove unnecessary zero-clearing codes from key_setsaval
key_setsaval is now used only for a newly-allocated sav. (It was
used to reset variables of an existing sav.)
---
Correct wrong assumption of sav->refcnt in key_delsah
A sav in a list is basically not to be sav->refcnt == 0. And also
KEY_FREESAV assumes sav->refcnt > 0.
---
Let key_getsavbyspi take a reference of a returning sav
---
Use time_mono_to_wall (NFC)
---
Separate sending message routine (NFC)
---
Simplify; remove unnecessary zero-clears
key_freesaval is used only when a target sav is being destroyed.
---
Omit NULL checks for sav->lft_c
sav->lft_c can be NULL only when initializing or destroying sav.
---
Omit unnecessary NULL checks for sav->sah
---
Omit unnecessary check of sav->state
key_allocsa_policy picks a sav of either MATURE or DYING so we
don't need to check its state again.
---
Simplify; omit unnecessary saidx passing
- ipsec_nextisr returns a saidx but no caller uses it
- key_checkrequest is passed a saidx but it can be gotton by
another argument (isr)
---
Fix splx isn't called on some error paths
---
Fix header size calculation of esp where sav is NULL
---
Fix header size calculation of ah in the case sav is NULL
This fix was also needed for esp.
---
Pass sav directly to opencrypto callback
In a callback, use a passed sav as-is by default and look up a sav
only if the passed sav is dead.
---
Avoid examining freshness of sav on packet processing
If a sav list is sorted (by lft_c->sadb_lifetime_addtime) in advance,
we don't need to examine each sav and also don't need to delete one
on the fly and send up a message. Fortunately every sav lists are sorted
as we need.
Added key_validate_savlist validates that each sav list is surely sorted
(run only if DEBUG because it's not cheap).
---
Add test cases for SAs with different SPIs
---
Prepare to stop using isr->sav
isr is a shared resource and using isr->sav as a temporal storage
for each packet processing is racy. And also having a reference from
isr to sav makes the lifetime of sav non-deterministic; such a reference
is removed when a packet is processed and isr->sav is overwritten by
new one. Let's have a sav locally for each packet processing instead of
using shared isr->sav.
However this change doesn't stop using isr->sav yet because there are
some users of isr->sav. isr->sav will be removed after the users find
a way to not use isr->sav.
---
Fix wrong argument handling
---
fix printf format.
---
Don't validate sav lists of LARVAL or DEAD states
We don't sort the lists so the validation will always fail.
Fix PR kern/52405
---
Make sure to sort the list when changing the state by key_sa_chgstate
---
Rename key_allocsa_policy to key_lookup_sa_bysaidx
---
Separate test files
---
Calculate ah_max_authsize on initialization as well as esp_max_ivlen
---
Remove m_tag_find(PACKET_TAG_IPSEC_PENDING_TDB) because nobody sets the tag
---
Restore a comment removed in previous
The comment is valid for the below code.
---
Make tests more stable
sleep command seems to wait longer than expected on anita so
use polling to wait for a state change.
---
Add tests that explicitly delete SAs instead of waiting for expirations
---
Remove invalid M_AUTHIPDGM check on ESP isr->sav
M_AUTHIPDGM flag is set to a mbuf in ah_input_cb. An sav of ESP can
have AH authentication as sav->tdb_authalgxform. However, in that
case esp_input and esp_input_cb are used to do ESP decryption and
AH authentication and M_AUTHIPDGM never be set to a mbuf. So
checking M_AUTHIPDGM of a mbuf on isr->sav of ESP is meaningless.
---
Look up sav instead of relying on unstable sp->req->sav
This code is executed only in an error path so an additional lookup
doesn't matter.
---
Correct a comment
---
Don't release sav if calling crypto_dispatch again
---
Remove extra KEY_FREESAV from ipsec_process_done
It should be done by the caller.
---
Don't bother the case of crp->crp_buf == NULL in callbacks
---
Hold a reference to an SP during opencrypto processing
An SP has a list of isr (ipsecrequest) that represents a sequence
of IPsec encryption/authentication processing. One isr corresponds
to one opencrypto processing. The lifetime of an isr follows its SP.
We pass an isr to a callback function of opencrypto to continue
to a next encryption/authentication processing. However nobody
guaranteed that the isr wasn't freed, i.e., its SP wasn't destroyed.
In order to avoid such unexpected destruction of isr, hold a reference
to its SP during opencrypto processing.
---
Don't make SAs expired on tests that delete SAs explicitly
---
Fix a debug message
---
Dedup error paths (NFC)
---
Use pool to allocate tdb_crypto
For ESP and AH, we need to allocate an extra variable space in addition
to struct tdb_crypto. The fixed size of pool items may be larger than
an actual requisite size of a buffer, but still the performance
improvement by replacing malloc with pool wins.
---
Don't use unstable isr->sav for header size calculations
We may need to optimize to not look up sav here for users that
don't need to know an exact size of headers (e.g., TCP segmemt size
caclulation).
---
Don't use sp->req->sav when handling NAT-T ESP fragmentation
In order to do this we need to look up a sav however an additional
look-up degrades performance. A sav is later looked up in
ipsec4_process_packet so delay the fragmentation check until then
to avoid an extra look-up.
---
Don't use key_lookup_sp that depends on unstable sp->req->sav
It provided a fast look-up of SP. We will provide an alternative
method in the future (after basic MP-ification finishes).
---
Stop setting isr->sav on looking up sav in key_checkrequest
---
Remove ipsecrequest#sav
---
Stop setting mtag of PACKET_TAG_IPSEC_IN_DONE because there is no users anymore
---
Skip ipsec_spi_*_*_preferred_new_timeout when running on qemu
Probably due to PR 43997
---
Add localcount to rump kernels
---
Remove unused macro
---
Fix key_getcomb_setlifetime
The fix adjusts a soft limit to be 80% of a corresponding hard limit.
I'm not sure the fix is really correct though, at least the original
code is wrong. A passed comb is zero-cleared before calling
key_getcomb_setlifetime, so
comb->sadb_comb_soft_addtime = comb->sadb_comb_soft_addtime * 80 / 100;
is meaningless.
---
Provide and apply key_sp_refcnt (NFC)
It simplifies further changes.
---
Fix indentation
Pointed out by knakahara@
---
Use pslist(9) for sptree
---
Don't acquire global locks for IPsec if NET_MPSAFE
Note that the change is just to make testing easy and IPsec isn't MP-safe yet.
---
Let PF_KEY socks hold their own lock instead of softnet_lock
Operations on SAD and SPD are executed via PF_KEY socks. The operations
include deletions of SAs and SPs that will use synchronization mechanisms
such as pserialize_perform to wait for references to SAs and SPs to be
released. It is known that using such mechanisms with holding softnet_lock
causes a dead lock. We should avoid the situation.
---
Make IPsec SPD MP-safe
We use localcount(9), not psref(9), to make the sptree and secpolicy (SP)
entries MP-safe because SPs need to be referenced over opencrypto
processing that executes a callback in a different context.
SPs on sockets aren't managed by the sptree and can be destroyed in softint.
localcount_drain cannot be used in softint so we delay the destruction of
such SPs to a thread context. To do so, a list to manage such SPs is added
(key_socksplist) and key_timehandler_spd deletes dead SPs in the list.
For more details please read the locking notes in key.c.
Proposed on tech-kern@ and tech-net@
---
Fix updating ipsec_used
- key_update_used wasn't called in key_api_spddelete2 and key_api_spdflush
- key_update_used wasn't called if an SP had been added/deleted but
a reply to userland failed
---
Fix updating ipsec_used; turn on when SPs on sockets are added
---
Add missing IPsec policy checks to icmp6_rip6_input
icmp6_rip6_input is quite similar to rip6_input and the same checks exist
in rip6_input.
---
Add test cases for setsockopt(IP_IPSEC_POLICY)
---
Don't use KEY_NEWSP for dummy SP entries
By the change KEY_NEWSP is now not called from softint anymore
and we can use kmem_zalloc with KM_SLEEP for KEY_NEWSP.
---
Comment out unused functions
---
Add test cases that there are SPs but no relevant SAs
---
Don't allow sav->lft_c to be NULL
lft_c of an sav that was created by SADB_GETSPI could be NULL.
---
Clean up clunky eval strings
- Remove unnecessary \ at EOL
- This allows to omit ; too
- Remove unnecessary quotes for arguments of atf_set
- Don't expand $DEBUG in eval
- We expect it's expanded on execution
Suggested by kre@
---
Remove unnecessary KEY_FREESAV in an error path
sav should be freed (unreferenced) by the caller.
---
Use pslist(9) for sahtree
---
Use pslist(9) for sah->savtree
---
Rename local variable newsah to sah
It may not be new.
---
MP-ify SAD slightly
- Introduce key_sa_mtx and use it for some list operations
- Use pserialize for some list iterations
---
Introduce KEY_SA_UNREF and replace KEY_FREESAV with it where sav will never be actually freed in the future
KEY_SA_UNREF is still key_freesav so no functional change for now.
This change reduces diff of further changes.
---
Remove out-of-date log output
Pointed out by riastradh@
---
Use KDASSERT instead of KASSERT for mutex_ownable
Because mutex_ownable is too heavy to run in a fast path
even for DIAGNOSTIC + LOCKDEBUG.
Suggested by riastradh@
---
Assemble global lists and related locks into cache lines (NFCI)
Also rename variable names from *tree to *list because they are
just lists, not trees.
Suggested by riastradh@
---
Move locking notes
---
Update the locking notes
- Add locking order
- Add locking notes for misc lists such as reglist
- Mention pserialize, key_sp_ref and key_sp_unref on SP operations
Requested by riastradh@
---
Describe constraints of key_sp_ref and key_sp_unref
Requested by riastradh@
---
Hold key_sad.lock on SAVLIST_WRITER_INSERT_TAIL
---
Add __read_mostly to key_psz
Suggested by riastradh@
---
Tweak wording (pserialize critical section => pserialize read section)
Suggested by riastradh@
---
Add missing mutex_exit
---
Fix setkey -D -P outputs
The outputs were tweaked (by me), but I forgot updating libipsec
in my local ATF environment...
---
MP-ify SAD (key_sad.sahlist and sah entries)
localcount(9) is used to protect key_sad.sahlist and sah entries
as well as SPD (and will be used for SAD sav).
Please read the locking notes of SAD for more details.
---
Introduce key_sa_refcnt and replace sav->refcnt with it (NFC)
---
Destroy sav only in the loop for DEAD sav
---
Fix KASSERT(solocked(sb->sb_so)) failure in sbappendaddr that is called eventually from key_sendup_mbuf
If key_sendup_mbuf isn't passed a socket, the assertion fails.
Originally in this case sb->sb_so was softnet_lock and callers
held softnet_lock so the assertion was magically satisfied.
Now sb->sb_so is key_so_mtx and also softnet_lock isn't always
held by callers so the assertion can fail.
Fix it by holding key_so_mtx if key_sendup_mbuf isn't passed a socket.
Reported by knakahara@
Tested by knakahara@ and ozaki-r@
---
Fix locking notes of SAD
---
Fix deadlock between key_sendup_mbuf called from key_acquire and localcount_drain
If we call key_sendup_mbuf from key_acquire that is called on packet
processing, a deadlock can happen like this:
- At key_acquire, a reference to an SP (and an SA) is held
- key_sendup_mbuf will try to take key_so_mtx
- Some other thread may try to localcount_drain to the SP with
holding key_so_mtx in say key_api_spdflush
- In this case localcount_drain never return because key_sendup_mbuf
that has stuck on key_so_mtx never release a reference to the SP
Fix the deadlock by deferring key_sendup_mbuf to the timer
(key_timehandler).
---
Fix that prev isn't cleared on retry
---
Limit the number of mbufs queued for deferred key_sendup_mbuf
It's easy to be queued hundreds of mbufs on the list under heavy
network load.
---
MP-ify SAD (savlist)
localcount(9) is used to protect savlist of sah. The basic design is
similar to MP-ifications of SPD and SAD sahlist. Please read the
locking notes of SAD for more details.
---
Simplify ipsec_reinject_ipstack (NFC)
---
Add per-CPU rtcache to ipsec_reinject_ipstack
It reduces route lookups and also reduces rtcache lock contentions
when NET_MPSAFE is enabled.
---
Use pool_cache(9) instead of pool(9) for tdb_crypto objects
The change improves network throughput especially on multi-core systems.
---
Update
ipsec(4), opencrypto(9) and vlan(4) are now MP-safe.
---
Write known issues on scalability
---
Share a global dummy SP between PCBs
It's never be changed so it can be pre-allocated and shared safely between PCBs.
---
Fix race condition on the rawcb list shared by rtsock and keysock
keysock now protects itself by its own mutex, which means that
the rawcb list is protected by two different mutexes (keysock's one
and softnet_lock for rtsock), of course it's useless.
Fix the situation by having a discrete rawcb list for each.
---
Use a dedicated mutex for rt_rawcb instead of softnet_lock if NET_MPSAFE
---
fix localcount leak in sav. fixed by ozaki-r@n.o.
I commit on behalf of him.
---
remove unnecessary comment.
---
Fix deadlock between pserialize_perform and localcount_drain
A typical ussage of localcount_drain looks like this:
mutex_enter(&mtx);
item = remove_from_list();
pserialize_perform(psz);
localcount_drain(&item->localcount, &cv, &mtx);
mutex_exit(&mtx);
This sequence can cause a deadlock which happens for example on the following
situation:
- Thread A calls localcount_drain which calls xc_broadcast after releasing
a specified mutex
- Thread B enters the sequence and calls pserialize_perform with holding
the mutex while pserialize_perform also calls xc_broadcast
- Thread C (xc_thread) that calls an xcall callback of localcount_drain tries
to hold the mutex
xc_broadcast of thread B doesn't start until xc_broadcast of thread A
finishes, which is a feature of xcall(9). This means that pserialize_perform
never complete until xc_broadcast of thread A finishes. On the other hand,
thread C that is a callee of xc_broadcast of thread A sticks on the mutex.
Finally the threads block each other (A blocks B, B blocks C and C blocks A).
A possible fix is to serialize executions of the above sequence by another
mutex, but adding another mutex makes the code complex, so fix the deadlock
by another way; the fix is to release the mutex before pserialize_perform
and instead use a condvar to prevent pserialize_perform from being called
simultaneously.
Note that the deadlock has happened only if NET_MPSAFE is enabled.
---
Add missing ifdef NET_MPSAFE
---
Take softnet_lock on pr_input properly if NET_MPSAFE
Currently softnet_lock is taken unnecessarily in some cases, e.g.,
icmp_input and encap4_input from ip_input, or not taken even if needed,
e.g., udp_input and tcp_input from ipsec4_common_input_cb. Fix them.
NFC if NET_MPSAFE is disabled (default).
---
- sanitize key debugging so that we don't print extra newlines or unassociated
debugging messages.
- remove unused functions and make internal ones static
- print information in one line per message
---
humanize printing of ip addresses
---
cast reduction, NFC.
---
Fix typo in comment
---
Pull out ipsec_fill_saidx_bymbuf (NFC)
---
Don't abuse key_checkrequest just for looking up sav
It does more than expected for example key_acquire.
---
Fix SP is broken on transport mode
isr->saidx was modified accidentally in ipsec_nextisr.
Reported by christos@
Helped investigations by christos@ and knakahara@
---
Constify isr at many places (NFC)
---
Include socketvar.h for softnet_lock
---
Fix buffer length for ipsec_logsastr
|
| 1.184.2.5 | 18-Jan-2019 |
pgoyette | Synch with HEAD
|
| 1.184.2.4 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.184.2.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.184.2.2 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.184.2.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.186.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.192.4.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.195.4.1 | 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.198.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.25 | 07-Oct-2024 |
jakllsch | Allow CACHE_LINE_SIZE 256 with uint64_t fatp_word_t
|
| 1.24 | 04-Nov-2022 |
ozaki-r | branches: 1.24.8;
inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.23 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.22 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.21 | 13-Aug-2021 |
andvar | fix typos in words "pointer" and s/fram /frame/
|
| 1.20 | 01-Oct-2019 |
chs | in many device attach paths, allocate memory with KM_SLEEP instead of KM_NOSLEEP
and remove code to handle failures that can no longer happen.
|
| 1.19 | 03-May-2018 |
maxv | branches: 1.19.2;
Remove now unused tcpip.h includes. Some were already unused before.
|
| 1.18 | 01-Jun-2017 |
chs | branches: 1.18.8;
remove checks for failure after memory allocation calls that cannot fail:
kmem_alloc() with KM_SLEEP
kmem_zalloc() with KM_SLEEP
percpu_alloc()
pserialize_create()
psref_class_create()
all of these paths include an assertion that the allocation has not failed,
so callers should not assert that again.
|
| 1.17 | 13-Dec-2016 |
ozaki-r | Remove unnecessary inclusions of nd6.h
|
| 1.16 | 28-Jul-2016 |
martin | PR kern/51371: avoid shifting negative values
|
| 1.15 | 26-Apr-2016 |
ozaki-r | branches: 1.15.2;
Sweep unnecessary route.h inclusions
|
| 1.14 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.13 | 31-Mar-2015 |
ozaki-r | Remove unnecessary opt_ipsec.h inclusions
|
| 1.12 | 10-Nov-2014 |
maxv | branches: 1.12.2;
Do not uselessly include <sys/malloc.h>.
|
| 1.11 | 05-Sep-2014 |
matt | Don't use C++ keywords (class, template) as variables
|
| 1.10 | 15-Sep-2013 |
martin | branches: 1.10.4;
ifdef a variable like its use
|
| 1.9 | 13-Apr-2012 |
yamt | branches: 1.9.2; 1.9.4;
add a big comment
(copy and paste from cvs log rev.1.1)
|
| 1.8 | 17-Jul-2011 |
joerg | branches: 1.8.2; 1.8.6;
Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.7 | 06-Jun-2011 |
dyoung | Don't allocate resources for vtw until/unless it is enabled. This will
further help those machines where memory is in short supply.
TBD: release resources after vtw is disabled and all entries have
expired.
|
| 1.6 | 03-Jun-2011 |
dyoung | branches: 1.6.2;
Don't sleep until memory becomes available.
Use kmem_zalloc() instead of kmem_alloc() + bzero().
During initialization, try to get all of the memory we need for the
vestigial time-wait structures before we set any of the structures up,
and if any single allocation fails, release all of the memory.
This should help low-memory hosts. A much better fix postpones
allocating any memory until vtw is enabled through the sysctl.
|
| 1.5 | 03-Jun-2011 |
dyoung | Defer scheduling vtw_tick() and setting the vtw hooks until
vtw_control() is called. In this way, vtw_tick() will be re-scheduled
repeatedly while vtw is in use.
Pay tcp_vtw_was_enabled no attention in vtw_earlyinit(), since it's
always going to be 0 during initialization.
|
| 1.4 | 17-May-2011 |
dholland | branches: 1.4.2; 1.4.4;
typo in comment
|
| 1.3 | 11-May-2011 |
drochner | use getmicrouptime(9) rather than microtime(9) for TIME_WAIT duration
calculation, because this doesn't get confused by system time changes,
and uses less CPU cycles
reviewed by dyoung
|
| 1.2 | 06-May-2011 |
drochner | remove an empty function
|
| 1.1 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.4.4.1 | 23-Jun-2011 |
cherry | Catchup with rmind-uvmplock merge.
|
| 1.4.2.3 | 12-Jun-2011 |
rmind | sync with head
|
| 1.4.2.2 | 31-May-2011 |
rmind | sync with head
|
| 1.4.2.1 | 17-May-2011 |
rmind | file tcp_vtw.c was added on branch rmind-uvmplock on 2011-05-31 03:05:08 +0000
|
| 1.6.2.2 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.6.2.1 | 03-Jun-2011 |
jruoho | file tcp_vtw.c was added on branch jruoho-x86intr on 2011-06-06 09:09:57 +0000
|
| 1.8.6.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.8.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.8.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.9.4.3 | 18-May-2014 |
rmind | sync with head
|
| 1.9.4.2 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.9.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.9.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.9.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.10.4.1 | 17-Jan-2015 |
martin | Pull up following revision(s) (requested by maxv in ticket #427):
sys/compat/svr4/svr4_schedctl.c: revision 1.8
sys/netinet/tcp_timer.c: revision 1.88
sys/miscfs/genfs/layer_vfsops.c: revision 1.45
sys/compat/svr4/svr4_ioctl.c: revision 1.37
sys/ufs/chfs/chfs_vfsops.c: revision 1.14
sys/miscfs/fdesc/fdesc_vfsops.c: revision 1.91
sys/compat/linux/arch/i386/linux_ptrace.c: revision 1.30
sys/compat/common/kern_time_50.c: revision 1.28
sys/netinet6/ip6_forward.c: revision 1.74
sys/miscfs/umapfs/umap_vnops.c: revision 1.57
sys/compat/svr4/svr4_fcntl.c: revision 1.74
distrib/sets/lists/comp/mi: revision 1.1931
sys/netinet6/udp6_output.c: revision 1.46
sys/fs/puffs/puffs_compat.c: revision 1.3
sys/fs/udf/udf_rename.c: revision 1.11
sys/compat/svr4/svr4_filio.c: revision 1.24
sys/fs/udf/udf_rename.c: revision 1.12
sys/netinet/tcp_usrreq.c: revision 1.202
sys/miscfs/umapfs/umap_subr.c: revision 1.29
sys/compat/linux/common/linux_fadvise64.c: revision 1.3
sys/netinet/if_atm.c: revision 1.34
sys/miscfs/procfs/procfs_subr.c: revision 1.106
sys/miscfs/genfs/layer_subr.c: revision 1.37
sys/netinet/tcp_sack.c: revision 1.30
sys/compat/freebsd/freebsd_misc.c: revision 1.33
sys/compat/freebsd/freebsd_file.c: revision 1.33
sys/ufs/chfs/chfs_vnode.c: revision 1.12
sys/compat/svr4/svr4_ttold.c: revision 1.34
sys/compat/linux/common/linux_file.c: revision 1.114
sys/compat/linux/arch/mips/linux_machdep.c: revision 1.43
sys/compat/linux/common/linux_signal.c: revision 1.76
sys/compat/common/compat_util.c: revision 1.46
sys/compat/linux/arch/arm/linux_ptrace.c: revision 1.18
sys/compat/svr4/svr4_sockio.c: revision 1.36
sys/compat/linux/arch/arm/linux_machdep.c: revision 1.32
sys/compat/svr4/svr4_signal.c: revision 1.66
sys/kern/kern_exec.c: revision 1.410
sys/fs/puffs/puffs_vfsops.c: revision 1.115
sys/compat/svr4/svr4_exec_elf64.c: revision 1.15
sys/compat/linux/arch/i386/linux_machdep.c: revision 1.159
sys/compat/linux/arch/alpha/linux_machdep.c: revision 1.50
sys/compat/linux32/common/linux32_misc.c: revision 1.24
sys/netinet/in_pcb.c: revision 1.153
sys/sys/malloc.h: revision 1.116
sys/compat/common/if_43.c: revision 1.9
share/man/man9/Makefile: revision 1.380
sys/netinet/tcp_vtw.c: revision 1.12
sys/miscfs/umapfs/umap_vfsops.c: revision 1.95
sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.186
sys/compat/common/uipc_syscalls_43.c: revision 1.46
sys/ufs/ext2fs/ext2fs_vnops.c: revision 1.115
sys/fs/puffs/puffs_msgif.c: revision 1.97
sys/compat/svr4/svr4_ipc.c: revision 1.27
sys/compat/linux/common/linux_exec.c: revision 1.117
sys/ufs/ext2fs/ext2fs_readwrite.c: revision 1.66
sys/netinet/tcp_output.c: revision 1.179
sys/compat/svr4/svr4_termios.c: revision 1.28
sys/fs/udf/udf_strat_bootstrap.c: revision 1.4
sys/fs/puffs/puffs_subr.c: revision 1.67
sys/fs/puffs/puffs_node.c: revision 1.36
sys/miscfs/overlay/overlay_vnops.c: revision 1.21
sys/fs/cd9660/cd9660_node.c: revision 1.34
sys/netinet/raw_ip.c: revision 1.146
sys/sys/mallocvar.h: revision 1.13
sys/miscfs/overlay/overlay_vfsops.c: revision 1.63
share/man/man9/malloc.9: revision 1.50
sys/netinet6/dest6.c: revision 1.18
sys/compat/linux/common/linux_uselib.c: revision 1.33
sys/compat/linux/common/linux_socket.c: revision 1.120
share/man/man9/malloc.9: revision 1.51
sys/netinet/tcp_subr.c: revision 1.257
sys/compat/linux/common/linux_socketcall.c: revision 1.45
sys/compat/linux/common/linux_fadvise64_64.c: revision 1.3
sys/compat/freebsd/freebsd_ipc.c: revision 1.17
sys/compat/linux/common/linux_misc_notalpha.c: revision 1.109
sys/compat/linux/arch/alpha/linux_pipe.c: revision 1.17
sys/netinet6/in6_pcb.c: revision 1.132
sys/netinet6/in6_ifattach.c: revision 1.94
sys/compat/svr4/svr4_exec_elf32.c: revision 1.15
sys/miscfs/nullfs/null_vfsops.c: revision 1.90
sys/fs/cd9660/cd9660_util.c: revision 1.12
sys/compat/linux/arch/powerpc/linux_machdep.c: revision 1.48
sys/compat/freebsd/freebsd_exec_elf32.c: revision 1.20
sys/miscfs/procfs/procfs_vfsops.c: revision 1.94
sys/compat/linux/arch/powerpc/linux_ptrace.c: revision 1.28
sys/compat/linux/common/linux_sched.c: revision 1.67
sys/compat/linux/common/linux_exec_aout.c: revision 1.67
sys/compat/linux/common/linux_pipe.c: revision 1.67
sys/compat/linux/common/linux_llseek.c: revision 1.34
sys/compat/linux/arch/mips/linux_ptrace.c: revision 1.10
Do not uselessly include <sys/malloc.h>.
Cleanup:
- remove struct kmembuckets (dead)
- correctly deadify MALLOC_XX
- remove MALLOC_DEFINE_LIMIT and MALLOC_JUSTDEFINE_LIMIT (dead)
- remove malloc_roundup(), malloc_type_setlimit(), MALLOC_DEFINE_LIMIT()
and MALLOC_JUSTDEFINE_LIMIT() from man 9 malloc
New sentence, new line. Bump date for previous.
Obsolete malloc_roundup(9), malloc_type_setlimit(9) and MALLOC_DEFINE_LIMIT(9)
man pages.
|
| 1.12.2.6 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.12.2.5 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.12.2.4 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.12.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.12.2.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.12.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.15.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.15.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
| 1.18.8.1 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.19.2.1 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.24.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.11 | 07-Oct-2024 |
jakllsch | Allow CACHE_LINE_SIZE 256 with uint64_t fatp_word_t
|
| 1.10 | 11-Dec-2022 |
mlelstv | branches: 1.10.8;
Need larger fat pointers for 128bit cache lines.
|
| 1.9 | 19-Apr-2018 |
christos | s/static inline/static __inline/g for consistency.
|
| 1.8 | 13-Dec-2016 |
ozaki-r | branches: 1.8.14;
Remove unnecessary inclusions of nd6.h
|
| 1.7 | 26-Apr-2016 |
ozaki-r | branches: 1.7.2;
Sweep unnecessary route.h inclusions
|
| 1.6 | 23-Nov-2012 |
joerg | branches: 1.6.2; 1.6.14;
Add RCS keyword. Avoid overflow in constant.
|
| 1.5 | 07-Jun-2011 |
joerg | branches: 1.5.2; 1.5.12;
Be a bit cleaner and reduce the amount of namespace pollution
|
| 1.4 | 06-Jun-2011 |
dyoung | Don't allocate resources for vtw until/unless it is enabled. This will
further help those machines where memory is in short supply.
TBD: release resources after vtw is disabled and all entries have
expired.
|
| 1.3 | 17-May-2011 |
dholland | branches: 1.3.2; 1.3.4; 1.3.6;
typo in comment
|
| 1.2 | 03-May-2011 |
dyoung | Remove #ifdef INET6 throughout.
|
| 1.1 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.3.6.2 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.3.6.1 | 17-May-2011 |
jruoho | file tcp_vtw.h was added on branch jruoho-x86intr on 2011-06-06 09:09:57 +0000
|
| 1.3.4.1 | 23-Jun-2011 |
cherry | Catchup with rmind-uvmplock merge.
|
| 1.3.2.3 | 12-Jun-2011 |
rmind | sync with head
|
| 1.3.2.2 | 31-May-2011 |
rmind | sync with head
|
| 1.3.2.1 | 17-May-2011 |
rmind | file tcp_vtw.h was added on branch rmind-uvmplock on 2011-05-31 03:05:08 +0000
|
| 1.5.12.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.5.12.1 | 25-Feb-2013 |
tls | resync with head
|
| 1.5.2.1 | 16-Jan-2013 |
yamt | sync with (a bit old) head
|
| 1.6.14.2 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.6.14.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.6.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.7.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.8.14.1 | 22-Apr-2018 |
pgoyette | Sync with HEAD
|
| 1.10.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.12 | 29-Apr-2018 |
maxv | Move struct tcpiphdr from tcpip.h to tcp_var.h, to match UDP (udpiphdr in
udp_var.h).
tcpip.h is now empty, and can be removed.
|
| 1.11 | 25-Dec-2007 |
perry | branches: 1.11.96;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.10 | 10-Dec-2005 |
elad | branches: 1.10.46; 1.10.52; 1.10.56; 1.10.60;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.9 | 07-Aug-2003 |
agc | branches: 1.9.16;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.8 | 20-Nov-1999 |
thorpej | branches: 1.8.28;
Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.7 | 10-Feb-1998 |
perry | branches: 1.7.14; 1.7.20;
add/cleanup multiple inclusion protection.
|
| 1.6 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.7.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.7.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.8.28.4 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.8.28.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.8.28.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.8.28.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.9.16.2 | 21-Jan-2008 |
yamt | sync with head
|
| 1.9.16.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.10.60.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.10.56.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.10.52.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.10.46.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.11.96.1 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.19 | 03-Feb-2021 |
roy | CTASSERT -> __CTASSERT to unbreak userland build.
While here move __packed in tcp_debug.h back to where it was and
note removal warrants more investigation.
|
| 1.18 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.17 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.16 | 22-Jun-2012 |
christos | branches: 1.16.52;
PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.15 | 07-Jan-2012 |
christos | u_intN -> uintN
make standalone
|
| 1.14 | 24-Sep-2011 |
christos | branches: 1.14.2; 1.14.6;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.13 | 25-Dec-2007 |
perry | Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.12 | 10-Dec-2005 |
elad | branches: 1.12.46; 1.12.52; 1.12.56; 1.12.60;
Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.11 | 12-Feb-2005 |
manu | branches: 1.11.6;
Add support for IPsec Network Address Translator traversal (NAT-T), as
described by RFC 3947 and 3948.
|
| 1.10 | 07-Aug-2003 |
agc | branches: 1.10.8; 1.10.10;
Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.9 | 20-Nov-1999 |
thorpej | branches: 1.9.28;
Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.8 | 10-Feb-1998 |
perry | branches: 1.8.12; 1.8.14; 1.8.20;
add/cleanup multiple inclusion protection.
|
| 1.7 | 25-Oct-1996 |
thorpej | Make length and offset fields unsigned. From Kevin M. Lahey <kml@nas.nasa.gov>
|
| 1.6 | 13-Apr-1995 |
cgd | branches: 1.6.6;
be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.6.6.1 | 10-Nov-1996 |
thorpej | Update from trunk:
- Make ip_len and ip_off unsigned.
- Make sure we don't accept or transmit packets larger than the
maximim IP packet size.
This fixes the so-called `death ping' bug.
Sum of work from Bill Fenner <fenner@parc.xerox.com>,
Kevin Lahey <kml@nas.nasa.gov>, and myself.
Thanks to Curt Sampson, Jukka Marin, and Kevin Lahey for testing
this under NetBSD 1.2
|
| 1.8.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.8.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.8.12.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.8.12.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.8.12.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.9.28.5 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.9.28.4 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.9.28.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.9.28.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.9.28.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.10.10.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.10.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.11.6.2 | 21-Jan-2008 |
yamt | sync with head
|
| 1.11.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.12.60.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.12.56.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.12.52.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.12.46.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.14.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.14.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.14.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.16.52.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.6 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.5 | 17-Feb-2021 |
christos | - pass the alignment instead of the mask (as Roy asked and to match the
other macro)
- use alignof to determine that alignment and CTASSERT what we expect
- remove unused macros
|
| 1.4 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.3 | 28-Apr-2008 |
martin | branches: 1.3.4; 1.3.102;
Remove clause 3 and 4 from TNF licenses
|
| 1.2 | 23-Apr-2008 |
thorpej | branches: 1.2.2;
Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.1 | 12-Apr-2008 |
thorpej | branches: 1.1.2;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.1.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.2.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.3.102.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.3.4.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.3.4.1 | 28-Apr-2008 |
mjf | file udp_private.h was added on branch mjf-devfs2 on 2008-06-02 13:24:25 +0000
|
| 1.266 | 08-Oct-2024 |
riastradh | udp(4): Clarify udp4/6_espinudp and inp_overudp_cb return.
Cleanup to detect problems like this earlier:
PR kern/58688: userland panic of kernel via wg(4)
|
| 1.265 | 05-Jul-2024 |
rin | sys: Drop redundant NULL check before m_freem(9)
m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c
Compile-tested on amd64/ALL.
Suggested by knakahara@
|
| 1.264 | 04-Nov-2022 |
ozaki-r | branches: 1.264.8;
inpcb: rename functions to inpcb_*
Inspired by rmind-smpnet patches.
|
| 1.263 | 28-Oct-2022 |
ozaki-r | inpcb: separate inpcb again to reduce the size of PCB for IPv4
The data size of PCB for IPv4 increased because of the merge of
struct in6pcb. The change decreases the size to the original size by
separating struct inpcb (again). struct in4pcb and in6pcb that embed
struct inpcb are introduced.
Even after the separation, users don't need to realize the separation
and only have to use some macros to access dedicated data. For example,
inp->inp_laddr is now accessed through in4p_laddr(inp).
|
| 1.262 | 28-Oct-2022 |
ozaki-r | inpcb: integrate data structures of PCB into one
Data structures of network protocol control blocks (PCBs), i.e.,
struct inpcb, in6pcb and inpcb_hdr, are not organized well. Users of
the data structures have to handle them separately and thus the code
is cluttered and duplicated.
The commit integrates the data structures into one, struct inpcb. As a
result, users of PCBs only have to handle just one data structure, so
the code becomes simple.
One drawback is that the data size of PCB for IPv4 increases by 40 bytes
(from 248 bytes to 288 bytes).
|
| 1.261 | 19-Feb-2021 |
christos | - Make ALIGNED_POINTER use __alignof(t) instead of sizeof(t). This is more
correct because it works with non-primitive types and provides the ABI
alignment for the type the compiler will use.
- Remove all the *_HDR_ALIGNMENT macros and asserts
- Replace POINTER_ALIGNED_P with ACCESSIBLE_POINTER which is identical to
ALIGNED_POINTER, but returns that the pointer is always aligned if the
CPU supports unaligned accesses.
[ as proposed in tech-kern ]
|
| 1.260 | 14-Feb-2021 |
christos | - centralize header align and pullup into a single inline function
- use a single macro to align pointers and expose the alignment, instead
of hard-coding 3 in 1/2 the macros.
- fix an issue in the ipv6 lt2p where it was aligning for ipv4 and pulling
for ipv6.
|
| 1.259 | 20-Aug-2020 |
riastradh | branches: 1.259.2;
[ozaki-r] Changes to the kernel core for wireguard
|
| 1.258 | 27-Dec-2018 |
maxv | Remove unused arguments.
|
| 1.257 | 22-Nov-2018 |
knakahara | Support IPv6 NAT-T. Implemented by hsuenaga@IIJ and ohishi@IIJ.
Add ATF later.
|
| 1.256 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.255 | 15-Jul-2018 |
maxv | Retire ipkdb entirely. The option was removed from the config files
yesterday.
ok kamil christos
|
| 1.254 | 31-May-2018 |
maxv | branches: 1.254.2;
Remove the non-IKE part of the computation, too.
|
| 1.253 | 31-May-2018 |
maxv | Remove support for non-IKE markers in the kernel. Discussed on tech-net@,
and now in PR/53334. Basically non-IKE markers come from a deprecated
draft, and our kernel code for them has never worked.
Setsockopt will now reject UDP_ENCAP_ESPINUDP_NON_IKE.
Perhaps we should also add a check in key_handle_natt_info(), to make
sure we also reject UDP_ENCAP_ESPINUDP_NON_IKE in the SADB.
|
| 1.252 | 18-May-2018 |
maxv | IP6_EXTHDR_GET -> M_REGION_GET, no functional change.
|
| 1.251 | 13-May-2018 |
maxv | Clarify ESP-in-UDP.
|
| 1.250 | 01-May-2018 |
maxv | Remove unused argument from udp4_espinudp, and remove unused includes.
|
| 1.249 | 28-Apr-2018 |
maxv | Remove unused ipsec_var.h includes.
|
| 1.248 | 13-Apr-2018 |
maxv | Improve the check, we want to have len >= udphdr all the time, and not
just when the packet size doesn't match the mbuf size.
Normally that's not a huge problem, since IP6_EXTHDR_GET gets called
earlier, so we can't have
(ip_len == iphlen + len) && (len < sizeof(struct udphdr))
|
| 1.247 | 12-Apr-2018 |
maxv | Remove misleading comment; we're just checking the SP, not verifying the
AH/ESP payload. While here style a bit.
|
| 1.246 | 19-Mar-2018 |
roy | socket: report receive buffer overflows
Add soroverflow() which increments the overflow counter, sets so_error
to ENOBUFS and wakes the receive socket up.
Replace all code that manually increments this counter with soroverflow().
Add soroverflow() to raw_input().
This allows userland to detect route(4) overflows so it can re-sync
with the current state.
|
| 1.245 | 28-Feb-2018 |
maxv | branches: 1.245.2;
Remove unused ipsec_private.h includes.
|
| 1.244 | 28-Feb-2018 |
maxv | Remove duplicate IPSEC_STATINC(IPSEC_STAT_IN_POLVIO), ipsec_in_reject
already increases it. IPSEC6_STATINC is now unused, so remove it too.
|
| 1.243 | 26-Feb-2018 |
maxv | Dedup: merge ipsec4_in_reject and ipsec6_in_reject into ipsec_in_reject.
While here fix misleading comment.
ok ozaki-r@
|
| 1.242 | 14-Feb-2018 |
maxv | Revert my two last changes in this file. They are apparently causing
problems with racoon, I'll investigate this later.
|
| 1.241 | 12-Feb-2018 |
maxv | Don't rebase the pointers. 'm' is only allowed to become NULL (which
means 'processed').
|
| 1.240 | 10-Feb-2018 |
maxv | If the socket wants a ESP-over-UDP packet, and the packet is incorrect,
stop processing it instead of giving it to udp4_sendup. It just doesn't
make any sense not to drop it.
I was already telling myself this the other day when I visited this place,
but I just saw PR/36782 (11 years old) that suggests the exact same thing,
so fix it.
Now, udp4_espinudp always frees the mbuf, and is made void. The packet is
not processed any further afterwards.
|
| 1.239 | 08-Feb-2018 |
maxv | More style, no functional change.
|
| 1.238 | 08-Feb-2018 |
maxv | Style, and remove printfs.
|
| 1.237 | 08-Feb-2018 |
maxv | Fix three pretty bad mistakes in NAT-T:
* If we got a keepalive packet, we need to call m_freem, not m_free.
Here the next mbufs in the chain are not freed. Seems easy to remotely
DoS the system by sending fragmented keepalives in a loop.
* If !ipsec_used, free the mbuf.
* In udp_input, we need to update 'uh', because udp4_realinput may have
modified the chain. Perhaps we also need to re-enforce alignment, so
add an XXX.
|
| 1.236 | 11-Dec-2017 |
ryo | As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.235 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.234 | 06-Jul-2017 |
christos | Merge the two copies SO_TIMESTAMP/SO_OTIMESTAMP processing to a single
function, and add a SOOPT_TIMESTAMP define reducing compat pollution from
5 places to 1.
|
| 1.233 | 20-Apr-2017 |
ozaki-r | branches: 1.233.4;
Remove unnecessary NULL checks for inp_socket and in6p_socket
They cannot be NULL except for programming errors.
|
| 1.232 | 20-Apr-2017 |
ozaki-r | Simplify logic of udp4_sendup and udp6_sendup
They are always passed a socket with the same protocol faimiliy
as its own: AF_INET for udp4_sendup and AF_INET6 for udp6_sendup.
|
| 1.231 | 03-Mar-2017 |
ozaki-r | Pass inpcb/in6pcb instead of socket to ip_output/ip6_output
- Passing a socket to Layer 3 is layer violation and even unnecessary
- The change makes codes of callers and IPsec a bit simple
|
| 1.230 | 24-Jan-2017 |
ozaki-r | Tweak softnet_lock and NET_MPSAFE
- Don't hold softnet_lock in some functions if NET_MPSAFE
- Add softnet_lock to sysctl_net_inet_icmp_redirtimeout
- Add softnet_lock to expire_upcalls of ip_mroute.c
- Restore softnet_lock for in{,6}_pcbpurgeif{,0} if NET_MPSAFE
- Mark some softnet_lock for future work
|
| 1.229 | 18-Nov-2016 |
knakahara | branches: 1.229.2;
fix: "ifconfig destory" can stalls when "ifconfig" is done parallel.
This problem occurs only if NET_MPSAFE on.
ifconfig destroy side:
kernel entry point is ifioctl => if_clone_destroy.
pr_purgeif() acquires softnet_lock, and then ifa_remove() calls
pserialize_perform() holding softnet_lock.
ifconfig side:
kernel entry point is socreate.
pr_attach()(udp_attach_wrapper()) calls sosetlock(). In this call path,
sosetlock() try to acquire softnet_lock.
These can cause dead lock.
|
| 1.228 | 15-Nov-2016 |
mlelstv | Enforce alignment requirements that are violated in some cases.
For machines that don't need strict alignment (i386,amd64,vax,m68k) this
is a no-op.
Fixes PR kern/50766 but should be improved.
|
| 1.227 | 19-Oct-2016 |
ozaki-r | Remove unnecessary #ifdef IPSEC
The entire function is already in #ifdef IPSEC.
No functional change.
|
| 1.226 | 10-Jun-2016 |
ozaki-r | branches: 1.226.2;
Avoid storing a pointer of an interface in a mbuf
Having a pointer of an interface in a mbuf isn't safe if we remove big
kernel locks; an interface object (ifnet) can be destroyed anytime in any
packet processing and accessing such object via a pointer is racy. Instead
we have to get an object from the interface collection (ifindex2ifnet) via
an interface index (if_index) that is stored to a mbuf instead of an
pointer.
The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9)
for sleep-able critical sections and m_{get,put}_rcvif that use
pserialize(9) for other critical sections. The change also adds another
API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition
moratorium, i.e., it is intended to be used for places where are not
planned to be MP-ified soon.
The change adds some overhead due to psref to performance sensitive paths,
however the overhead is not serious, 2% down at worst.
Proposed on tech-kern and tech-net.
|
| 1.225 | 26-Apr-2016 |
ozaki-r | Sweep unnecessary route.h inclusions
|
| 1.224 | 15-Feb-2016 |
rtr | Reduce code duplication.
Split creation of IPv4-Mapped IPv6 addresses into its own function
and use it.
No functional change intended. As posted to tech-net@
|
| 1.223 | 20-Jan-2016 |
riastradh | Give proper prototype to udp_output.
|
| 1.222 | 24-Aug-2015 |
pooka | sprinkle _KERNEL_OPT
|
| 1.221 | 02-May-2015 |
rtr | make connect syscall use sockaddr_big and modify pr_{send,connect}
nam parameter type from buf * to sockaddr *.
final commit for parameter type changes to protocol user requests
* bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
|
| 1.220 | 26-Apr-2015 |
rtr | remove pr_generic from struct pr_usrreqs and all implementations of
pr_generic in protocols.
bump to 7.99.13
approved by rmind@
|
| 1.219 | 24-Apr-2015 |
rtr | make accept, getsockname and getpeername syscalls use sockaddr_big and modify
pr_{accept,sockname,peername} nam parameter type from mbuf * to sockaddr *.
* retained use of mbuftypes[MT_SONAME] for now.
* bump to netbsd version 7.99.12 for parameter type change.
patch posted to tech-net@ 2015/04/19
|
| 1.218 | 03-Apr-2015 |
rtr | * change pr_bind to accept struct sockaddr * instead of struct mbuf *
* update protocol bind implementations to use/expect sockaddr *
instead of mbuf *
* introduce sockaddr_big struct for storage of addr data passed via
sys_bind; sockaddr_big is of sufficient size and alignment to
accommodate all addr data sizes received.
* modify sys_bind to allocate sockaddr_big instead of using an mbuf.
* bump kernel version to 7.99.9 for change to pr_bind() parameter type.
Patch posted to tech-net@
http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html
The choice to use a new structure sockaddr_big has been retained since
changing sockaddr_storage size would lead to unnecessary ABI change. The
use of the new structure does not preclude future work that increases
the size of sockaddr_storage and at that time sockaddr_big may be
trivially replaced.
Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
|
| 1.217 | 09-Aug-2014 |
rtr | branches: 1.217.2; 1.217.4; 1.217.6; 1.217.10;
split PRU_CONNECT2 & PRU_PURGEIF function out of pr_generic() usrreq
switches and put into separate functions
- always KASSERT(solocked(so)) even if not implemented
(for PRU_CONNECT2 only)
- replace calls to pr_generic() with req = PRU_CONNECT2 with calls to
pr_connect2()
- replace calls to pr_generic() with req = PRU_PURGEIF with calls to
pr_purgeif()
put common code from unp_connect2() (used by unp_connect() into
unp_connect1() and call out to it when needed
patch only briefly reviewed by rmind@
|
| 1.216 | 08-Aug-2014 |
rtr | split PRU_RCVD function out of pr_generic() usrreq switches and put into
separate functions
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_RCVD with calls to
pr_rcvd()
|
| 1.215 | 05-Aug-2014 |
rtr | split PRU_SEND function out of pr_generic() usrreq switches and put into
separate functions
xxx_send(struct socket *, struct mbuf *, struct mbuf *,
struct mbuf *, struct lwp *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_SEND with calls to
pr_send()
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_send() PRUs
- l2cap_send() -> l2cap_send_pcb()
- sco_send() -> sco_send_pcb()
- rfcomm_send() -> rfcomm_send_pcb()
patch reviewed by rmind
|
| 1.214 | 05-Aug-2014 |
rtr | revert the removal of struct lwp * parameter from bind, listen and connect
user requests.
this should resolve the issue relating to nfs client hangs presented
recently by wiz on current-users@
|
| 1.213 | 02-Aug-2014 |
rtr | restore splsoftnet() in various usrreqs that were removed during the PRU
splits. we will properly review removal after the PRU split work is
complete.
|
| 1.212 | 31-Jul-2014 |
rtr | split PRU_DISCONNECT, PRU_SHUTDOWN and PRU_ABORT function out of
pr_generic() usrreq switches and put into separate functions
xxx_disconnect(struct socket *)
xxx_shutdown(struct socket *)
xxx_abort(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req =
PRU_{DISCONNECT,SHUTDOWN,ABORT}
with calls to pr_{disconnect,shutdown,abort}() respectively
rename existing internal functions used to implement above functionality
to permit use of the names for xxx_{disconnect,shutdown,abort}().
- {l2cap,sco,rfcomm}_disconnect() ->
{l2cap,sco,rfcomm}_disconnect_pcb()
- {unp,rip,tcp}_disconnect() -> {unp,rip,tcp}_disconnect1()
- unp_shutdown() -> unp_shutdown1()
patch reviewed by rmind
|
| 1.211 | 30-Jul-2014 |
rtr | split PRU_CONNECT function out of pr_generic() usrreq switches and put
into seaparate functions
xxx_listen(struct socket *, struct mbuf *)
- always KASSERT(solocked(so)) and KASSERT(nam != NULL)
- replace calls to pr_generic() with req = PRU_CONNECT with
pr_connect()
- rename existin {l2cap,sco,rfcomm}_connect() to
{l2cap,sco,rfcomm}_connect_pcb() respectively to permit
naming consistency with other protocols functions.
- drop struct lwp * parameter from unp_connect() and at_pcbconnect()
and use curlwp instead where appropriate.
patch reviewed by rmind
|
| 1.210 | 24-Jul-2014 |
rtr | split PRU_BIND and PRU_LISTEN function out of pr_generic() usrreq
switches and put into separate functions
xxx_bind(struct socket *, struct mbuf *)
xxx_listen(struct socket *)
- always KASSERT(solocked(so)) even if not implemented
- replace calls to pr_generic() with req = PRU_BIND with call to
pr_bind()
- replace calls to pr_generic() with req = PRU_LISTEN with call to
pr_listen()
- drop struct lwp * parameter from at_pcbsetaddr(), in_pcbbind() and
unp_bind() and always use curlwp.
rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_{bind,listen}() PRUs
- l2cap_{bind,listen}() -> l2cap_{bind,listen}_pcb()
- sco_{bind,listen}() -> sco_{bind,listen}_pcb()
- rfcomm_{bind,listen}() -> rfcomm_{bind,listen}_pcb()
patch reviewed by rmind
welcome to netbsd 6.99.48
|
| 1.209 | 23-Jul-2014 |
rtr | split PRU_SENDOOB and PRU_RCVOOB function out of pr_generic() usrreq
switches and put into separate functions
xxx_sendoob(struct socket *, struct mbuf *, struct mbuf *)
xxx_recvoob(struct socket *, struct mbuf *, int)
- always KASSERT(solocked(so)) even if request is not implemented
- replace calls to pr_generic() with req = PRU_{SEND,RCV}OOB with
calls to pr_{send,recv}oob() respectively.
there is still some tweaking of m_freem(m) and m_freem(control) to come
for consistency. not performed with this commit for clarity.
reviewed by rmind
|
| 1.208 | 09-Jul-2014 |
rtr | * split PRU_ACCEPT function out of pr_generic() usrreq switches and put
into a separate function xxx_accept(struct socket *, struct mbuf *)
note: future cleanup will take place to remove struct mbuf parameter
type and replace it with a more appropriate type.
patch reviewed by rmind
|
| 1.207 | 09-Jul-2014 |
rtr | * split PRU_PEERADDR and PRU_SOCKADDR function out of pr_generic()
usrreq switches and put into separate functions
xxx_{peer,sock}addr(struct socket *, struct mbuf *).
- KASSERT(solocked(so)) always in new functions even if request
is not implemented
- KASSERT(pcb != NULL) and KASSERT(nam) if the request is
implemented and not for tcp.
* for tcp roll #ifdef KPROF and #ifdef DEBUG code from tcp_usrreq() into
easier to cut & paste functions tcp_debug_capture() and
tcp_debug_trace()
- functions provided by rmind
- remaining use of PRU_{PEER,SOCK}ADDR #define to be removed in a
future commit.
* rename netbt functions to permit consistency of pru function names
(as has been done with other requests already split out).
- l2cap_{peer,sock}addr() -> l2cap_{peer,sock}_addr_pcb()
- rfcomm_{peer,sock}addr() -> rfcomm_{peer,sock}_addr_pcb()
- sco_{peer,sock}addr() -> sco_{peer,sock}_addr_pcb()
* split/refactor do_sys_getsockname(lwp, fd, which, nam) into
two functions do_sys_get{peer,sock}name(fd, nam).
- move PRU_PEERADDR handling into do_sys_getpeername() from
do_sys_getsockname()
- have svr4_stream directly call do_sys_get{sock,peer}name()
respectively instead of providing `which' & fix a DPRINTF string
that incorrectly wrote "getpeername" when it meant "getsockname"
- fix sys_getpeername() and sys_getsockname() to call
do_sys_get{sock,peer}name() without `which' and `lwp' & adjust
comments
- bump kernel version for removal of lwp & which parameters from
do_sys_getsockname()
note: future cleanup to remove struct mbuf * abuse in
xxx_{peer,sock}name()
still to come, not done in this commit since it is easier to do post
split.
patch reviewed by rmind
welcome to 6.99.47
|
| 1.206 | 07-Jul-2014 |
rtr | * sprinkle KASSERT(solocked(so)); in all pr_stat() functions.
* fix remaining inconsistent struct socket parameter names.
|
| 1.205 | 07-Jul-2014 |
rtr | backout change that made pr_stat return EOPNOTSUPP for protocols that
were not filling in struct stat.
decision made after further discussion with rmind and investigation of
how other operating systems behave. soo_stat() is doing just enough to
be able to call what gets returned valid and thus justifys a return of
success.
additional review will be done to determine of the pr_stat functions
that were already returning EOPNOTSUPP can be considered successful with
what soo_stat() is doing.
|
| 1.204 | 07-Jul-2014 |
rtr | * have pr_stat return EOPNOTSUPP consistently for all protocols that do
not fill in struct stat instead of returning success.
* in pr_stat remove all checks for non-NULL so->so_pcb except where the
pcb is actually used (i.e. cases where we don't return EOPNOTSUPP).
proposed on tech-net@
|
| 1.203 | 06-Jul-2014 |
rtr | * split PRU_SENSE functionality out of xxx_usrreq() switches and place into
separate xxx_stat(struct socket *, struct stat *) functions.
* replace calls using pr_generic with req == PRU_SENSE with pr_stat().
further change will follow that cleans up the pattern used to extract the
pcb and test for its presence.
reviewed by rmind
|
| 1.202 | 01-Jul-2014 |
rtr | fix parameter types in pr_ioctl, called xx_control() functions and remove
abuse of pointer to struct mbuf type.
param2 changed to u_long type and uses parameter name 'cmd' (ioctl command)
param3 changed to void * type and uses parameter name 'data'
param4 changed to struct ifnet * and uses parameter name 'ifp'
param5 has been removed (formerly struct lwp *) and uses of 'l' have been
replaced with curlwp from curproc(9).
callers have had (now unnecessary) casts to struct mbuf * removed, called
code has had (now unnecessary) casts to u_long, void * and struct ifnet *
respectively removed.
reviewed by rmind@
|
| 1.201 | 23-Jun-2014 |
rtr | where appropriate rename xxx_ioctl() struct mbuf * parameters from
`control' to `ifp' after split from xxx_usrreq().
sys_socket.c
fix wrapping of arguments to be consistent with other function calls
in the file after replacing pr_usrreq() call with pr_ioctl() which
required one less argument.
link_proto.c
fix indentation of parameters in link_ioctl() prototype to be
consistent with the rest of the file.
discussed with rmind@
|
| 1.200 | 22-Jun-2014 |
rtr | * split PRU_CONTROL functionality out of xxx_userreq() switches and place
into separate xxx_ioctl() functions.
* place KASSERT(req != PRU_CONTROL) inside xxx_userreq() as it is now
inappropriate for req = PRU_CONTROL in xxx_userreq().
* replace calls to pr_generic() with req = PRU_CONTROL with pr_ioctl().
* remove & fixup references to PRU_CONTROL xxx_userreq() function comments.
* fix various comments references for xxx_userreq() that mentioned
PRU_CONTROL as xxx_userreq() no longer handles the request.
a further change will follow to fix parameter and naming inconsistencies
retained from original code.
Reviewed by rmind@
|
| 1.199 | 30-May-2014 |
christos | Introduce 2 new variables: ipsec_enabled and ipsec_used.
Ipsec enabled is controlled by sysctl and determines if is allowed.
ipsec_used is set automatically based on ipsec being enabled, and
rules existing.
|
| 1.198 | 22-May-2014 |
rmind | Move udp6_input(), udp6_sendup(), udp6_realinput() and udp6_input_checksum()
from udp_usrreq.c to udp6_usrreq.c where they belong. No functional change.
|
| 1.197 | 20-May-2014 |
rmind | Adjust PR_WRAP_USRREQS() to include the attach/detach functions.
We still need the kernel-lock for some corner cases.
|
| 1.196 | 19-May-2014 |
rmind | - Split off PRU_ATTACH and PRU_DETACH logic into separate functions.
- Replace malloc with kmem and eliminate M_PCB while here.
- Sprinkle more asserts.
|
| 1.195 | 18-May-2014 |
rmind | Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.194 | 25-Feb-2014 |
pooka | branches: 1.194.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.
Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.
|
| 1.193 | 04-Jan-2014 |
pooka | IPv6 UDP uses the IPv4 pcb tables, and therefore the stats, so need
to create the percpu UDPv4 counters even in a v6-only system.
|
| 1.192 | 02-Jan-2014 |
pooka | Allow kernels compiled with INET+INET6 to be booted as IPv4-only or IPv6-only.
|
| 1.191 | 23-Nov-2013 |
christos | convert from CIRCLEQ to TAILQ.
|
| 1.190 | 05-Jun-2013 |
christos | branches: 1.190.2;
IPSEC has not come in two speeds for a long time now (IPSEC == kame,
FAST_IPSEC). Make everything refer to IPSEC to avoid confusion.
|
| 1.189 | 05-Jun-2013 |
christos | conditionalize the net traversal code on FAST_IPSEC to make rump build.
|
| 1.188 | 04-Jun-2013 |
christos | PR/47886: Dr. Wolfgang Stukenbrock: IPSEC_NAT_T enabled kernels may access
outdated pointers and pass ESP data to UPD-sockets.
While here, simplify the code and remove the IPSEC_NAT_T option; always
compile nat-traversal in so that it does not bitrot.
|
| 1.187 | 22-Jun-2012 |
christos | branches: 1.187.2;
PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.186 | 22-Mar-2012 |
drochner | remove KAME IPSEC, replaced by FAST_IPSEC
|
| 1.185 | 09-Jan-2012 |
liamjfoy | minor typo fix
|
| 1.184 | 19-Dec-2011 |
drochner | rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.
|
| 1.183 | 24-Sep-2011 |
christos | branches: 1.183.2; 1.183.6;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.182 | 17-Jul-2011 |
joerg | Retire varargs.h support. Move machine/stdarg.h logic into MI
sys/stdarg.h and expect compiler to provide proper builtins, defaulting
to the GCC interface. lint still has a special fallback.
Reduce abuse of _BSD_VA_LIST_ by defining __va_list by default and
derive va_list as required by standards.
|
| 1.181 | 03-Jul-2011 |
mrg | avoid an uninitialised variable warning. this one seems a false
positive, but since it's for some hacky workaround code anyway...
|
| 1.180 | 03-May-2011 |
dyoung | Reduces the resources demanded by TCP sessions in TIME_WAIT-state using
methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime
Truncation (MSLT).
MSLT and VTW were contributed by Coyote Point Systems, Inc.
Even after a TCP session enters the TIME_WAIT state, its corresponding
socket and protocol control blocks (PCBs) stick around until the TCP
Maximum Segment Lifetime (MSL) expires. On a host whose workload
necessarily creates and closes down many TCP sockets, the sockets & PCBs
for TCP sessions in TIME_WAIT state amount to many megabytes of dead
weight in RAM.
Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to
a class based on the nearness of the peer. Corresponding to each class
is an MSL, and a session uses the MSL of its class. The classes are
loopback (local host equals remote host), local (local host and remote
host are on the same link/subnet), and remote (local host and remote
host communicate via one or more gateways). Classes corresponding to
nearer peers have lower MSLs by default: 2 seconds for loopback, 10
seconds for local, 60 seconds for remote. Loopback and local sessions
expire more quickly when MSLT is used.
Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket
dead weight with a compact representation of the session, called a
"vestigial PCB". VTW data structures are designed to be very fast and
memory-efficient: for fast insertion and lookup of vestigial PCBs,
the PCBs are stored in a hash table that is designed to minimize the
number of cacheline visits per lookup/insertion. The memory both
for vestigial PCBs and for elements of the PCB hashtable come from
fixed-size pools, and linked data structures exploit this to conserve
memory by representing references with a narrow index/offset from the
start of a pool instead of a pointer. When space for new vestigial PCBs
runs out, VTW makes room by discarding old vestigial PCBs, oldest first.
VTW cooperates with MSLT.
It may help to think of VTW as a "FIN cache" by analogy to the SYN
cache.
A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT
sessions as fast as it can is approximately 17% idle when VTW is active
versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM
when VTW is active (approximately 64k vestigial PCBs are created) than
when it is inactive.
|
| 1.179 | 16-Sep-2009 |
pooka | branches: 1.179.4; 1.179.6;
Replace a large number of link set based sysctl node creations with
calls from subsystem constructors. Benefits both future kernel
modules and rump.
no change to sysctl nodes on i386/MONOLITHIC & build tested i386/ALL
|
| 1.178 | 19-Jul-2009 |
minskim | Enable IP_MINTTL option for SOCK_DGRAM sockets.
|
| 1.177 | 18-Apr-2009 |
tsutsui | Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch
|
| 1.176 | 18-Mar-2009 |
cegger | bcopy -> memcpy
|
| 1.175 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.174 | 19-Jan-2009 |
christos | branches: 1.174.2;
Provide compatibility to the old timeval SCM_TIMESTAMP messages.
|
| 1.173 | 06-Aug-2008 |
plunky | branches: 1.173.2;
Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.172 | 04-May-2008 |
thorpej | branches: 1.172.2; 1.172.6;
Simplify the interface to netstat_sysctl() and allocate space for
the collated counters using kmem_alloc().
PR kern/38577
|
| 1.171 | 26-Apr-2008 |
yamt | branches: 1.171.2;
udp_init: don't forget to allocate udp6stat_percpu.
|
| 1.170 | 24-Apr-2008 |
ad | Merge the socket locking patch:
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
|
| 1.169 | 23-Apr-2008 |
thorpej | Make IPSEC and FAST_IPSEC stats per-cpu. Use <net/net_stats.h> and
netstat_sysctl().
|
| 1.168 | 15-Apr-2008 |
thorpej | branches: 1.168.2;
Make udp6 stats per-cpu.
|
| 1.167 | 15-Apr-2008 |
thorpej | Make ip6 and icmp6 stats per-cpu.
|
| 1.166 | 12-Apr-2008 |
thorpej | Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.165 | 08-Apr-2008 |
thorpej | Change IPv6 stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old ip6stat structure; old netstat
binaries will continue to work properly.
|
| 1.164 | 06-Apr-2008 |
thorpej | Change UDP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old icmpstat structure; old netstat
binaries will continue to work properly.
|
| 1.163 | 27-Nov-2007 |
christos | branches: 1.163.14;
require that the options argument is the right size, not that it is greater
or equal to the requested size. Suggested by Matt Thomas.
|
| 1.162 | 02-Sep-2007 |
dyoung | branches: 1.162.6;
m_copym(..., 0, M_COPYALL, ...) -> m_copypacket(..., ...).
|
| 1.161 | 02-Sep-2007 |
dyoung | m_copy() was deprecated, apparently, long ago. m_copy(...) ->
m_copym(..., M_DONTWAIT).
|
| 1.160 | 27-Jun-2007 |
degroote | branches: 1.160.2; 1.160.6; 1.160.8;
Add support for options IPSEC_NAT_T (RFC 3947 and 3948) for fast_ipsec(4).
No objection on tech-net@
|
| 1.159 | 12-May-2007 |
dyoung | Use sockaddr_in_init().
|
| 1.158 | 04-Mar-2007 |
christos | branches: 1.158.2; 1.158.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.157 | 17-Feb-2007 |
dyoung | KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.156 | 14-Nov-2006 |
rpaulo | branches: 1.156.4;
Remove ifndef COMPAT_42. No objections in tech-net.
|
| 1.155 | 10-Nov-2006 |
yamt | udp_ctloutput: plug a memory leak.
|
| 1.154 | 10-Nov-2006 |
yamt | remove some __unused in function parameters.
|
| 1.153 | 10-Nov-2006 |
yamt | udp_ctloutput: remove unnecessary goto and break.
|
| 1.152 | 10-Nov-2006 |
yamt | udp_ctloutput: ansify.
|
| 1.151 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.150 | 10-Oct-2006 |
dogcow | change the MOWNER_INIT define to take two args; fix extant struct mowner
decls to use it. Makes options MBUFTRACE compile again and not whinge about
missing structure declarations. (Also makes initialization consistent.)
|
| 1.149 | 05-Oct-2006 |
tls | Protect calls to pool_put/pool_get that may occur in interrupt context
with spl used to protect other allocations and frees, or datastructure
element insertion and removal, in adjacent code.
It is almost unquestionably the case that some of the spl()/splx() calls
added here are superfluous, but it really seems wrong to see:
s=splfoo();
/* frob data structure */
splx(s);
pool_put(x);
and if we think we need to protect the first operation, then it is hard
to see why we should not think we need to protect the next. "Better
safe than sorry".
It is also almost unquestionably the case that I missed some pool
gets/puts from interrupt context with my strategy for finding these
calls; use of PR_NOWAIT is a strong hint that a pool may be used from
interrupt context but many callers in the kernel pass a "can wait/can't
wait" flag down such that my searches might not have found them. One
notable area that needs to be looked at is pf.
See also:
http://mail-index.netbsd.org/tech-kern/2006/07/19/0003.html
http://mail-index.netbsd.org/tech-kern/2006/07/19/0009.html
|
| 1.148 | 23-Jul-2006 |
ad | branches: 1.148.4; 1.148.6;
Use the LWP cached credentials where sane.
|
| 1.147 | 23-Feb-2006 |
christos | branches: 1.147.2;
Handle IPSEC_NAT_T in the FAST_IPSEC case.
XXX: need to fix the FAST_IPSEC code now.
|
| 1.146 | 21-Jan-2006 |
rpaulo | branches: 1.146.2; 1.146.4;
Better support of IPv6 scoped addresses.
- most of the kernel code will not care about the actual encoding of
scope zone IDs and won't touch "s6_addr16[1]" directly.
- similarly, most of the kernel code will not care about link-local
scoped addresses as a special case.
- scope boundary check will be stricter. For example, the current
*BSD code allows a packet with src=::1 and dst=(some global IPv6
address) to be sent outside of the node, if the application do:
s = socket(AF_INET6);
bind(s, "::1");
sendto(s, some_global_IPv6_addr);
This is clearly wrong, since ::1 is only meaningful within a single
node, but the current implementation of the *BSD kernel cannot
reject this attempt.
- and, while there, don't try to remove the ff02::/32 interface route
entry in in6_ifdetach() as it's already gone.
This also includes some level of support for the standard source
address selection algorithm defined in RFC3484, which will be
completed on in the future.
From the KAME project via JINMEI Tatuya.
Approved by core@.
|
| 1.145 | 11-Dec-2005 |
christos | branches: 1.145.2;
merge ktrace-lwp.
|
| 1.144 | 09-Dec-2005 |
manu | Fix a bug in ESP over UDP: because udp4_espinudp() called m_pullup, it
could modify the struct mbuf and calling functions (udp_input() and
udp4_realinput()) would have used a garbled local copy of the pointer.
The fix is not perfect. udp4_espinudp() should use m_pulldown()...
|
| 1.143 | 15-Nov-2005 |
dsl | Pass the current process structure to in_pcbconnect() so that it can
pass it to in_pcbbind() so that can allocate a low numbered port
if setsockopt() has been used to set IP_PORTRANGE to IP_PORTRANGE_LOW.
While there, fail in_pcbconnect() if the in_pcbbind() fails - rather
than sending the request out from a port of zero.
This has been largely broken since the socket option was added in 1998.
|
| 1.142 | 03-Sep-2005 |
kleink | branches: 1.142.6;
udp4_espinudp(): don't assume that the Non-ESP marker (or UDP payload)
is aligned on a 64-bit boundary.
|
| 1.141 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.140 | 10-Aug-2005 |
yamt | device independent part of ipv6 rx checksum offloading.
|
| 1.139 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.138 | 29-Apr-2005 |
manu | branches: 1.138.2;
Fix memory leak
|
| 1.137 | 25-Apr-2005 |
manu | Don't sleep when handling ESP over UDP packets.
|
| 1.136 | 23-Apr-2005 |
manu | Enhance IPSEC_NAT_T so that it can work with multiple machines behind the
same NAT.
|
| 1.135 | 18-Apr-2005 |
yamt | fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.134 | 11-Mar-2005 |
atatat | branches: 1.134.2;
Revert the change that made kern.file2 and net.*.*.pcblist into nodes
instead of structs. It had other deleterious side-effects that are
rather nasty. Another solution must be found.
|
| 1.133 | 10-Mar-2005 |
atatat | Change types of kern.file2 and net.*.*.pcblist to NODE
|
| 1.132 | 09-Mar-2005 |
atatat | Add the following nodes to the sysctl tree:
net.local.stream.pcblist
net.local.dgram.pcblist
net.inet.tcp.pcblist
net.inet.udp.pcblist
net.inet.raw.pcblist
net.inet6.tcp6.pcblist
net.inet6.udp6.pcblist
net.inet6.raw6.pcblist
which allow retrieval of the pcbs in use for those protocols. The
struct involved is 32/64 bit clean and incorporates parts of struct
inpcb, struct unpcb, a bit of struct tcpcb, and two socket addresses.
|
| 1.131 | 26-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.130 | 12-Feb-2005 |
manu | Add support for IPsec Network Address Translator traversal (NAT-T), as
described by RFC 3947 and 3948.
|
| 1.129 | 21-Dec-2004 |
yamt | branches: 1.129.2; 1.129.4;
factor out receive side tcp/udp checksum handling code so that they
can be used by eg. packet filters.
reviewed by Christos Zoulas on tech-net@.
(slightly tweaked since then to make tcp and udp similar.)
|
| 1.128 | 19-Dec-2004 |
christos | yamt's changes seem to fix all the checksumming issues. Turn the loopback
checksums back off so we can make sure that everything works.
|
| 1.127 | 18-Dec-2004 |
yamt | udp6_input: correct loopback test.
|
| 1.126 | 17-Dec-2004 |
christos | Turn checksumming on loopback back on until we fix the bugs in it.
Connect over tcp on the loopback is broken:
4729 amq 0.000007 CALL connect(4,0x804f2a0,0x1c)
4729 amq 75.007420 RET connect -1 errno 60 Connection timed out
|
| 1.125 | 15-Dec-2004 |
thorpej | Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.124 | 03-Sep-2004 |
darrenr | add a per-socket counter for dropped UDP packets when the internal buffers
are full.
|
| 1.123 | 02-Jul-2004 |
heas | Adjust description for net.inet.udp.checksum; it does not controll checking,
only computing.
|
| 1.122 | 25-May-2004 |
atatat | Sysctl descriptions under net subtree (net.key not done)
|
| 1.121 | 07-May-2004 |
jonathan | Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.120 | 01-May-2004 |
matt | Use EVCNT_ATTACH_STATIC{,2}
|
| 1.119 | 18-Apr-2004 |
matt | ANSI'fy and de __P
|
| 1.118 | 31-Mar-2004 |
itojun | clean previous commit (uh_sum != 0 check in IPv6)
|
| 1.117 | 31-Mar-2004 |
itojun | drop packet if IPv6 udp packet does not have checksum (checksum is mandatory
in IPv6).
|
| 1.116 | 24-Mar-2004 |
atatat | branches: 1.116.2;
Tango on sysctl_createv() and flags. The flags have all been renamed,
and sysctl_createv() now uses more arguments.
|
| 1.115 | 10-Mar-2004 |
drochner | fix tcp/udp checksum test in the M_CSUM_NO_PSEUDOHDR case
(this can never have worked)
now I can use a "bge" gigabit interface with hw checksumming
ttcp-t: 2147483648 bytes in 18.31 real seconds = 114527.11 KB/sec +++
woow!
|
| 1.114 | 04-Dec-2003 |
atatat | Dynamic sysctl.
Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(),
vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all
nodes are registered with the tree, and nodes can be added (or
removed) easily, and I/O to and from the tree is handled generically.
Since the nodes are registered with the tree, the mapping from name to
number (and back again) can now be discovered, instead of having to be
hard coded. Adding new nodes to the tree is likewise much simpler --
the new infrastructure handles almost all the work for simple types,
and just about anything else can be done with a small helper function.
All existing nodes are where they were before (numerically speaking),
so all existing consumers of sysctl information should notice no
difference.
PS - I'm sorry, but there's a distinct lack of documentation at the
moment. I'm working on sysctl(3/8/9) right now, and I promise to
watch out for buses.
|
| 1.113 | 23-Oct-2003 |
mycroft | Remove all the code to maintain ia_inpcbs. This information was only used to
close sockets on address changes, which was deemed to be a bad idea and was
summarily removed, so there is no point in wasting effort on maintaining it
any more.
|
| 1.112 | 18-Oct-2003 |
enami | Fix indent.
|
| 1.111 | 25-Sep-2003 |
mycroft | Fix glaring errors in recent changes.
|
| 1.110 | 12-Sep-2003 |
itojun | send icmp admin prohibit if socket policy mismatches.
|
| 1.109 | 04-Sep-2003 |
itojun | revamp inpcb/in6pcb so that they are more aligned with each other.
in6pcb lookup now uses hash(9).
|
| 1.108 | 22-Aug-2003 |
itojun | remove ipsec_set/getsocket. now we explicitly pass socket * to ip{,6}_output.
|
| 1.107 | 22-Aug-2003 |
itojun | change the additional arg to be passed to ip{,6}_output to struct socket *.
this fixes KAME policy lookup which was broken by the previous commit.
|
| 1.106 | 21-Aug-2003 |
jonathan | Honour the M_CSUM_NO_PSEUDOHDR, if set on inbound TCP and UDP packets.
Tested against bcm5700 with patched if_bge.c.
|
| 1.105 | 15-Aug-2003 |
jonathan | (fast-ipsec): Add hooks to pass IPv4 IPsec traffic into fast-ipsec, if
configured with ``options FAST_IPSEC''. Kernels with KAME IPsec or
with no IPsec should work as before.
All calls to ip_output() now always pass an additional compulsory
argument: the inpcb associated with the packet being sent,
or 0 if no inpcb is available.
Fast-ipsec tested with ICMP or UDP over ESP. TCP doesn't work, yet.
|
| 1.104 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.103 | 29-Jun-2003 |
fvdl | branches: 1.103.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.102 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.101 | 23-Jun-2003 |
martin | Make sure to include opt_foo.h if a defflag option FOO is used.
|
| 1.100 | 15-Jun-2003 |
matt | Change the way multicasts are kept. They now use a hash table in the same
manner as the ifaddr hash table. By doing this, the mkludge code can go
away. At the same time, keep track of what pcbs are using what ifaddr and
when an address is deleted from an interface, notify/abort all sockets
that have that address as a source. Switch IGMP and multicasts to use pools
for allocation. Fix a number of potential problems in the igmp code where
allocation failures could cause a trap/panic.
|
| 1.99 | 14-May-2003 |
itojun | always use PULLDOWN_TEST codepath.
|
| 1.98 | 26-Feb-2003 |
matt | Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.
|
| 1.97 | 20-Jan-2003 |
simonb | Remove variables that are only assigned too but not referenced.
|
| 1.96 | 14-Aug-2002 |
itojun | avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.
IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.
|
| 1.95 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.94 | 09-Jun-2002 |
itojun | whitespace
|
| 1.93 | 12-May-2002 |
matt | branches: 1.93.2; 1.93.4;
Eliminate commons.
|
| 1.92 | 21-Dec-2001 |
itojun | comment and whitespace. sync with kame
|
| 1.91 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.90 | 07-Nov-2001 |
itojun | do not grab packet to joined multicast group, when ip6_dst and in6p_laddr
mismatches. it makes the behavior more closer to 4.4BSD IPv4 code.
sync with kame
|
| 1.89 | 04-Nov-2001 |
matt | Convert netinet to not use the internal <sys/queue.h> field names
but instead the access macros. Use the FOREACH macros where appropriate.
|
| 1.88 | 02-Nov-2001 |
itojun | array boundary overflow on the use of IPv4 mapped address. from simonb
|
| 1.87 | 29-Oct-2001 |
simonb | Don't need to include <uvm/uvm_extern.h> just to include <sys/sysctl.h>
anymore.
|
| 1.86 | 24-Oct-2001 |
itojun | remove unused codepath (unifdef -UUDP6)
|
| 1.85 | 15-Oct-2001 |
itojun | branches: 1.85.2;
implement IPV6_V6ONLY socket option from draft-ietf-ipngwg-rfc2553bis-03.txt.
IPV6_BINDV6ONLY (netbsd only) is deprecated, but still work just like before.
|
| 1.84 | 17-Sep-2001 |
thorpej | Split the pre-computed ifnet checksum flags into Tx and Rx directions.
Add capabilities bits that indicate an interface can only perform
in-bound TCPv4 or UDPv4 checksums. There is at least one Gig-E chip
for which this is true (Level One LXT-1001), and this is also the
case for the Intel i82559 10/100 Ethernet chips.
|
| 1.83 | 25-Jul-2001 |
itojun | branches: 1.83.2;
allocate ipsec policy buffer attached to pcb in in*_pcballoc, before
giving anyone accesses to pcb (do not reveal an inconsistent ones).
sync with kame
|
| 1.82 | 25-Jul-2001 |
itojun | g/c #if 0'ed fragment. sync with kame.
|
| 1.81 | 17-Jul-2001 |
enami | Add missing counting up of ``socket buffer is full'' counter when
failed to sbappendaddr().
|
| 1.80 | 03-Jul-2001 |
itojun | branches: 1.80.2;
call in{,6}_pcbpurgeif0() before in{,6}_purgeif().
|
| 1.79 | 27-Jun-2001 |
itojun | fix udp reception to sockets bound to linklocal address (like fe80::1%lo0).
sync with kame
|
| 1.78 | 02-Jun-2001 |
thorpej | Implement support for IP/TCP/UDP checksum offloading provided by
network interfaces. This works by pre-computing the pseudo-header
checksum and caching it, delaying the actual checksum to ip_output()
if the hardware cannot perform the sum for us. In-bound checksums
can either be fully-checked by hardware, or summed up for final
verification by software. This method was modeled after how this
is done in FreeBSD, although the code is significantly different in
most places.
We don't delay checksums for IPv6/TCP, but we do take advantage of the
cached pseudo-header checksum.
Note: hardware-assisted checksumming defaults to "off". It is
enabled with ifconfig(8). See the manual page for details.
Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet,
3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
|
| 1.77 | 31-May-2001 |
soda | missing opt_inet.h
|
| 1.76 | 08-May-2001 |
itojun | correct faith prefix determination. use sys/netinet/if_faith.c:faithprefix()
to determine. sync with kame.
(without this change, non-faith socket may mistakenly accept for-faith traffic)
|
| 1.75 | 24-Jan-2001 |
itojun | branches: 1.75.2;
- record IPsec packet history into m_aux structure.
- let ipfilter look at wire-format packet only (not the decapsulated ones),
so that VPN setting can work with NAT/ipfilter settings.
sync with kame.
TODO: use header history for stricter inbound validation
|
| 1.74 | 04-Dec-2000 |
itojun | more on previous (udp4 multicast fix)
|
| 1.73 | 04-Dec-2000 |
itojun | fix multicast inbound packet processing.
NetBSD PR 11629 From: salvet@ics.muni.cz
|
| 1.72 | 17-Oct-2000 |
itojun | be more friendly with INET-less build.
XXX we need to do more to do a working INET-less build
|
| 1.71 | 30-Aug-2000 |
itojun | minor typo. s/iPsec/IPsec/
|
| 1.70 | 24-Jul-2000 |
sommerfeld | Drop packet, increment udps_badlen if the udp header length field
reports a size smaller than the udp header; defends against bogosity
detected by Assar Westerlund.
This patch and the previous ip_icmp.c change were the joint work of
assar, itojun, and myself.
|
| 1.69 | 07-Jul-2000 |
itojun | sync with kame.
introduce in6_{recover,embed}scope, for in-kernel scoped-address manipulation.
improve in6_pcbnotify.
|
| 1.68 | 06-Jul-2000 |
itojun | remove unnecessary #include <netkey/key_debug.h>. from kame.
|
| 1.67 | 28-Jun-2000 |
mrg | <vm/vm.h> -> <uvm/uvm_extern.h>
|
| 1.66 | 30-Mar-2000 |
augustss | branches: 1.66.4;
Remove register declarations.
|
| 1.65 | 30-Mar-2000 |
simonb | Delete reduncdant decl of inetctlerrmap - it's in <netinet/in_var.h>.
|
| 1.64 | 22-Mar-2000 |
ws | Make IPKDB working again.
Add support for i386 debugging and pci-based ne2000 boards.
|
| 1.63 | 01-Mar-2000 |
itojun | introduce m->m_pkthdr.aux to hold random data which needs to be passed
between protocol handlers.
ipsec socket pointers, ipsec decryption/auth information, tunnel
decapsulation information are in my mind - there can be several other usage.
at this moment, we use this for ipsec socket pointer passing. this will
avoid reuse of m->m_pkthdr.rcvif in ipsec code.
due to the change, MHLEN will be decreased by sizeof(void *) - for example,
for i386, MHLEN was 100 bytes, but is now 96 bytes.
we may want to increase MSIZE from 128 to 256 for some of our architectures.
take caution if you use it for keeping some data item for long period
of time - use extra caution on M_PREPEND() or m_adj(), as they may result
in loss of m->m_pkthdr.aux pointer (and mbuf leak).
this will bump kernel version.
(as discussed in tech-net, tested in kame tree)
|
| 1.62 | 29-Feb-2000 |
itojun | avoid copy-overwrite-copy on incoming udp4 checksum. use in4_cksum
which takes care of pseudo header checksum without overwrites.
|
| 1.61 | 11-Feb-2000 |
itojun | don't increase both "no port on broadcast packet" and "no port" stat.
increasing both of them will result in negative number on udp
"delivered" stat on netstat(8), since netstat computes number of delivered
packet by subtracting them from number of inbound packets.
|
| 1.60 | 02-Feb-2000 |
thorpej | PRU_PURGEADDR -> PRU_PURGEIF, per a discussion w/ itojun. In the IPv4
and IPv6 code, also use this to traverse PCB tables, looking for cached
routes referencing the dying ifnet, forcing them to be refreshed.
|
| 1.59 | 01-Feb-2000 |
thorpej | First-draft if_detach() implementation, originally from Bill Studnemund,
although this version has been changed somewhat:
- reference counting on ifaddrs isn't as complete as Bill's original
work was. This is hard to get right, and we should attack one
protocol at a time.
- This doesn't do reference counting or dynamic allocation of ifnets yet.
- This version introduces a new PRU -- PRU_PURGEADDR, which is used to
purge an ifaddr from a protocol. The old method Bill used didn't work
on all protocols, and it only worked on some because it was Very Lucky.
This mostly works ... i.e. works for my USB Ethernet, except for a dangling
ifaddr reference left by the IPv6 code; have not yet tracked this down.
|
| 1.58 | 31-Jan-2000 |
itojun | bring in latest KAME ipsec tree.
- interop issues in ipcomp is fixed
- padding type (after ESP) is configurable
- key database memory management (need more fixes)
- policy specification is revisited
XXX m->m_pkthdr.rcvif is still overloaded - hope to fix it soon
|
| 1.57 | 31-Jan-2000 |
itojun | destination port == 0 is illegal based on RFC768.
(NetBSD PR: 9137 - I thought I committed this already but I wasn't)
|
| 1.56 | 06-Jan-2000 |
itojun | remove extra portability #ifdef (like #ifdef __FreeBSD__) in KAME IPv6/IPsec
code, from netbsd-current repository.
#ifdef'ed version is always available from ftp.kame.net.
XXX please do not make too many diff-unfriendly changes, we'll need to take
bunch of diffs on upgrade...
|
| 1.55 | 06-Jan-2000 |
itojun | make IPV6_BINDV6ONLY setsockopt available. it controls behavior of
AF_INET6 wildcard listening socket. heavily documented in ip6(4).
net.inet6.ip6.bindv6only defines default value. default is 1.
"options INET6_BINDV6ONLY" removes any code fragment that supports
IPV6_BINDV6ONLY == 0 case (not defopt'ed as use of this is rare).
|
| 1.54 | 22-Dec-1999 |
itojun | drop IPv6 packets with v4 mapped address on src/dst. they are illegal
and may be used to fool IPv6 implementations (by using ::ffff:127.0.0.1 as
source you may be able to pretend the packet is from local node)
|
| 1.53 | 13-Dec-1999 |
itojun | sync IPv6 part with latest KAME tree. IPsec part is left unmodified
due to massive changes in KAME side.
- IPv6 output goes through nd6_output
- faith can capture IPv4 packets as well - you can run IPv4-to-IPv6 translator
using heavily modified DNS servers
- per-interface statistics (required for IPv6 MIB)
- interface autoconfig is revisited
- udp input handling has a big change for mapped address support.
- introduce in4_cksum() for non-overwriting checksumming
- introduce m_pulldown()
- neighbor discovery cleanups/improvements
- netinet/in.h strictly conforms to RFC2553 (no extra defs visible to userland)
- IFA_STATS is fixed a bit (not tested)
- and more more more.
TODO:
- cleanup os-independency #ifdef
- avoid rcvif dual use (for IPsec) to help ifdetach
(sorry for jumbo commit, I can't separate this any more...)
|
| 1.52 | 13-Sep-1999 |
itojun | branches: 1.52.2; 1.52.8;
- Call in{,6}_pcbdetach if ipsec initialization is failed during PRU_ATTACH.
This situation happens on severe memory shortage. We may need more
improvements here and there.
- Grab IEEE802 address from IFT_ETHER card, even if the card is
inserted after bootup time. Is there any other card that can be
inserted afterwards? pcmcia fddi card? :-P
- RFC2373 u bit handling suggests that we SHOULD NOT copy interface id from
ethernet card to pseudo interface, when ethernet card has IEEE802/EUI64
with u bit != 0 (this means that IEEE802/EUI64 is not universally unique).
Do not use such address as, for example, interface id for gif interface.
(I have such an ethernet card myself)
This may change interface id for your gif interface. be careful upgrading
rc files.
(sync with recent KAME)
|
| 1.51 | 09-Aug-1999 |
itojun | return with doing nothing from xx_ctlinput(), when sa->sa_family
is not the expected one.
I see PRC_REDIRECT_HOST with sa->sa_family == AF_UNIX coming to
{tcp,udp}_ctlinput() when I use dhclient, and I feel like adding
more sanity checks, without logging - if we log it it is too noisy.
|
| 1.50 | 09-Jul-1999 |
thorpej | defopt IPSEC and IPSEC_ESP (both into opt_ipsec.h).
|
| 1.49 | 06-Jul-1999 |
drochner | for incoming broadcasts, strip IP/UDP header correctly
wrap a line
|
| 1.48 | 01-Jul-1999 |
itojun | IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628.
(Sorry for a big commit, I can't separate this into several pieces...)
Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details.
- sys/kern: do not assume single mbuf, accept chained mbuf on passing
data from userland to kernel (or other way round).
- "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ
package (ftp://ftp.csl.sony.co.jp/pub/kjc/).
- sys/netinet/tcp*: IPv4/v6 dual stack tcp support.
- sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those
file to be there so we patch it up.
- sys/netinet: IPsec additions are here and there.
- sys/netinet6/*: most of IPv6 code sits here.
- sys/netkey: IPsec key management code
- dev/pci/pcidevs: regen
In my understanding no code here is subject to export control so it
should be safe.
|
| 1.47 | 19-Jan-1999 |
mycroft | branches: 1.47.4; 1.47.6;
Don't screw with ip_len; just subtract from it where we actually use the
value.
|
| 1.46 | 19-Jan-1999 |
mycroft | Don't overwrite the checksum fields when checking them. There's no reason to
do this, and it screws up ICMP replies.
XXX The returned IP checksum and length are still wrong.
|
| 1.45 | 11-Jan-1999 |
thorpej | Fix byte order and ip_len inconsistencies in ICMP reply code. Also, fix
some formatting and HTONS(foo) vs. foo = htons(foo) inconsistencies.
PR #6602, Darren Reed.
|
| 1.44 | 05-Jan-1998 |
thorpej | Finishing merging 4.4BSD-Lite2 netinet. At this point, the only changes
left were SCCS IDs and Copyright dates.
|
| 1.43 | 12-Sep-1997 |
drochner | Adjust packet len in mbuf header for incoming broadcasts.
Closes PR kern/4087 (by myself).
|
| 1.42 | 28-Jul-1997 |
thorpej | branches: 1.42.2;
Make the following tunable via sysctl, inspired by BSD/OS:
- udp_sendspace
- udp_recvspace
|
| 1.41 | 24-Jun-1997 |
thorpej | Don't adjust ip->ip_len before calling icmp_error(); icmp_error() already
does this. Per Stevens in TCP/IP Illustrated Vol. 2, p.774, submitted
by Koji Imada <koji@math.human.nagoya-u.ac.jp>.
|
| 1.40 | 11-Jan-1997 |
thorpej | Implement the IP_RECVIF socket option: supply a datagram packet's incoming
interface using a sockaddr_dl in a control mbuf.
Implement SO_TIMESTAMP for IP datagrams.
Move packet information option processing into a generic function
so that they work with multicast UDP and raw IP as well as unicast UDP.
Contributed by Bill Fenner <fenner@parc.xerox.com>.
|
| 1.39 | 25-Oct-1996 |
thorpej | In udp_output(), sanity check the length of the packet to be transmitted.
If it's larger than IP_MAXPACKET, return an error condition.
Based on a patch from Bill Fenner <fenner@parc.xerox.com>
|
| 1.38 | 16-Oct-1996 |
ws | Rename recently checked in KGDB to IPKDB to resolve conflicts with older KGDB
|
| 1.37 | 30-Sep-1996 |
ws | Add (and change) machine independent files for KGDB support
|
| 1.36 | 16-Sep-1996 |
mycroft | Make sure the sin_zero fields are filled.
|
| 1.35 | 15-Sep-1996 |
mycroft | Hash unconnected PCBs.
|
| 1.34 | 09-Sep-1996 |
mycroft | Add in_nullhost() and in_hosteq() macros, to hide some protocol
details. Also, fix a bug in TCP wrt SYN+URG packets.
|
| 1.33 | 23-May-1996 |
mycroft | udp_output() doesn't actually take control data, so don't pretend it does.
|
| 1.32 | 23-May-1996 |
mycroft | Make sure the control mbufs are freed in all cases.
|
| 1.31 | 23-May-1996 |
mycroft | Fix a race condition in PRU_DISCONNECT.
Rearrange the code to deal with unconnected sockets slightly.
Other minor changes.
|
| 1.30 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.29 | 20-May-1996 |
mrg | branches: 1.29.2;
if the sender set a cksum, check it, regardless if we care to
generate and send them ourselves. from rich stevens.
|
| 1.28 | 16-Mar-1996 |
christos | Fix printf format args.
|
| 1.27 | 13-Feb-1996 |
christos | netinet prototypes
|
| 1.26 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.25 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.24 | 12-Aug-1995 |
mycroft | branches: 1.24.2;
splnet --> splsoftnet
|
| 1.23 | 26-Jun-1995 |
cgd | fix typo
|
| 1.22 | 18-Jun-1995 |
cgd | convert pcb lists to CIRCLEQs, so that the end can be looked at more
easily, and so that the original (insque/remque) logic can be effectively
mimiced. (This fixes a bug in the previous set of list changes.)
also (since terminator is no longer null) reinstate uninitted list checks,
but mark them XXX.
|
| 1.21 | 12-Jun-1995 |
mycroft | Fix bogon in previous.
|
| 1.20 | 12-Jun-1995 |
mycroft | Change in_pcbnotify*() to take an errno value. Make inetctlerrmap[] an
array on ints, not u_chars.
|
| 1.19 | 12-Jun-1995 |
mycroft | Oops. Make source quench work again.
|
| 1.18 | 12-Jun-1995 |
mycroft | Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.17 | 04-Jun-1995 |
mycroft | Clean up many more casts.
|
| 1.16 | 01-Jun-1995 |
mycroft | Avoid byte-swapping IP addresses at run time.
|
| 1.15 | 13-Apr-1995 |
cgd | be a bit more careful and explicit with types. (basically a large no-op.)
|
| 1.14 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.13 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.12 | 10-Feb-1994 |
mycroft | Format police.
|
| 1.11 | 02-Feb-1994 |
hpeyerl | Multicast is no longer optional.
|
| 1.10 | 10-Jan-1994 |
mycroft | Should compile now with or without `options MULTICAST'.
|
| 1.9 | 08-Jan-1994 |
mycroft | More prototypes.
|
| 1.8 | 08-Jan-1994 |
mycroft | Slight rearrangement.
|
| 1.7 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.6 | 08-Jan-1994 |
mycroft | Fix some inconsistent spacing; spaces at the end of lines, etc.
|
| 1.5 | 18-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 06-Dec-1993 |
hpeyerl | multicast support.
>From Chris Maeda, cmaeda@cs.washington.edu
These patches are derived from the IP Multicast patches for BSDI.
|
| 1.3 | 22-May-1993 |
cgd | branches: 1.3.4;
add include of select.h if necessary for protos, or delete if extraneous
|
| 1.2 | 18-May-1993 |
cgd | make kernel select interface be one-stop shopping & clean it all up.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.3 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite2 for reference purposes.
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.3.4.1 | 24-Sep-1993 |
mycroft | Make all files using spl*() #include cpu.h. Changes from trunk.
|
| 1.24.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.29.2.2 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.29.2.1 | 10-Nov-1996 |
thorpej | Update from trunk:
- Make ip_len and ip_off unsigned.
- Make sure we don't accept or transmit packets larger than the
maximim IP packet size.
This fixes the so-called `death ping' bug.
Sum of work from Bill Fenner <fenner@parc.xerox.com>,
Kevin Lahey <kml@nas.nasa.gov>, and myself.
Thanks to Curt Sampson, Jukka Marin, and Kevin Lahey for testing
this under NetBSD 1.2
|
| 1.42.2.1 | 16-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.47.6.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.47.6.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.47.6.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.47.4.2 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.47.4.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.52.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.52.2.3 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.52.2.2 | 08-Dec-2000 |
bouyer | Sync with HEAD.
|
| 1.52.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.66.4.6 | 12-Apr-2004 |
jmc | Pullup patch (requested by itojun in ticket #134)
Drop packet if IPv6 udp packet does not have checksum.
|
| 1.66.4.5 | 09-May-2001 |
he | Pull up revision 1.76 (requested by itojun):
Correct faith prefix determintaion.
|
| 1.66.4.4 | 06-Apr-2001 |
he | Pull up revision 1.75 (requested by itojun):
Record IPsec packet history in m_aux structure. Let ipfilter
look at wire-format packet only (not the decapsulated ones), so
that VPN setting can work with NAT/ipfilter settings.
|
| 1.66.4.3 | 15-Dec-2000 |
he | Pull up revision 1.74 (requested by hubertf):
Fix multicast inbound packet processing. Fixes PR#11629.
(Required continuation of previous pullup.)
|
| 1.66.4.2 | 13-Dec-2000 |
he | Pull up revision 1.73 (requested by itojun):
Fix multicast inbound packet processing. Fixes PR#11629.
|
| 1.66.4.1 | 28-Jul-2000 |
sommerfeld | Pull up UDP, ICMP fixes:
- Drop packet, increment udps_badlen if the udp header length field
reports a size smaller than the udp header; defends against bogus
packets seen by by Assar Westerlund.
- allow icmp_error() to work when icmpreturndatabytes is sufficiently
large that the icmp error message doesn't fit in a header mbuf.
- defend against mbuf chains shorter than their contained ip->ip_len.
Joint work of myself, itojun, and assar
Approved by thorpej
revisions pulled up:
sys/netinet/ip_icmp.c 1.52
sys/netinet/udp_usrreq.c 1.70
|
| 1.75.2.9 | 27-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.75.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.75.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.75.2.6 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.75.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.75.2.4 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.75.2.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.75.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.75.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.80.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.80.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.80.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.80.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.83.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.85.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.93.4.2 | 31-Mar-2004 |
tron | Pull up revision 1.118 (requested by itojun in ticket #1645):
clean previous commit (uh_sum != 0 check in IPv6)
|
| 1.93.4.1 | 31-Mar-2004 |
tron | Pull up revision 1.117 (requested by itojun in ticket #1645):
drop packet if IPv6 udp packet does not have checksum (checksum is mandatory
in IPv6).
|
| 1.93.2.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.93.2.2 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.93.2.1 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.103.2.11 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.103.2.10 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.103.2.9 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.103.2.8 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.103.2.7 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.103.2.6 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.103.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.103.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.103.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.103.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.103.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.116.2.4 | 28-May-2004 |
tron | Pull up revision 1.122 (requested by atatat in ticket #391):
Sysctl descriptions under net subtree (net.key not done)
|
| 1.116.2.3 | 10-May-2004 |
tron | Pull up revision 1.121 (requested by jonathan in ticket #280):
Redo net.inet.* sysctl subtree for fast-ipsec from scratch.
Attach FAST-IPSEC statistics with 64-bit counters to new sysctl MIB.
Rework netstat to show FAST_IPSEC statistics, via sysctl, for
netstat -p ipsec.
New kernel files:
sys/netipsec/Makefile (new file; install *_var.h includes)
sys/netipsec/ipsec_var.h (new 64-bit mib counter struct)
Changed kernel files:
sys/Makefile (recurse into sys/netipsec/)
sys/netinet/in.h (fake IP_PROTO name for fast_ipsec
sysctl subtree.)
sys/netipsec/ipsec.h (minimal userspace inclusion)
sys/netipsec/ipsec_osdep.h (minimal userspace inclusion)
sys/netipsec/ipsec_netbsd.c (redo sysctl subtree from scratch)
sys/netipsec/key*.c (fix broken net.key subtree)
sys/netipsec/ah_var.h (increase all counters to 64 bits)
sys/netipsec/esp_var.h (increase all counters to 64 bits)
sys/netipsec/ipip_var.h (increase all counters to 64 bits)
sys/netipsec/ipcomp_var.h (increase all counters to 64 bits)
sys/netipsec/ipsec.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_mbuf.c (add #include netipsec/ipsec_var.h)
sys/netipsec/ipsec_output.c (add #include netipsec/ipsec_var.h)
sys/netinet/raw_ip.c (add #include netipsec/ipsec_var.h)
sys/netinet/tcp_input.c (add #include netipsec/ipsec_var.h)
sys/netinet/udp_usrreq.c (add #include netipsec/ipsec_var.h)
Changes to usr.bin/netstat to print the new fast-ipsec sysctl tree
for "netstat -s -p ipsec":
New file:
usr.bin/netstat/fast_ipsec.c (print fast-ipsec counters)
Changed files:
usr.bin/netstat/Makefile (add fast_ipsec.c)
usr.bin/netstat/netstat.h (declarations for fast_ipsec.c)
usr.bin/netstat/main.c (call KAME-vs-fast-ipsec dispatcher)
|
| 1.116.2.2 | 31-Mar-2004 |
tron | Pull up revision 1.118 (requested by itojun in ticket #28):
clean previous commit (uh_sum != 0 check in IPv6)
|
| 1.116.2.1 | 31-Mar-2004 |
tron | Pull up revision 1.117 (requested by itojun in ticket #28):
drop packet if IPv6 udp packet does not have checksum (checksum is mandatory
in IPv6).
|
| 1.129.4.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.129.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.129.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.134.2.6 | 29-Dec-2005 |
riz | Pull up following revision(s) (requested by manu in ticket #1052):
sys/netinet/udp_usrreq.c: revision 1.144
Fix a bug in ESP over UDP: because udp4_espinudp() called m_pullup, it
could modify the struct mbuf and calling functions (udp_input() and
udp4_realinput()) would have used a garbled local copy of the pointer.
The fix is not perfect. udp4_espinudp() should use m_pulldown()...
|
| 1.134.2.5 | 08-Sep-2005 |
tron | Pull up following revision(s) (requested by kleink in ticket #744):
sys/netinet/udp_usrreq.c: revision 1.142
udp4_espinudp(): don't assume that the Non-ESP marker (or UDP payload)
is aligned on a 64-bit boundary.
|
| 1.134.2.4 | 06-May-2005 |
tron | Pull up revision 1.135 (requested by yamt in ticket #251):
fix problems related to loopback interface checksum omission. PR/29971.
- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)
ok'ed by Jason Thorpe.
|
| 1.134.2.3 | 01-May-2005 |
tron | Pull up revision 1.138 (requested by manu in ticket #216):
Fix memory leak
|
| 1.134.2.2 | 28-Apr-2005 |
tron | Pull up revision 1.137 (requested by manu in ticket #203):
Don't sleep when handling ESP over UDP packets.
|
| 1.134.2.1 | 28-Apr-2005 |
tron | Pull up revision 1.136 (requested by man in ticket #201):
Enhance IPSEC_NAT_T so that it can work with multiple machines behind
the same NAT.
|
| 1.138.2.5 | 07-Dec-2007 |
yamt | sync with head
|
| 1.138.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.138.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.138.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.138.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.142.6.1 | 22-Nov-2005 |
yamt | sync with head.
|
| 1.145.2.2 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.145.2.1 | 01-Feb-2006 |
yamt | sync with head.
|
| 1.146.4.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.146.2.3 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.146.2.2 | 07-Feb-2006 |
rpaulo | in6pcb -> inpcb.
|
| 1.146.2.1 | 05-Feb-2006 |
rpaulo | <netinet6/in6_pcb.h> went away. Bye!
|
| 1.147.2.1 | 11-Aug-2006 |
yamt | sync with head
|
| 1.148.6.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.148.6.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.148.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.156.4.3 | 17-May-2007 |
yamt | sync with head.
|
| 1.156.4.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.156.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.158.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.158.2.3 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.158.2.2 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.158.2.1 | 08-Jun-2007 |
ad | Sync with head.
|
| 1.160.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.160.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.160.6.2 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.160.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.160.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.162.6.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.163.14.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.163.14.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.168.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.171.2.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.171.2.3 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.171.2.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.171.2.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.172.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.172.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.173.2.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.173.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.174.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.174.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.179.6.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.179.4.1 | 31-May-2011 |
rmind | sync with head
|
| 1.183.6.2 | 05-Apr-2012 |
mrg | sync to latest -current.
|
| 1.183.6.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.183.2.3 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.183.2.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.183.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.187.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.187.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.187.2.1 | 23-Jun-2013 |
tls | resync from head
|
| 1.190.2.5 | 18-May-2014 |
rmind | sync with head
|
| 1.190.2.4 | 17-Oct-2013 |
rmind | Eliminate some of the splsoftnet() calls, misc clean up.
|
| 1.190.2.3 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.190.2.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.190.2.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.194.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.217.10.1 | 06-Jun-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1607):
sys/netinet/udp_usrreq.c: revision 1.237 (via patch)
Fix three pretty bad mistakes in NAT-T:
* If we got a keepalive packet, we need to call m_freem, not m_free.
Here the next mbufs in the chain are not freed. Seems easy to remotely
DoS the system by sending fragmented keepalives in a loop.
* If !ipsec_used, free the mbuf.
* In udp_input, we need to update 'uh', because udp4_realinput may have
modified the chain. Perhaps we also need to re-enforce alignment, so
add an XXX.
|
| 1.217.6.1 | 07-Jun-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1607):
sys/netinet/udp_usrreq.c: revision 1.237 (via patch)
Fix three pretty bad mistakes in NAT-T:
* If we got a keepalive packet, we need to call m_freem, not m_free.
Here the next mbufs in the chain are not freed. Seems easy to remotely
DoS the system by sending fragmented keepalives in a loop.
* If !ipsec_used, free the mbuf.
* In udp_input, we need to update 'uh', because udp4_realinput may have
modified the chain. Perhaps we also need to re-enforce alignment, so
add an XXX.
|
| 1.217.4.9 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.217.4.8 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.217.4.7 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.217.4.6 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.217.4.5 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.217.4.4 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.217.4.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.217.4.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.217.4.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.217.2.1 | 06-Jun-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #1607):
sys/netinet/udp_usrreq.c: revision 1.237 (via patch)
Fix three pretty bad mistakes in NAT-T:
* If we got a keepalive packet, we need to call m_freem, not m_free.
Here the next mbufs in the chain are not freed. Seems easy to remotely
DoS the system by sending fragmented keepalives in a loop.
* If !ipsec_used, free the mbuf.
* In udp_input, we need to update 'uh', because udp4_realinput may have
modified the chain. Perhaps we also need to re-enforce alignment, so
add an XXX.
|
| 1.226.2.4 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.226.2.3 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.226.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.226.2.1 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.229.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.233.4.3 | 07-Jun-2018 |
martin | Pull up following revision(s) (requested by maxv in ticket #837):
sys/netinet/udp_usrreq.c: revision 1.237
Fix three pretty bad mistakes in NAT-T:
* If we got a keepalive packet, we need to call m_freem, not m_free.
Here the next mbufs in the chain are not freed. Seems easy to remotely
DoS the system by sending fragmented keepalives in a loop.
* If !ipsec_used, free the mbuf.
* In udp_input, we need to update 'uh', because udp4_realinput may have
modified the chain. Perhaps we also need to re-enforce alignment, so
add an XXX.
|
| 1.233.4.2 | 09-Apr-2018 |
bouyer | Pull up following revision(s) (requested by roy in ticket #724):
tests/net/icmp/t_ping.c: revision 1.19
sys/netinet6/raw_ip6.c: revision 1.166
sys/netinet6/ip6_input.c: revision 1.195
sys/net/raw_usrreq.c: revision 1.59
sys/sys/socketvar.h: revision 1.151
sys/kern/uipc_socket2.c: revision 1.128
tests/lib/libc/sys/t_recvmmsg.c: revision 1.2
lib/libc/sys/recv.2: revision 1.38
sys/net/rtsock.c: revision 1.239
sys/netinet/udp_usrreq.c: revision 1.246
sys/netinet6/icmp6.c: revision 1.224
tests/net/icmp/t_ping.c: revision 1.20
sys/netipsec/keysock.c: revision 1.63
sys/netinet/raw_ip.c: revision 1.172
sys/kern/uipc_socket.c: revision 1.260
tests/net/icmp/t_ping.c: revision 1.22
sys/kern/uipc_socket.c: revision 1.261
tests/net/icmp/t_ping.c: revision 1.23
sys/netinet/ip_mroute.c: revision 1.155
sbin/route/route.c: revision 1.159
sys/netinet6/ip6_mroute.c: revision 1.123
sys/netatalk/ddp_input.c: revision 1.31
sys/netcan/can.c: revision 1.3
sys/kern/uipc_usrreq.c: revision 1.184
sys/netinet6/udp6_usrreq.c: revision 1.138
tests/net/icmp/t_ping.c: revision 1.18
socket: report receive buffer overflows
Add soroverflow() which increments the overflow counter, sets so_error
to ENOBUFS and wakes the receive socket up.
Replace all code that manually increments this counter with soroverflow().
Add soroverflow() to raw_input().
This allows userland to detect route(4) overflows so it can re-sync
with the current state.
socket: clear error even when peeking
The error has already been reported and it's pointless requiring another
recv(2) call just to clear it.
socket: remove now incorrect comment that so_error is only udp
As it can be affected by route(4) sockets which are raw.
rtsock: log dropped messages that we cannot report to userland
Handle ENOBUFS when receiving messages.
Don't send messages if the receiver has died.
Sprinkle more soroverflow().
Handle ENOBUFS in recv
Handle ENOBUFS in sendto
Note value received. Harden another sendto for ENOBUFS.
Handle the routing socket overflowing gracefully.
Allow a valid sendto .... duh
Handle errors better.
Fix test for checking we sent all the data we asked to.
|
| 1.233.4.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.245.2.9 | 18-Jan-2019 |
pgoyette | Synch with HEAD
|
| 1.245.2.8 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.245.2.7 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.245.2.6 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.245.2.5 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.245.2.4 | 21-May-2018 |
pgoyette | Sync with HEAD
|
| 1.245.2.3 | 02-May-2018 |
pgoyette | Synch with HEAD
|
| 1.245.2.2 | 16-Apr-2018 |
pgoyette | Sync with HEAD, resolve some conflicts
|
| 1.245.2.1 | 22-Mar-2018 |
pgoyette | Synch with HEAD, resolve conflicts
|
| 1.254.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.259.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.264.8.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.48 | 03-Feb-2021 |
roy | Sprinkle CTASSERT to enforce on-wire layout without __packed
|
| 1.47 | 03-Feb-2021 |
roy | Remove __packed from various network structures
They are already network aligned and adding the __packed attribute
just causes needless compiler warnings about accssing members of packed
objects.
|
| 1.46 | 20-Aug-2020 |
riastradh | branches: 1.46.2;
[ozaki-r] Changes to the kernel core for wireguard
|
| 1.45 | 14-Sep-2018 |
maxv | Use non-variadic function pointer in protosw::pr_input.
|
| 1.44 | 22-Aug-2018 |
msaitoh | - Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.
|
| 1.43 | 08-Feb-2018 |
maxv | branches: 1.43.2; 1.43.4;
Style, and remove prototype of udp_sysctl (does not exist).
|
| 1.42 | 10-Aug-2017 |
ryo | Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
|
| 1.41 | 20-Jan-2016 |
riastradh | branches: 1.41.10;
Give proper prototype to udp_output.
|
| 1.40 | 18-May-2014 |
rmind | branches: 1.40.4;
Add struct pr_usrreqs with a pr_generic function and prepare for the
dismantling of pr_usrreq in the protocols; no functional change intended.
PRU_ATTACH/PRU_DETACH changes will follow soon.
Bump for struct protosw. Welcome to 6.99.62!
|
| 1.39 | 02-Jan-2014 |
pooka | branches: 1.39.2;
Allow kernels compiled with INET+INET6 to be booted as IPv4-only or IPv6-only.
|
| 1.38 | 22-Jun-2012 |
christos | branches: 1.38.2; 1.38.4;
PR/46602: Move the rfc6056 port randomization to the IP layer.
|
| 1.37 | 24-Sep-2011 |
christos | branches: 1.37.2;
Add inet4 part of the rfc6056 code contributed by Vlad Balan as part of
Google SoC-2011
|
| 1.36 | 06-Aug-2008 |
plunky | Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.
approved by core
|
| 1.35 | 12-Apr-2008 |
thorpej | branches: 1.35.4; 1.35.6; 1.35.10;
Make IP, TCP, UDP, and ICMP statistics per-CPU. The stats are collated
when the user requests them via sysctl.
|
| 1.34 | 06-Apr-2008 |
thorpej | Change UDP stats from a structure to an array of uint64_t's.
Note: This is ABI-compatible with the old icmpstat structure; old netstat
binaries will continue to work properly.
|
| 1.33 | 25-Dec-2007 |
perry | branches: 1.33.6;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.32 | 17-Feb-2007 |
dyoung | branches: 1.32.18; 1.32.24; 1.32.26; 1.32.30;
KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous
parentheses in return statements.
Cosmetic: don't open-code TAILQ_FOREACH().
Cosmetic: change types of variables to avoid oodles of casts: in
in6_src.c, avoid casts by changing several route_in6 pointers
to struct route pointers. Remove unnecessary casts to caddr_t
elsewhere.
Pave the way for eliminating address family-specific route caches:
soon, struct route will not embed a sockaddr, but it will hold
a reference to an external sockaddr, instead. We will set the
destination sockaddr using rtcache_setdst(). (I created a stub
for it, but it isn't used anywhere, yet.) rtcache_free() will
free the sockaddr. I have extracted from rtcache_free() a helper
subroutine, rtcache_clear(). rtcache_clear() will "forget" a
cached route, but it will not forget the destination by releasing
the sockaddr. I use rtcache_clear() instead of rtcache_free()
in rtcache_update(), because rtcache_update() is not supposed
to forget the destination.
Constify:
1 Introduce const accessor for route->ro_dst, rtcache_getdst().
2 Constify the 'dst' argument to ifnet->if_output(). This
led me to constify a lot of code called by output routines.
3 Constify the sockaddr argument to protosw->pr_ctlinput. This
led me to constify a lot of code called by ctlinput routines.
4 Introduce const macros for converting from a generic sockaddr
to family-specific sockaddrs, e.g., sockaddr_in: satocsin6,
satocsin, et cetera.
|
| 1.31 | 11-Dec-2005 |
christos | branches: 1.31.26;
merge ktrace-lwp.
|
| 1.30 | 10-Dec-2005 |
elad | Multiple inclusion protection, as suggested by christos@ on tech-kern@
few days ago.
|
| 1.29 | 10-Aug-2005 |
yamt | move {tcp,udp}_do_loopback_cksum back to tcp/udp
so that they can be referenced by ipv6.
|
| 1.28 | 05-Aug-2005 |
elad | Add sysctls for IP, ICMP, TCP, and UDP statistics.
|
| 1.27 | 12-Feb-2005 |
manu | branches: 1.27.6;
Add support for IPsec Network Address Translator traversal (NAT-T), as
described by RFC 3947 and 3948.
|
| 1.26 | 21-Dec-2004 |
yamt | branches: 1.26.2; 1.26.4;
factor out receive side tcp/udp checksum handling code so that they
can be used by eg. packet filters.
reviewed by Christos Zoulas on tech-net@.
(slightly tweaked since then to make tcp and udp similar.)
|
| 1.25 | 15-Dec-2004 |
thorpej | Don't perform checksums on loopback interfaces. They can be reenabled with
the net.inet.*.do_loopback_cksum sysctl.
Approved by: groo
|
| 1.24 | 21-Apr-2004 |
itojun | no space between function name and paren: foo (blah) -> foo(blah)
|
| 1.23 | 18-Apr-2004 |
matt | De __P()
|
| 1.22 | 07-Aug-2003 |
agc | Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22364, verified by myself.
|
| 1.21 | 29-Jun-2003 |
fvdl | branches: 1.21.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.20 | 28-Jun-2003 |
darrenr | Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.
Bump the kernel rev up to 1.6V
|
| 1.19 | 30-Jun-2002 |
thorpej | Changes to allow the IPv4 and IPv6 layers to align headers themseves,
as necessary:
* Implement a new mbuf utility routine, m_copyup(), is is like
m_pullup(), except that it always prepends and copies, rather
than only doing so if the desired length is larger than m->m_len.
m_copyup() also allows an offset into the destination mbuf, which
allows space for packet headers, in the forwarding case.
* Add *_HDR_ALIGNED_P() macros for IP, IPv6, ICMP, and IGMP. These
macros expand to 1 if __NO_STRICT_ALIGNMENT is defined, so that
architectures which do not have strict alignment constraints don't
pay for the test or visit the new align-if-needed path.
* Use the new macros to check if a header needs to be aligned, or to
assert that it already is, as appropriate.
Note: This code is still somewhat experimental. However, the new
code path won't be visited if individual device drivers continue
to guarantee that packets are delivered to layer 3 already properly
aligned (which are rules that are already in use).
|
| 1.18 | 12-May-2002 |
matt | branches: 1.18.2;
Eliminate commons.
|
| 1.17 | 20-Nov-1999 |
thorpej | branches: 1.17.6; 1.17.8;
Add the `packed' attribute to structures which describe wire protocol data.
|
| 1.16 | 19-Nov-1999 |
bouyer | Update protocoles and interfaces stats counters to 64bit.
RTM_IFINFO is now 0xf, 0xe is RTM_OIFINFO which returns the old (if_msghdr14)
struct with 32bit counters (binary compat, conditioned on COMPAT_14).
Same for sysctl: node 3 is renamed NET_RT_OIFLIST, NET_RT_IFLIST is now node 4.
Change rt_msg1() to add an mbuf to the mbuf chain instead of just panic()
when the message is larger than MHLEN.
|
| 1.15 | 10-Feb-1998 |
perry | branches: 1.15.12; 1.15.14; 1.15.20;
add/cleanup multiple inclusion protection.
|
| 1.14 | 28-Jul-1997 |
thorpej | Make the following tunable via sysctl, inspired by BSD/OS:
- udp_sendspace
- udp_recvspace
|
| 1.13 | 22-May-1996 |
mycroft | Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and
PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait.
Remove SS_PRIV completely.
|
| 1.12 | 13-Feb-1996 |
christos | branches: 1.12.4;
netinet prototypes
|
| 1.11 | 31-Jan-1996 |
mycroft | Build a hash table of PCBs. Hash function needs tweaking.
|
| 1.10 | 21-Nov-1995 |
cgd | make netinet work on systems where pointers and longs are 64 bits
(like the alpha). Biggest problem: IP headers were overlayed with
structure which included pointers, and which therefore didn't overlay
properly on 64-bit machines. Solution: instead of threading pointers
through IP header overlays, add a "queue element" structure to do
the threading, and point it at the ip headers.
|
| 1.9 | 12-Jun-1995 |
mycroft | branches: 1.9.2;
Various cleanup, including:
* Convert several data structures to use queue.h.
* Split in_pcbnotify() into two parts; one for notifying a specific PCB, and
one for notifying all PCBs for a particular foreign address.
|
| 1.8 | 26-Mar-1995 |
jtc | KERNEL -> _KERNEL
|
| 1.7 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.6 | 13-May-1994 |
mycroft | Update to 4.4-Lite networking code, with a few local changes.
|
| 1.5 | 10-Jan-1994 |
mycroft | Change the counters to be all the same type -- u_long.
|
| 1.4 | 08-Jan-1994 |
mycroft | Prototypes.
|
| 1.3 | 20-May-1993 |
cgd | more rcsid additions and file header cleanups
|
| 1.2 | 19-Apr-1993 |
mycroft | Add consistent multiple-inclusion protection.
|
| 1.1 | 21-Mar-1993 |
cgd | branches: 1.1.1;
Initial revision
|
| 1.1.1.2 | 05-Jan-1998 |
thorpej | Import sys/netinet from 4.4BSD-Lite for reference purposes.
|
| 1.1.1.1 | 21-Mar-1993 |
cgd | initial import of 386bsd-0.1 sources
|
| 1.9.2.1 | 02-Feb-1996 |
mycroft | Bring in changes for mondo patch 2.
|
| 1.12.4.1 | 11-Dec-1996 |
mycroft | From trunk:
Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods
that need it.
Fix numerous memory leaks and bogus return values.
|
| 1.15.20.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.15.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
|
| 1.15.12.3 | 30-Nov-1999 |
itojun | bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch
just for reference purposes.
This commit includes 1.4 -> 1.4.1 sync for kame branch.
The branch does not compile at all (due to the lack of ALTQ and some other
source code). Please do not try to modify the branch, this is just for
referenre purposes.
synchronization to latest KAME will take place on HEAD branch soon.
|
| 1.15.12.2 | 06-Jul-1999 |
itojun | KAME/NetBSD 1.4, SNAP kit 1999/07/05.
NOTE: this branch is just for reference purposes (i.e. for taking cvs diff).
do not touch anything on the branch. actual work must be done on HEAD branch.
|
| 1.15.12.1 | 28-Jun-1999 |
itojun | KAME/NetBSD 1.4 SNAP kit, dated 19990628.
NOTE: this branch (kame) is used just for refernce. this may not compile
due to multiple reasons.
|
| 1.17.8.2 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.17.8.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.17.6.2 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.17.6.1 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.18.2.1 | 15-Jul-2002 |
gehenna | catch up with -current.
|
| 1.21.2.9 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.21.2.8 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.21.2.7 | 15-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.21.2.6 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.21.2.5 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.21.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.21.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.21.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.21.2.1 | 02-Jul-2003 |
darrenr | Apply the aborted ktrace-lwp changes to a specific branch. This is just for
others to review, I'm concerned that patch fuziness may have resulted in some
errant code being generated but I'll look at that later by comparing the diff
from the base to the branch with the file I attempt to apply to it. This will,
at the very least, put the changes in a better context for others to review
them and attempt to tinker with removing passing of 'struct lwp' through
the kernel.
|
| 1.26.4.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.26.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.27.6.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.27.6.2 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.27.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.31.26.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.32.30.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.32.26.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.32.24.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.32.18.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.33.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.33.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.35.10.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.35.6.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.35.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.37.2.2 | 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.37.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.38.4.4 | 18-May-2014 |
rmind | sync with head
|
| 1.38.4.3 | 23-Sep-2013 |
rmind | - Add some initial locking to the IPv4 PCB.
- Rename inpcb_lookup_*() routines to be more accurate and add comments.
- Add some comments about connection life-cycle WRT socket layer.
|
| 1.38.4.2 | 28-Aug-2013 |
rmind | Checkpoint work in progress:
- Initial split of the protocol user-request method into the following
methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq.
- Adjust socreate(9) and sonewconn(9) to call pr_attach without the
socket lock held (as a preparation for the locking scheme adjustment).
- Adjust all pr_attach routines to assert that PCB is not set.
- Sprinkle various comments, document some routines and their locking.
- Remove M_PCB, replace with kmem(9).
- Fix few bugs spotted on the way.
|
| 1.38.4.1 | 17-Jul-2013 |
rmind | Checkpoint work in progress:
- Move PCB structures under __INPCB_PRIVATE, adjust most of the callers
and thus make IPv4 PCB structures mostly opaque. Any volunteers for
merging in6pcb with inpcb (see rpaulo-netinet-merge-pcb branch)?
- Move various global vars to the modules where they belong, make them static.
- Some preliminary work for IPv4 PCB locking scheme.
- Make raw IP code mostly MP-safe. Simplify some of it.
- Rework "fast" IP forwarding (ipflow) code to be mostly MP-safe. It should
run from a software interrupt, rather than hard.
- Rework tun(4) pseudo interface to be MP-safe.
- Work towards making some other interfaces more strict.
|
| 1.38.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.38.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.39.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.40.4.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.40.4.1 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.41.10.1 | 21-Dec-2017 |
snj | Pull up following revision(s) (requested by ryo in ticket #445):
distrib/sets/lists/debug/mi: revision 1.222
distrib/sets/lists/tests/mi: revision 1.760
share/man/man4/ip.4: revision 1.38
sys/netinet/in.c: revision 1.207
sys/netinet/in.h: revision 1.101
sys/netinet/in_pcb.c: revision 1.179
sys/netinet/in_pcb.h: revision 1.64
sys/netinet/ip_output.c: revision 1.284, 1.286
sys/netinet/ip_var.h: revision 1.120-1.121
sys/netinet/raw_ip.c: revision 1.166-1.167
sys/netinet/udp_usrreq.c: revision 1.235-1.236
sys/netinet/udp_var.h: revision 1.42
tests/net/net/Makefile: revision 1.21
tests/net/net/t_pktinfo_send.c: revision 1.1-1.2
Add support IP_PKTINFO for sendmsg(2).
The source address or output interface can be specified by adding IP_PKTINFO
to the control part of the message on a SOCK_DGRAM or SOCK_RAW socket.
Reviewed by ozaki-r@ and christos@. thanks.
--
As is the case with IPV6_PKTINFO, IP_PKTINFO can be sent without EADDRINUSE
even if the UDP address:port in use is specified.
|
| 1.43.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.43.2.2 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.43.2.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.46.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.8 | 07-Feb-2020 |
thorpej | Use percpu_foreach_xcall() to gather volatile per-cpu counters. These
must be serialized against the interrupts / soft-interrupts in which
they're manipulated, as well as protected from non-atomic 64-bit memory
loads on 32-bit platforms.
|
| 1.7 | 01-Feb-2020 |
riastradh | Switch sys/net to percpu_create.
|
| 1.6 | 19-Sep-2019 |
ozaki-r | branches: 1.6.2;
wqinput: avoid having struct wqinput_worklist directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Input handlers of wqinput normally involves sleepable operations so we must
avoid dereferencing a percpu data (struct wqinput_worklist) after executing
an input handler. Address this situation by having just a pointer to the data
in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
|
| 1.5 | 10-Aug-2018 |
msaitoh | branches: 1.5.4;
Change the type of wqinput's drop counter to uint64_t. OK'd by ozaki-r@.
|
| 1.4 | 24-Feb-2018 |
ozaki-r | branches: 1.4.2; 1.4.4;
Avoid a deadlock between softnet_lock and IFNET_LOCK
A deadlock occurs because there is a violation of the rule of lock ordering;
softnet_lock is held with hodling IFNET_LOCK, which violates the rule.
To avoid the deadlock, replace softnet_lock in in_control and in6_control
with KERNEL_LOCK.
We also need to add some KERNEL_LOCKs to protect the network stack surely.
This is required, for example, for PR kern/51356.
Fix PR kern/53043
|
| 1.3 | 02-Jun-2017 |
para | branches: 1.3.2; 1.3.8;
pool_init does not copy its name argument
therefore don't pass in a stack allocated buffer
vmstat -mv shows pool(s) with broken name(s)
use the name argument passed into wqinput_create directly
which is a static string in all 4 callee cases
(workqueue_create/workqueue_init copies the name argument)
|
| 1.2 | 21-May-2017 |
ozaki-r | Add missing NULL check for pool_get call with PR_NOWAIT
This should fix a kernel panic reported by wiz@ on current-users ML:
http://mail-index.netbsd.org/current-users/2017/05/03/msg031646.html
|
| 1.1 | 02-Feb-2017 |
ozaki-r | branches: 1.1.2; 1.1.4; 1.1.8;
Defer some pr_input to workqueue
pr_input is currently called in softint. Some pr_input such as ICMP, ICMPv6
and CARP can add/delete/update IP addresses and routing table entries. For
example, icmp6_redirect_input updates an a routing table entry and
nd6_ra_input may delete an IP address.
Basically such operations shouldn't be done in softint. That aside, we have
a reason to avoid the situation; psz/psref waits cannot be used in softint,
however they are required to work in such pr_input in the MP-safe world.
The change implements the workqueue pr_input framework called wqinput which
provides a means to defer pr_input of a protocol to workqueue easily.
Currently icmp_input, icmp6_input, carp_proto_input and carp6_proto_input
are deferred to workqueue by the framework.
Proposed and discussed on tech-kern and tech-net
|
| 1.1.8.2 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.1.8.1 | 02-Feb-2017 |
bouyer | file wqinput.c was added on branch bouyer-socketcan on 2017-04-21 16:54:06 +0000
|
| 1.1.4.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.1.4.1 | 02-Feb-2017 |
pgoyette | file wqinput.c was added on branch pgoyette-localcount on 2017-03-20 06:57:51 +0000
|
| 1.1.2.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.1.2.2 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 02-Feb-2017 |
skrll | file wqinput.c was added on branch nick-nhusb on 2017-02-05 13:40:59 +0000
|
| 1.3.8.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.3.8.1 | 02-Jun-2017 |
jdolecek | file wqinput.c was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.3.2.2 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by knakahara in ticket #1385):
sys/net/if.c 1.461
sys/net/if.h 1.277
sys/net/if_gif.c 1.149
sys/net/if_gif.h 1.33
sys/net/if_ipsec.c 1.19,1.20,1.24
sys/net/if_ipsec.h 1.5
sys/net/if_l2tp.c 1.33,1.36-1.39
sys/net/if_l2tp.h 1.7,1.8
sys/net/route.c 1.220,1.221
sys/net/route.h 1.125
sys/netinet/in_gif.c 1.95
sys/netinet/in_l2tp.c 1.17
sys/netinet/ip_input.c 1.391,1.392
sys/netinet/wqinput.c 1.6
sys/netinet6/in6_gif.c 1.94
sys/netinet6/in6_l2tp.c 1.18
sys/netinet6/ip6_forward.c 1.97
sys/netinet6/ip6_input.c 1.210,1.211
sys/netipsec/ipsec_output.c 1.82,1.83 (patched)
sys/netipsec/ipsecif.c 1.12,1.13,1.15,1.17 (patched)
sys/netipsec/key.c 1.259,1.260
ipsecif(4) support input drop packet counter.
ipsecif(4) should not increment drop counter by errors not related to if_snd. Pointed out by ozaki-r@n.o, thanks.
Remove unnecessary addresses in PF_KEY message.
MOBIKE Extensions for PF_KEY draft-schilcher-mobike-pfkey-extension-01.txt says
|
| 1.3.2.1 | 26-Feb-2018 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #588):
sys/netinet6/in6.c: revision 1.260
sys/netinet/in.c: revision 1.219
sys/netinet/wqinput.c: revision 1.4
sys/rump/net/lib/libnetinet/netinet_component.c: revision 1.11
sys/netinet/ip_input.c: revision 1.376
sys/netinet6/ip6_input.c: revision 1.193
Avoid a deadlock between softnet_lock and IFNET_LOCK
A deadlock occurs because there is a violation of the rule of lock ordering;
softnet_lock is held with hodling IFNET_LOCK, which violates the rule.
To avoid the deadlock, replace softnet_lock in in_control and in6_control
with KERNEL_LOCK.
We also need to add some KERNEL_LOCKs to protect the network stack surely.
This is required, for example, for PR kern/51356.
Fix PR kern/53043
|
| 1.4.4.2 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.4.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.4.2.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.5.4.1 | 24-Sep-2019 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #238):
sys/netipsec/ipsec_output.c: revision 1.83
sys/net/route.h: revision 1.125
sys/netinet6/ip6_input.c: revision 1.210
sys/netinet6/ip6_input.c: revision 1.211
sys/net/if.c: revision 1.461
sys/net/if_gif.h: revision 1.33
sys/net/route.c: revision 1.220
sys/net/route.c: revision 1.221
sys/net/if.h: revision 1.277
sys/netinet6/ip6_forward.c: revision 1.97
sys/netinet/wqinput.c: revision 1.6
sys/net/if_ipsec.h: revision 1.5
sys/netinet6/in6_l2tp.c: revision 1.18
sys/netinet6/in6_gif.c: revision 1.94
sys/net/if_l2tp.h: revision 1.7
sys/net/if_gif.c: revision 1.149
sys/net/if_l2tp.h: revision 1.8
sys/netinet/in_gif.c: revision 1.95
sys/netinet/in_l2tp.c: revision 1.17
sys/netipsec/ipsecif.c: revision 1.17
sys/net/if_ipsec.c: revision 1.24
sys/net/if_l2tp.c: revision 1.37
sys/netinet/ip_input.c: revision 1.391
sys/net/if_l2tp.c: revision 1.38
sys/netinet/ip_input.c: revision 1.392
sys/net/if_l2tp.c: revision 1.39
Avoid having a rtcache directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
wqinput: avoid having struct wqinput_worklist directly in a percpu storage
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Input handlers of wqinput normally involves sleepable operations so we must
avoid dereferencing a percpu data (struct wqinput_worklist) after executing
an input handler. Address this situation by having just a pointer to the data
in a percpu storage instead.
Reviewed by knakahara@ and yamaguchi@
-
Add missing #include <sys/kmem.h>
-
Divide Tx context of l2tp(4) to improve performance.
It seems l2tp(4) call path is too long for instruction cache. So, dividing
l2tp(4) Tx context improves CPU use efficiency.
After this commit, l2tp(4) throughput gains 10% on my machine(Atom C3000).
-
Apply some missing changes lost on the previous commit
-
Avoid having a rtcache directly in a percpu storage for tunnel protocols.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Using rtcache, i.e., packet processing, typically involves sleepable operations
such as rwlock so we must avoid dereferencing a rtcache that is directly stored
in a percpu storage during packet processing. Address this situation by having
just a pointer to a rtcache in a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
-
l2tp(4): avoid having struct ifqueue directly in a percpu storage.
percpu(9) has a certain memory storage for each CPU and provides it by the piece
to users. If the storages went short, percpu(9) enlarges them by allocating new
larger memory areas, replacing old ones with them and destroying the old ones.
A percpu storage referenced by a pointer gotten via percpu_getref can be
destroyed by the mechanism after a running thread sleeps even if percpu_putref
has not been called.
Tx processing of l2tp(4) uses normally involves sleepable operations so we
must avoid dereferencing a percpu data (struct ifqueue) after executing Tx
processing. Address this situation by having just a pointer to the data in
a percpu storage instead.
Reviewed by ozaki-r@ and yamaguchi@
|
| 1.6.2.1 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.1 | 02-Feb-2017 |
ozaki-r | branches: 1.1.2; 1.1.4; 1.1.8; 1.1.18;
Defer some pr_input to workqueue
pr_input is currently called in softint. Some pr_input such as ICMP, ICMPv6
and CARP can add/delete/update IP addresses and routing table entries. For
example, icmp6_redirect_input updates an a routing table entry and
nd6_ra_input may delete an IP address.
Basically such operations shouldn't be done in softint. That aside, we have
a reason to avoid the situation; psz/psref waits cannot be used in softint,
however they are required to work in such pr_input in the MP-safe world.
The change implements the workqueue pr_input framework called wqinput which
provides a means to defer pr_input of a protocol to workqueue easily.
Currently icmp_input, icmp6_input, carp_proto_input and carp6_proto_input
are deferred to workqueue by the framework.
Proposed and discussed on tech-kern and tech-net
|
| 1.1.18.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.1.18.1 | 02-Feb-2017 |
jdolecek | file wqinput.h was added on branch tls-maxphys on 2017-12-03 11:39:04 +0000
|
| 1.1.8.2 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.1.8.1 | 02-Feb-2017 |
bouyer | file wqinput.h was added on branch bouyer-socketcan on 2017-04-21 16:54:06 +0000
|
| 1.1.4.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.1.4.1 | 02-Feb-2017 |
pgoyette | file wqinput.h was added on branch pgoyette-localcount on 2017-03-20 06:57:51 +0000
|
| 1.1.2.2 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.1.2.1 | 02-Feb-2017 |
skrll | file wqinput.h was added on branch nick-nhusb on 2017-02-05 13:40:59 +0000
|
