| History log of /src/sys/dev/scsipi |
| Revision | Date | Author | Comments |
| 1.6 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.5 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.4 | 31-May-2001 |
tron | branches: 1.4.22; 1.4.30; 1.4.32;
Install "dev/scsipi/scsipi_cd.h" again because it is e.g. used by the
CD player include in version 2.x of the "kdemultimedia" package.
|
| 1.3 | 11-Apr-2001 |
jdolecek | Only install headers which are actually used by our userland. This
saves about 2.2MB under /usr/include/dev/. Discussed on tech-kern@
recently.
I HOPE to get the list right. The headers I left in are ones
used for MI tools and those whose usage I discovered by grep over tree sources.
Feel free to put needed includes back in if you encounter anything which
should not be removed from lists.
|
| 1.2 | 20-Feb-2000 |
mjacob | branches: 1.2.6;
add ses.h to list of includes
|
| 1.1 | 12-Jun-1998 |
cgd | branches: 1.1.14;
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.
|
| 1.1.14.2 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.1.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.2.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.4.32.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.4.30.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.4.22.1 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.4 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.3 | 13-Feb-1998 |
enami | branches: 1.3.26;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.2 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.1 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.3.26.1 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.30 | 03-Feb-2019 |
mrg | - add or adjust /* FALLTHROUGH */ where appropriate
- add __unreachable() after functions that can return but won't in
this case, and thus can't be marked __dead easily
|
| 1.29 | 18-Oct-2014 |
snj | branches: 1.29.20;
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.28 | 28-Apr-2008 |
martin | branches: 1.28.44;
Remove clause 3 and 4 from TNF licenses
|
| 1.27 | 05-Apr-2008 |
cegger | branches: 1.27.2; 1.27.4;
use aprint_*_dev and device_xname
|
| 1.26 | 11-Dec-2005 |
christos | branches: 1.26.70;
merge ktrace-lwp.
|
| 1.25 | 29-May-2005 |
christos | - Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.24 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.23 | 17-Sep-2004 |
mycroft | branches: 1.23.4; 1.23.6;
Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.22 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.21 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.20 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.19 | 10-Mar-2004 |
bouyer | branches: 1.19.2;
Decrease xs_retries before retrying aborted commands, and report EIO if it
reaches 0. Avoids looping on aborded command in some special cases.
|
| 1.18 | 15-Nov-2001 |
lukem | branches: 1.18.16;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.17 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.16 | 14-May-2001 |
bouyer | branches: 1.16.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.15 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.14 | 30-Sep-1999 |
thorpej | branches: 1.14.2; 1.14.14;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.13 | 28-Aug-1999 |
thorpej | If scsipi_make_xs() fails, make sure to flag the buffer as error'd,
if provided, and biodone() it. Callers of scsipi_command() expect
this to be done.
|
| 1.12 | 25-Jun-1999 |
thorpej | Go to splbio() when calling scsipi_free_xs(), just like the SCSI counterpart.
Fixes PR #7853, Paul Dokas <dokas@cs.umn.edu>.
|
| 1.11 | 17-Nov-1998 |
bouyer | branches: 1.11.4; 1.11.6; 1.11.8;
In atapi_wdc.c, issue a 'REQUEST SENSE' command when appropriate.
Return XS_SENSE when the full sense info has been retrieved, or
XS_SHORTSENSE if only the sense key was available (from the error register)
Make atapi_interpret_sense() deal with this, and call scsipi_interpret_sense()
for XS_SENSE. (XXX sd_interpret_sense() and the ioctl code needs to be made
aware of XS_SHORTSENSE too ! sense hanlding for these is now less broken for
devices that support 'REQUEST SENSE')
All the ATAPI devices I have access to seems to honnor the SENSE_REQUEST
command, but I suspect some ATAPI devices will not (althouh it's mandatory).
The code should be able to deal with this, but is untested ...
|
| 1.10 | 12-Oct-1998 |
bouyer | Merge bouyer-ide
|
| 1.9 | 15-Aug-1998 |
mycroft | branches: 1.9.2;
Assign my copyrights to TNF.
|
| 1.8 | 15-Jul-1998 |
mjacob | part of fix for kern/3835: use of enumerated returns from target sense handlers
|
| 1.7 | 11-Jul-1998 |
mjacob | Clarify that Sense Key 1 is a "RECOVERED ERROR"- not just an error.
|
| 1.6 | 01-Jul-1998 |
mjacob | Use new Sense Key defines (for clarity)
|
| 1.5 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.4 | 30-Dec-1997 |
is | EACCESS->EROFS, for the write on readonly medium error.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file atapi_base.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: some spaces -> tabs, indenting
consistency in prototypes.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file atapi_base.c was added on branch marc-pcmcia on 1997-08-27 23:32:53 +0000
|
| 1.9.2.2 | 02-Oct-1998 |
bouyer | When printing the error message, also print the channel with the contrller
name and drive number.
|
| 1.9.2.1 | 15-Aug-1998 |
bouyer | file atapi_base.c was added on branch bouyer-ide on 1998-10-02 18:12:49 +0000
|
| 1.11.8.1 | 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.
|
| 1.11.6.1 | 01-Jul-1999 |
thorpej | Sync w/ -current.
|
| 1.11.4.1 | 25-Jun-1999 |
perry | pullup 1.11->1.12 (thorpej): Go to splbio() when calling scsipi_free_xs()
|
| 1.14.14.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.14.14.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.14.14.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.14.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.14.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.16.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.18.16.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.18.16.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.18.16.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.18.16.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.18.16.3 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.18.16.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.18.16.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.19.2.1 | 11-Sep-2004 |
he | Pull up revisions 1.21-1.22 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.23.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.23.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.26.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.27.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.27.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.28.44.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.29.20.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.11 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.10 | 14-May-2001 |
bouyer | branches: 1.10.22;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.9 | 13-Jul-1998 |
thorpej | branches: 1.9.26;
Those tabs sure are excitable!!
|
| 1.8 | 13-Jul-1998 |
hpeyerl | Add support for ATA CD changer devices like the NEC CDR-251.
Evidence of managerial coding removed by Victor T. Cleaner (thorpej)
|
| 1.7 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.6 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.5 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.4 | 03-Oct-1997 |
enami | Don't define AUDIO_PLAY twice (accidently add in last commit).
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file atapi_cd.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: some spaces -> tabs, indenting
consistency in prototypes.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file atapi_cd.h was added on branch marc-pcmcia on 1997-08-27 23:32:54 +0000
|
| 1.9.26.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.10.22.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.6 | 12-Oct-2003 |
bouyer | Remove this file; ATAPI_READ_FORMAT_CAPACITIES is the same as
READ_FORMAT_CAPACITIES; atapi_flex_geometry_page should probably be the same as
scsi_disk's page_flex_geometry.
|
| 1.5 | 17-Sep-2003 |
mycroft | Remove PQUIRK_BYTE5_ZERO.
|
| 1.4 | 14-May-2001 |
bouyer | branches: 1.4.22;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.3 | 13-Feb-1998 |
enami | branches: 1.3.26;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.2 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.1 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.3.26.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.4.22.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.6 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.5 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.4 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.3 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.2 | 01-Dec-2001 |
bouyer | branches: 1.2.16; 1.2.24; 1.2.26;
Various quirks for the ATAPI OnStream DI-30, mostly from the FreeBSD driver.
Many thanks to Chris Pinnock for giving me remote access to his hardware.
|
| 1.1 | 18-Jun-2001 |
bouyer | branches: 1.1.2; 1.1.4;
Snapshot of ATAPI tapes support. Known to be able to read/write to tape with:
st0 at atapibus0 drive 1: <Seagate STT8000A, , 5.51> type 1 sequential
removable
Major changes may still happen in order to properly support other ATAPI
tape drives.
|
| 1.1.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.1.2.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.1.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.1.2.1 | 18-Jun-2001 |
nathanw | file atapi_tape.h was added on branch nathanw_sa on 2001-06-21 20:05:40 +0000
|
| 1.2.26.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.2.24.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.2.16.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.2.16.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.141 | 05-Oct-2021 |
rin | PR kern/56403
Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ:
(1) Change ata_xfer_ops:c_poll from void to int function. When it returns
ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again.
(2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is
achieved.
A similar change has been made for mvsata(4) (see mvsata_bio_poll()),
and no functional changes for other devices.
This is how the drivers worked before jdolecek-ncq branch was merged.
Note that this changes are less likely to cause infinite recursion:
(1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE
in its error handling paths via wdc_ata_bio_done().
(2) Return value from c_start (= wdc_ata_bio_start()) is checked in
ata_xfer_start().
Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines
terminate the recursion for wdc(4). The situation is similar for mvsata(4).
Still, there is a possibility where ata_xfer_start() takes long time to
finish a normal operation. This can result in a delayed response for lower
priority interrupts. But, I've never observed such a situation, even when
heavy thrashing takes place for swap partition in wd(4).
"Go ahead" by jdolecek@.
|
| 1.140 | 07-Aug-2021 |
thorpej | Merge thorpej-cfargs2.
|
| 1.139 | 24-Apr-2021 |
thorpej | branches: 1.139.8;
Merge thorpej-cfargs branch:
Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.
Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)
Remove unnecessary or redundant interface attributes where they're not
needed.
There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)
...and a sentinel value CFARG_EOL.
Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.
Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).
|
| 1.138 | 13-Apr-2020 |
jdolecek | branches: 1.138.4;
fix use-after-free for ata xfer on bio submission found by KASAN
driver ata_bio hooks read parts of the xfer after ata_exec_xfer()
call in order to determine return value, change so that the hook
doesn't return any value - callers do not care already,
as all I/O requests are asynchronous
this problem was uncovered by recent change for wd(4) to not hold
wd mutex during ata_bio call, the interrupt for the xfer might
thus actually fire immediately
adjust also ata_exec_command driver hooks similarily - remove all
completion and waiting logic from drivers, upper layer ata code
using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself
PR kern/55169 by Nick Hudson
|
| 1.137 | 04-Apr-2020 |
jdolecek | branches: 1.137.2;
fix deadlock in wdcwait() when xfer timeout happens while the atabus
thread sleeps in wdcwait() - check current lwp rather than relying
on global ATACH_TH_RUN channel flag
should fix the hang part of the problem reported in
http://mail-index.netbsd.org/netbsd-users/2020/03/12/msg024249.html
thanks to Paul Ripke for providing extensive debugging info
|
| 1.136 | 19-Feb-2020 |
riastradh | C99 initializers for scsipi_bustype. No functional change intended.
|
| 1.135 | 12-Feb-2020 |
chs | the number of possible ATAPI devices on an ATA bus is not always 2,
it is however many devices the underlying ATA bus can have (eg. 1 for SATA),
so initialize the scsipi chan_ntargets from the ATA ch_ndrives.
this fixes a memory read overrun detected by KASAN.
discussed with mlelstv@ and jdolecek@
|
| 1.134 | 10-Nov-2019 |
chs | branches: 1.134.2;
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.
|
| 1.133 | 12-Nov-2018 |
jdolecek | branches: 1.133.4;
pass correct status + error to *_atapi_phase_complete(), so that
the function is actually able to recognize when there was an error;
tested via reading a cd0 device in QEMU with ejected cdrom
bug was introduced with jdolecek-ncq branch
fixes PR kern/53724 by Andreas Gustafsson
|
| 1.132 | 12-Nov-2018 |
jdolecek | hold channel lock during whole ata_dmaerr()/ata_downgrade_mode() -
according to code inspection this is safe, none of the set_modes
hooks execute anything which would be taking the lock
adresses PR kern/53714 by Andreas Gustafsson
|
| 1.131 | 22-Oct-2018 |
jdolecek | Merge jdolecek-ncqfixes branch
- ata_xfer's are dynamicall allocated as needed using a pool, no longer
limited to number of possible openings supported by controller; dump
and recovery paths use dedicated pre-allocated storage
- moved callouts and condvars from ata_xfer to queue or channel, so that
ata_xfer does not need special initialization
- slot allocation now done when xfer is being activated, uncoupled
from memory allocation; active slots are no longer tracked by controller
code
- channel and drive reset is done always via the atabus thread, and
now executes with channel locked the whole time
- NCQ recovery moved to shared function, and run via the thread also
- added some workarounds for buggy error recovery AHCI emulation in QEMU
and Parallels
designed to primarily fix kern/52614, but might also help with kern/47041
and kern/53183
|
| 1.130 | 01-Sep-2018 |
mlelstv | Make wdcdebug_atapi_mask configurable like the other debug variables.
|
| 1.129 | 17-Oct-2017 |
jdolecek | branches: 1.129.2; 1.129.4; 1.129.6;
reintroduce ATACH_IRQ_WAIT flag for attachments using wdcintr(), only
process the interrupt when the flag is set - this fixes spurious interrupt
during post-reset drive setup in wdc_ata_bio_start(), and wdc_atapi_start()
while those functions set WDCTL_IDS, this seems to be ignored by certain
(maybe all) PCI-IDE controllers; usually the implicit KERNEL_LOCK() would
prevent the interrupt anyway, but not when the start routine is started
from the atabus thread, which doesn't take it
fixes 'panic: wdc_ata_bio_intr: bad state' reported on current-users
by Chavdar Ivanov
|
| 1.128 | 10-Oct-2017 |
jdolecek | revert the logic in wdc_atapi_intr() for wdc_wait_for_unbusy() to what it
was before NCQ merge; it got broken during the efford to remove ch_status
and ch_error on the branch
fixes atapi timeouts in vbox and with real harware reported separately
by Abhinav Upadhyay, Pault Goyette, Chavdar Ivanov, and Rares
Aioanei; with a bit of luck it could also fix PR kern/52605 and/or PR
kern/52606 by Martin Husemann
|
| 1.127 | 08-Oct-2017 |
christos | don't initialize tfd twice.
|
| 1.126 | 08-Oct-2017 |
christos | initialize tfd
|
| 1.125 | 08-Oct-2017 |
joerg | Initialize tfd even when not waiting for DSC.
|
| 1.124 | 07-Oct-2017 |
jdolecek | Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch
ATA subsystem was changed to support several outstanding commands, and use
NCQ xfers if supported by both the controller and the disk, including NCQ
error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers
if supported. Added FUA support.
Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced
by condvars, and switched most code from spl* to mutexes (separate
wd(4) and ata channel lock).
Introduced new option WD_CHAOS_MONKEY to facilitate testing of error
handling, fixed several uncovered issues. Also fixed several problems
with kernel dump to wd(4) disk.
Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64,
with and without port multiplier, both disk and ATAPI devices; other
drivers and archs mechanically adjusted and compile-tested. NCQ is
supported for ahcisata(4) and siisata(4) for any controller, for
mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in
mvsata(4).
Thanks to Matt Thomas for initial ATA infrastructure patch, and
Jonathan A.Kollasch for siisata(4) NCQ changes and general testing.
Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041,
PR kern/51979 (kernel dump)
|
| 1.123 | 20-Nov-2016 |
mlelstv | branches: 1.123.4;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.122 | 02-May-2016 |
christos | branches: 1.122.2;
move scsipi_strvis -> libkern:strnvisx()
change the prototype to match userland
fix sizes of strings passed to it
|
| 1.121 | 20-Mar-2016 |
tsutsui | Pass a correct residual size to wdcbit_bucket() that drains extra bytes.
Seems broken (but almost harmless?) since rev 1.101 committed ~10 years ago:
http://mail-index.netbsd.org/source-changes/2006/09/07/0043.html
|
| 1.120 | 20-Mar-2016 |
tsutsui | Remove extra spaces.
|
| 1.119 | 31-Jul-2012 |
bouyer | branches: 1.119.2; 1.119.16;
Move a check for the drive type to the right place. This bug would cause
ATA drives to not be detected when the PATA (or emulated PATA) bus is shared
with an ATAPI device.
Should fix the problem reported by various peoples.
Thanks to Onno van der Linden and vanoven@free.fr for pointing out the
code in error.
|
| 1.118 | 31-Jul-2012 |
bouyer | Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc
cleanup and SATA PMP support), now that I'm back to fix the fallouts.
|
| 1.117 | 26-Jul-2012 |
jakllsch | Revert, with intention of restoring in a less invasive way, the SATA Port
Multiplier code.
ok christos@
|
| 1.116 | 24-Jul-2012 |
jakllsch | Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and
Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump.
First step in reverting regressions to ata(4) subsystem during the addition of
port multiplier support.
|
| 1.115 | 15-Jul-2012 |
dsl | Some namespace protection (and add greppablity).
Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_.
Don't use an enum for drive_type - you don't know how big it will be.
Move driver_type to avoid implicit structure padding (esp on arm).
This change is purely lexical and mechanical.
Update to 6.99.9 - this wasn't done when the SATA PMP changes
were made - I'm sure they warranted a bump.
|
| 1.114 | 02-Jul-2012 |
bouyer | Add sata Port MultiPlier (PMP) support to the ata bus layer,
as described in
http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html
PMP support in integrated to the atabus layer.
struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive
(renamed to ch_ndrives) closely reflects the size of the ch_drive[] array.
Add helper functions atabus_alloc_drives() and atabus_free_drives()
to manage ch_drive[]/ch_ndrives.
Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify
how much drive they really support (master/slave or single).
ata_reset_drive() callback gains a uint32_t *sigp argument which,
when not NULL, will contain the signature of the device being reset.
While there, some cosmetic changes:
- added a drive_type enum to ata_drive_datas, and stop encoding the
probed drive type in drive_flags (we were out of drive flags anyway).
- rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this
really is
- remove ata_channel->ata_drives, it's redundant with the pointer in
ata_drive_datas
- factor out the interpretation of SATA signatures in sata_interpet_sig()
propagate these changes to the ATA HBA drivers, and add support for PMP
to ahcisata(4) and siisata(4).
Thanks to:
- Protocase (http://www.protocase.com/) which provided a system
with lots of controllers, SATA PMP and drive slots
- Conservation Genomics Laboratory, Department of Biology, New Mexico State
University for hosting the above system
- Brook Milligan, who set up remote access and has been very responsive
when SATA cable move was needed
|
| 1.113 | 20-Apr-2012 |
bouyer | Add a bustype_async_event_xfer_mode() callback to scsipi_bustype (which can
be NULL), so that transport-specific details of transfer mode setting/printing
can be handled more easily.
Move scsipi_async_event_xfer_mode() and scsipi_print_xfer_mode() to
scsi_base.c and split in parallel scsi and FC/SAS parts.
size of struct scsipi_bustype has changed, welcome to 6.99.5
|
| 1.112 | 24-May-2011 |
joerg | branches: 1.112.4; 1.112.8;
Fix obvious condition snafu
|
| 1.111 | 28-Mar-2010 |
snj | branches: 1.111.2;
Spell "enough" properly.
|
| 1.110 | 19-Oct-2009 |
bouyer | branches: 1.110.2; 1.110.4;
Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !
|
| 1.109 | 16-Nov-2008 |
bouyer | cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so
we can't use it here. Rssurect ATACH_TH_RUN, backing out
src/sys/dev/ata/ata.c 1.101
src/sys/dev/ata/ata_wdc.c 1.90
src/sys/dev/ata/atavar.h 1.77
src/sys/dev/ic/wdc.c 1.255
src/sys/dev/scsipi/atapi_wdc.c 1.108
Should fix kern/39927 and kern/39725.
|
| 1.108 | 02-Oct-2008 |
bouyer | branches: 1.108.2; 1.108.4;
Kill ATACH_TH_RUN and use cpu_intr_p() instead.
|
| 1.107 | 24-Mar-2008 |
cube | branches: 1.107.4; 1.107.6; 1.107.10;
Split device_t and softc for atapibus(4).
|
| 1.106 | 18-Mar-2008 |
cube | Split device_t and softc for ATA devices, as well as wd(4). Other
cosmetic changes where appropriate.
|
| 1.105 | 19-Oct-2007 |
ad | branches: 1.105.12; 1.105.16;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
|
| 1.104 | 23-Jan-2007 |
reinoud | branches: 1.104.6; 1.104.18; 1.104.20; 1.104.24;
Fix for atapi devices that have 32 bit databusses selected. The devices
won't error out but will return garbage data. Implemented by generalising
the odd-check for PIO selection to check for 4 byte alignment in offset and
size.
Note that this check could be relaxed by checking if the ATAPI/ide bus
width is 16 for the device but its unknown to me if that information is
available here.
|
| 1.103 | 01-Oct-2006 |
reinoud | branches: 1.103.2;
In gcc4 structures are differently aligned compared to gcc3. Lots of
structures can thus be misaligned for DMA when declared on stack.
The current atapi code accepts the misaligned dma's only screw's them up.
This small patch disables DMA for the rare transactions that are
misaligned. Note that all bulk-transfers are aligned due to page sizes.
See PR kern/34689
|
| 1.102 | 30-Sep-2006 |
itohy | A little effort against kernel bloat....
Exclude ATA DMA support if no ATA DMA capable drivers are compiled in.
|
| 1.101 | 07-Sep-2006 |
itohy | branches: 1.101.2; 1.101.4;
Add PIOBM (busmastering transfer using ATA PIO mode) support.
The PIOBM is used by only one driver (will be added later,
stay tuned) and intruduce an attribute "ata_piobm" so that
it will be conditionally compiled in.
The "ata_dma" (busmastering transfer using ATA DMA mode) and
"ata_udma" (busmastering transfer using ATA Ultra DMA mode)
attributes are also added for consistency, but unused for now.
|
| 1.100 | 27-Jun-2006 |
tron | Make this build with GCC 4.x.
|
| 1.99 | 21-May-2006 |
christos | branches: 1.99.2;
void casts to functions whose return values are ignored.
|
| 1.98 | 20-Feb-2006 |
thorpej | branches: 1.98.2; 1.98.6; 1.98.8;
Use device_is_active() rather than testing dv_flags for DVF_ACTIVE
directly.
|
| 1.97 | 11-Dec-2005 |
christos | branches: 1.97.2; 1.97.4; 1.97.6;
merge ktrace-lwp.
|
| 1.96 | 07-Jul-2005 |
drochner | tighten the autoconf constraints by passing the atapi/ata/ata_hl attribute
to config_found()
(after some cleanup we might be able to kill SCSIPI_BUSTYPE_*)
|
| 1.95 | 06-Jul-2005 |
thorpej | branches: 1.95.2;
- When starting an ATA or ATAPI transfer, handle the case where (*dma_init)()
returns EINVAL, indicating that DMA cannot be done for this transfer.
Fall back to PIO in this case.
- Add a geodeide_dma_init() routine that checks to make sure that transfers
start on a 16 byte boundary, returning EINVAL if not. Works around a chip
bug that causes a hard system hang.
Problem reported and patch tested by Erik Fair.
|
| 1.94 | 07-Jun-2005 |
peter | Fix typos.
|
| 1.93 | 29-May-2005 |
christos | - Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.92 | 27-Feb-2005 |
perry | branches: 1.92.2;
nuke trailing whitespace
|
| 1.91 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.90 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.89 | 21-Aug-2004 |
thorpej | branches: 1.89.4; 1.89.6;
atastart() (called only at splbio(), and from interrupts) can change
drive_flags, to make sure all drive_flags manipulations are done at
splbio().
|
| 1.88 | 20-Aug-2004 |
thorpej | Move most of wdc_softc into a new atac_softc structure that contains
info common to all types of ATA controllers.
|
| 1.87 | 14-Aug-2004 |
thorpej | - Split the register handles out of struct wdc_channel into a separate
wdc_regs structure, and array of which (indexed per channel) is pointed
to by struct wdc_softc.
- Move the resulting wdc_channel structure to atavar.h and rename it to
ata_channel. Rename the corresponding flags.
- Add a "ch_ndrive" member to struct ata_channel, which indicates the
maximum number of drives that can be present on the channel. For now,
this is always 2. Add an ATA_MAXDRIVES constant that places an upper
limit on this value, also currently 2.
|
| 1.86 | 13-Aug-2004 |
thorpej | WDCDEBUG -> ATADEBUG.
|
| 1.85 | 13-Aug-2004 |
thorpej | Don't bother with bits that tell of the presence of optional callbacks;
just check the function pointers for NULL.
|
| 1.84 | 13-Aug-2004 |
thorpej | Move wdcstart() to ata.c and rename it to atastart().
|
| 1.83 | 13-Aug-2004 |
thorpej | Move wdc_exec_xfer() to ata.c and rename it ata_exec_xfer().
|
| 1.82 | 12-Aug-2004 |
thorpej | Move wdc_addref() and wdc_delref() to ata.c and rename them to
ata_*.
|
| 1.81 | 12-Aug-2004 |
thorpej | - Move wdc_xfer_pool, wdc_get_xfer(), wdc_free_xfer() to ata.c, and rename
to ata_*.
- Use a static initializer for the ata_xfer_pool.
|
| 1.80 | 12-Aug-2004 |
thorpej | Move wdc_probe_caps() to ata.c and rename it ata_probe_caps().
|
| 1.79 | 12-Aug-2004 |
thorpej | Rename some constants:
WDC_COMPLETE -> ATACMD_COMPLETE
WDC_QUEUED -> ATACMD_QUEUED
WDC_TRY_AGAIN -> ATACMD_TRY_AGAIN
|
| 1.78 | 12-Aug-2004 |
thorpej | Rename "struct wdc_command" to "struct ata_command".
|
| 1.77 | 11-Aug-2004 |
mycroft | Make datain_pio and dataout_pio function pointers in wdc_softc, which can
be overridden by the backend if desired. Add experimental code to wdc_pcmcia
to use this in memory-mapped mode, disabled by default.
|
| 1.76 | 11-Aug-2004 |
mycroft | Change the behavior when a transfer mode setting is rejected. For PIO, fall
back to mode 3. For DMA, fall back to mode 0. These are the minimums for
ATAPI. For UDMA, fail.
|
| 1.75 | 11-Aug-2004 |
mycroft | Add two helper functions -- wdc_datain_pio() and wdc_dataout_pio() -- which
encapsulate the logic for the various methods of transferring data. Use these
throughout.
|
| 1.74 | 04-Aug-2004 |
bouyer | Change wdc_kill_pending() to take a struct ata_drive_datas * as argument,
and kill only pending requests for this drive.
Implement a DRIVE_WAITDRAIN flag, which will cause the active command to
be killed once complete.
Other minor fixes.
Now it's possible to detach a ATA or ATAPI device from ioctl even when
a dd on the raw char partition is running.
|
| 1.73 | 04-Aug-2004 |
bouyer | Add
struct ata_xfer *active_xfer
to ata_queue. Now the active xfer isn't the head of the queue any more,
this makes a few things easier (this will also help for tagged queuing
support).
Remove the WDCF_ACTIVE flag, test active_xfer != NULL instead.
clean up wdc_free_xfer() and kill_xfer().
Clean up wdc_reset_channel(), and make it issue a ATAPI_SOFT_RESET if the
active command is ATAPI.
In wdc_atapi_get_params(), use AT_WAIT | AT_POLL for ATAPI_SOFT_RESET,
so that we'll use tsleep() instead of delay().
In wdc_atapi_start(), call wdc_dmawait() at the right place.
|
| 1.72 | 01-Aug-2004 |
bouyer | Implement an atabus control device, and define some ATA bus control
IOCTLS. Implement ATABUSIORESET, which will reset the given ATA bus.
|
| 1.71 | 31-Jul-2004 |
bouyer | Implement asynchronous channel reset.
Use this to reset the channel before doing a dump, instead of the hack in
wdc_exec_xfer() based on C_POLL. This hack was causing problems on
controllers with a shared queue, because we now can have C_POLL set during
concurent channels probes (problem found and analysed on sparc64 by
Martin Husemann).
This should even make core dumps marginally more reliable on ATA drives.
|
| 1.70 | 08-May-2004 |
bouyer | Add a delay(10) after re-enabling interrupts in the control register.
Some controllers/drives (e.g. SataLink 3114 with WD Raptor) require
it. Should fix kern/23808 by Chris Gilbert, patch suplied by Chris Gilbert
on tech-kern, extended to all places enabling interrupts by me.
|
| 1.69 | 03-Feb-2004 |
bouyer | branches: 1.69.2;
Clear drive_flags when wdc_atapi_get_params() fails. Avoid a panic later
when probing a slave-only bus in some cases.
|
| 1.68 | 03-Jan-2004 |
thorpej | More structure member namespace cleanup: thread -> ch_thread
|
| 1.67 | 03-Jan-2004 |
thorpej | More wdc_channel structure member namespace cleanup:
- channel -> ch_channel
- wdc -> ch_wdc
|
| 1.66 | 03-Jan-2004 |
thorpej | Rename "struct channel_softc" to "struct wdc_channel".
|
| 1.65 | 01-Jan-2004 |
thorpej | Tidy this file up.
|
| 1.64 | 01-Jan-2004 |
thorpej | Prepend "wdc_" to wait_for_drq, wait_for_unbusy, and wait_for_ready.
|
| 1.63 | 01-Jan-2004 |
thorpej | Rename:
- wdc_xfer to ata_xfer
- channel_queue to ata_queue
and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA
controllers. Clean up the member names of these structures while at it.
|
| 1.62 | 27-Nov-2003 |
fvdl | There are some cards that map the ATA control and IDE DMA registers
in a different fashion. Individually, they have the same functionality,
but their layout is different. An example of such a chipset is
the Promise 203xx.
To be able to deal with this, transform the cmd and dma bus_space handles
into an array of handles, each seperately created with bus_space_subregion.
The code generated by using the extra indirection shouldn't change much,
since the extra indirection is negated by having the offset calculation
already done in bus_space_subregion. E.g.
bus_space_write_4(tag, handle, offset, value)
becomes
bus_space_write_4(tag, handles[offset], 0, value)
Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386)
and on cmdide (sparc64).
|
| 1.61 | 29-Oct-2003 |
bouyer | freese->freeze, as pointed out by Frederick Bruck.
|
| 1.60 | 15-Oct-2003 |
bouyer | Clear chp->ch_drive[i].drive_flags if we didn't find a driver for the
drive.
|
| 1.59 | 08-Oct-2003 |
bouyer | Make the ATA mid-layer appears as atabus, as proposed in
http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html
This adds a device (atabus) between IDE controllers and wd or atapibus, to
have each ATA channel show up in the device tree. Later there will be atabus
devices in /dev, so that we can do IOCTL on them.
Each atabus has its own kernel thread, to handle operations that needs polling,
e.g. reset and others.
Device probing on each bus it defered to the atabus thread creation.
This allows to do the reset and basic device probes in parallel, which reduce
boot time on systems with several pciide controllers.
|
| 1.58 | 05-Oct-2003 |
bouyer | Remove references to University of California from my copyright notices.
|
| 1.57 | 19-Sep-2003 |
mycroft | 1) Use config_interrupts() to attach IDE and ATAPI drives. This eliminates
most polling.
2) Clean up some goofiness in pciide -- get rid of the whole "candisable" path
(it's gratuitous) and simplify the code by calling pciide_map_compat_intr(),
*_set_modes() and wdc_print_modes() from central locations.
3) Add a register writability and register ghost test to eliminate phantom
drives more quickly.
|
| 1.56 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.55 | 27-Sep-2002 |
provos | branches: 1.55.6;
remove trailing \n in panic(). approved perry.
|
| 1.54 | 26-Jul-2002 |
wiz | enouth -> enough.
|
| 1.53 | 22-Jul-2002 |
bouyer | Do PIO transfers for len < 4 bytes. From Matthias Drochner in private mail.
|
| 1.52 | 18-May-2002 |
bouyer | branches: 1.52.2;
If the data len if not a multiple of 2 bytes, force PIO mode. DMA can't
deal with such xfers, and can wedge the system with some controllers.
It's a bug to request such xfers for ATAPI, but as the request may come
from userland we have to protect against it.
|
| 1.51 | 23-Apr-2002 |
bouyer | branches: 1.51.2;
More copyright fixes, pointed out by Thomas. Thanks !
|
| 1.50 | 05-Apr-2002 |
bouyer | Implement mstohz() as discussed on tech-kern, and use it in SCSI drivers
to convert xs->timeout to callout() parameter.
|
| 1.49 | 31-Mar-2002 |
martin | Don't clear the adapt structure of the attaching channel - this nukes
all function pointers passed in from the adapter driver.
This partly fixes PR 13480, i.e. the FREECOM CD driver works now in pcmcia
adapters.
The remaining issue (timing problems with slow cards and cardbus bridges)
is probably the cause of several other PRs too.
|
| 1.48 | 27-Jan-2002 |
bouyer | IF we change the ATA modes, also reset the device, so that we can retry
the command with the new mode.
|
| 1.47 | 03-Dec-2001 |
bouyer | Update my copyrigth.
|
| 1.46 | 02-Dec-2001 |
bouyer | Clean up attach of wd/atapibus:
kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead
of ata_atapi_attach as attach arch. Create a ata_device, compatible with
scsipi_channel, to attach wd.
|
| 1.45 | 01-Dec-2001 |
bouyer | For ATAPI tape drives, poll for DSC (using a callout), to get the real
status of the command, and make sure the drive is ready for the next one.
|
| 1.44 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.43 | 27-Jun-2001 |
bouyer | branches: 1.43.2;
Some ATAPI CD occasionally reports "ATA illegal indication length" for
a request sense command, but the sense is properly transfered anyway.
Ignore the error for this case.
|
| 1.42 | 13-Jun-2001 |
bjh21 | Add explicit support for IDE and SCSI adaptors which don't support interrupts.
On such adaptors, all transfers are done in polling mode.
OK'ed by Manuel on tech-kern.
|
| 1.41 | 15-May-2001 |
lukem | delint: can't have an empty case item at the end of a switch; need a ; at least
|
| 1.40 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.39 | 22-Jan-2001 |
mycroft | branches: 1.39.2;
Force certain commands (i.e. DVD authentication) to PIO mode. They are
explicitly not supported in DMA mode, and generally return an `illegal field
in CDB' error.
This should have been done ages ago.
|
| 1.38 | 03-Aug-2000 |
bouyer | If the SET_FEATURE for piomode is rejected, fallback to mode 0 instead of
reporting an error.
This fixes the "piomode error (0x4)" problem with some ATAPI ZIP drives
reported on port-i386.
|
| 1.37 | 28-Jun-2000 |
mrg | remove include of <vm/vm.h>
|
| 1.36 | 12-Jun-2000 |
bouyer | branches: 1.36.2;
Add a callback (*irqack), for controllers that need special action to ack
the interrupt once it has been ack'd on the drive.
|
| 1.35 | 15-May-2000 |
bouyer | branches: 1.35.2;
Sync copyrigth notice
|
| 1.34 | 02-Apr-2000 |
augustss | Let the device thet gets an atapibus attached specify how to kill pending
transfers by giving a function pointer.
The old method always called wdc specific code.
|
| 1.33 | 01-Apr-2000 |
bouyer | - DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op
if an IRQ was not detected, unless the force flag was given. Use this to
detect if the IRQ was for us (closer to shared IRQ for controllers which
don't have their own IRQ handler in pciide.c) and to poll for DMA xfer.
Also makes the timeout recovery code simpler.
- ATAPI cleanup: don't call controller-specific functions from atapiconf.c
(wdc_*), so that it's possible to attach an atapibus to something else
than a wdc/pciide (Hi Lennart :).
Overload struct scsi_adapter with struct atapi_adapter, defined
as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points
to an scsi_adapter, atapi code casts it to atapi_adapter if needed.
Move atapi_softc to atapiconf.h so that it can be used by the underlying
controller code (e.g. atapi_wdc.c).
Add an atapi-specific callback *atapi_probedev(), which probe a drive
in a controller-specific way, allocate the sc_link and fills in the
ataparams if needed. It then calls atapi_probedev() (from atapiconf.c)
to do the generic initialisations and attach the device.
- While I'm there merge and centralise the state definitions in atavar.h.
It should now be possible to use a common ata/atapi routine to set the
drive's modes (will do later).
|
| 1.32 | 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.
|
| 1.31 | 20-Mar-2000 |
enami | Cancel requested command if controller isn't active.
|
| 1.30 | 17-Jan-2000 |
bouyer | Improve the downgrade logic:
- If UDMA 2 is failing try UDMA 1 first, it helps in some cases
- downgrade if we get an error in the first 4000 xfers, or if we get
4 errors in 4000 xfers if the first 4000 went without troubles.
While I'm there commit a local change I have since some time to get my CD
probed: issue a "blanck" IDENTIFY before the one used to detect slave ghosts,
with my drive the first IDENTIFY following a controller reset fails with an
aborted command ...
|
| 1.29 | 04-Nov-1999 |
bouyer | Pass xfer->c_bcount instead of sc_xfer->datalen to the device, so that
it gets the rigth len for a request sense (from OpenBSD).
|
| 1.28 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.27 | 30-Sep-1999 |
thorpej | branches: 1.27.2; 1.27.4; 1.27.6;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.26 | 23-Sep-1999 |
enami | Allow to detach wdc, atapibus, wd and cd.
|
| 1.25 | 09-Aug-1999 |
bouyer | Add a missing 'return' in previous commit.
Don't use C_NEEDDONE anymore (it's safe to always call scsipi_done()).
|
| 1.24 | 06-Aug-1999 |
bouyer | - Add some debug printf (WDCPROBE) in _wdcreset_wait(), I've needed these
2 times in the past
- Set up timeout per xfer instead of per interrupt. This helps with
PIO transfer (we would call timeout()/untimout() several times for a
transfer).
- If we missed an interrupt for a PIO transfer, reset and restart it
immedialy, don't try to recover and continue. If we missed an interrupt we
may have lost a read/write cycle on the IDE bus. If this happens 1) we
corrupt data and 2) we enter an interrupt loop at the end of the xfer, as
the drive has some more data to read/write, but the host thinks the xfer is
done.
This last change fix the (or at last some of the) 'lookup after lost interrupt'
some peoples have been experiencing.
|
| 1.23 | 05-May-1999 |
perry | Add patch from Manuel Bouyer -- his mail message said:
By code inspection I found 2 bugs in the ATAPI code, one may be the cause
of your problem: A counter is not reset when issuing the request_sense
command. This is c_skip, the counter used to track the offset in the data
buffer when a data tranfer needs multiple phases. The effect of this is that
the sense data transfered may be stored outside the sense buffer (sense buffer
+ some, potentially several KB, offset). This can only occur in PIO mode,
DMA is not affected.
This doesn't occur in "normal" use because when reading a data CD, either
the sense is issued for a non-data command (in which case c_skip stays to
0), or an error occured and no data has been transfered, and c_skip is still 0.
I can't see a case where a data READ/WRITE could cause a sense tranfer without
error.
The second problem is that b_resid can be set to a false value (resulting of
the sense tranfer and not the data transfer). Again this is not a problem with
usual data tranfers because both values ends up being 0 when no error occurs.
|
| 1.22 | 15-Apr-1999 |
bouyer | If setting to pio mode < 3 fails, try to set pio mode 3 before returning
EIO. The spec says ATAPI devices should support "PIO 3 or better".
They are supposed to support less as well. Setting the device to a highter
mode than the controller shoul'nt be a problem, and this is likely what
happens with legaty ISA controllers.
Solve problem reported by Ruey-Shyang Guo.
|
| 1.21 | 08-Apr-1999 |
bouyer | Fix a long-standing bug with the ACAP_* flags:
- the cap field is a u_int8_t, so none of the defined flags would fit in.
Looks like nobody had a drive using 16 bytes commands.
- the ACAP_DRQ_* flags are all wrong. Just remove them and use the definitions
from ata/atareg.h, there's no need to duplicate theses. The effect of this
was that we were always polling for the command phase, even for drives
with interrupt DRQ. This didn't break until the code was changed to support
shared interrupts.
Should fix the lookup problems or 'boot hangs' reported by some users, and
kern/7111.
|
| 1.20 | 01-Apr-1999 |
bouyer | branches: 1.20.2;
- change the interrupt routines to take a 3rd arguments, set to 1 if we
are called from the interrupt or timeout handler, 0 otherwise.
- use this to know if we can busy-wait for wait_for_unbusy or wait_for_ready
This fixes a bug where CDs withot the DRQ_INTR capability would not busy-wait
for the CMDOUT phase.
While I'm there change 2 delay() to DELAY() for consistency, and
garbage-collect some old code from wdcintr() which has been ifdef'd out
for some time now.
|
| 1.19 | 25-Mar-1999 |
bouyer | Avoid busy-waiting when possible if we are not polling (real IRQ or timeout
callback). Shared PCI IRQ should now work (but still untested).
|
| 1.18 | 21-Feb-1999 |
hubertf | Only define WDCDEBUG if not already define (e.g. by kernel option)
|
| 1.17 | 15-Feb-1999 |
bouyer | If the REQUEST_SENSE command fails, issue a reset instead of XS_DRIVER_STUFFUP.
If REQUEST_SENSE succeeds but doens't return any data, return XS_SHORTSENSE
as the sense field will just contain a copy or the error register.
|
| 1.16 | 02-Feb-1999 |
bouyer | For IDE/ATAPI protocols botches, reset the bus and return XS_RESET
instead of XS_DRIVER_STUFFUP.
|
| 1.15 | 29-Jan-1999 |
bouyer | Cosmetic changes, from Soren S. Jorvan.
|
| 1.14 | 09-Jan-1999 |
thorpej | Make this compile again after the void * arith changes.
|
| 1.13 | 08-Jan-1999 |
augustss | Avoid arithmetic on `void *' since that's not ANSI C.
|
| 1.12 | 17-Dec-1998 |
bouyer | Add a new quirk flags, "ADEV_NOSENSE", for devices that don't
handle properly the request sense command. Add <FX320S, , q01> as being
sense-unfriendly in the quirk table.
|
| 1.11 | 16-Dec-1998 |
bouyer | Keep track of DMA errors, and downgrade the transert mode (UDMA ->DMA,
DMA->PIO) in case of 2 consecutive errors. Don't downgrade if the
PIO/DMA/UDMA modes were forced by a config flag.
|
| 1.10 | 21-Nov-1998 |
drochner | in wdc_softc: access the per-channel data via a pointer array instead of
an array of fixed-sized channel_softc elements. This way IDE controllers
which more than 1 channel (pciide) can extend the channel data easily
for private needs.
To avoid the double dereference at runtime, change the argument of
wdcstart() to the channel data pointer instead of the array index.
|
| 1.9 | 19-Nov-1998 |
thorpej | Adapt to the new scsipi_adapter interface.
|
| 1.8 | 17-Nov-1998 |
bouyer | In atapi_wdc.c, issue a 'REQUEST SENSE' command when appropriate.
Return XS_SENSE when the full sense info has been retrieved, or
XS_SHORTSENSE if only the sense key was available (from the error register)
Make atapi_interpret_sense() deal with this, and call scsipi_interpret_sense()
for XS_SENSE. (XXX sd_interpret_sense() and the ioctl code needs to be made
aware of XS_SHORTSENSE too ! sense hanlding for these is now less broken for
devices that support 'REQUEST SENSE')
All the ATAPI devices I have access to seems to honnor the SENSE_REQUEST
command, but I suspect some ATAPI devices will not (althouh it's mandatory).
The code should be able to deal with this, but is untested ...
|
| 1.7 | 19-Oct-1998 |
bouyer | Move the "warning: bcount value is %d after io" message in a WDCDEBUG_PRINT.
Ifdef DIAGNOSTIC, print the above message only if bcount is < 0.
|
| 1.6 | 13-Oct-1998 |
bouyer | Add a new drive flag, DRIVE_MODE, set when the drive properly reported its
PIO/DMA modes. Don't try a SET_FEATURE if it didn't report its mode.
|
| 1.5 | 13-Oct-1998 |
bouyer | In debug message, always print controller:channel:drive
|
| 1.4 | 13-Oct-1998 |
bouyer | Properly separate DEBUF_XFERS and DEBUG_FUNCS: DEBUF_XFERS is for data
transfers (may be called often), where DEBUG_FUNCS is for setup functions
(not used for normal contitions).
|
| 1.3 | 13-Oct-1998 |
bouyer | pciide.c: don't define WDCDEBUG, so it compiles on alpha. Correct a bogon
in the printing of DMA mode (piix3/4 only)
others: set the debug_mask to 0, so that debug messages are turned off by
default but can be easily turned on.
Reset drive_flags to 0 for unconfigured devices, so that they are ignored
later. For configured devices, reset state to 0 after probe/attach.
|
| 1.2 | 12-Oct-1998 |
bouyer | Merge bouyer-ide
|
| 1.1 | 04-Jun-1998 |
bouyer | branches: 1.1.2;
file atapi_wdc.c was initially added on branch bouyer-ide.
|
| 1.1.2.13 | 12-Oct-1998 |
bouyer | Sync with HEAD.
|
| 1.1.2.12 | 05-Oct-1998 |
bouyer | Rename WDC_CAPABILITY_PIO to WDC_CAPABILITY_MODE, as it's used to tell wether
the controller's driver can tell which timing mode it uses. Check this
before setting DMA modes too.
This allow the generic DMA code to work again.
|
| 1.1.2.11 | 04-Oct-1998 |
bouyer | New flag for wdc_xfer: C_DMA, set when a cmd will use DMA. This reduce some
if() in ata_wdc.c and atapi_wdc.c from 3 tests to one.
|
| 1.1.2.10 | 02-Oct-1998 |
bouyer | Cleanup a few error message, remove some dead code. Re-add a reset at end of
attach, unless we have WDC_NO_EXTRA_RESETS.
Add support for data32iot/data32ioh from -current: each controller pass
WDC_CAPABILITY_DATA32 if they can do 32bit, and WDC_CAPABILITY_DATA16 if they
can do 16bit. For controller that support both, the usual autodetect
mechanism is still used.
|
| 1.1.2.9 | 02-Oct-1998 |
bouyer | Use channel provided by your caller, not xfer->channel which is unitialized
at this point.
|
| 1.1.2.8 | 20-Sep-1998 |
bouyer | Init up xfer->drive in wdc_ata_bio(), or bad things will happen for drive 1.
No need to init xfer->channel, it's done in wdc_exec_xfer().
|
| 1.1.2.7 | 20-Sep-1998 |
bouyer | - always call config_found() from wdcattach(), instead of printing our own
the "not configured" message ourselve. When no atapibus is configured,
use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we
don't have to include the whole atapi support in an ata-only config.
- the dma_* function pointers take a 'int flags' as last argument, instead
of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write)
and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps).
- Reworked wdcprobe() so look more like the old one. A status of 0xff is
interpreted as "no drive" (freebsd does this); this this speed up
the probe for non-IDE machines (a reset timeout has been bumped to 31s,
to match the specs). The probe set ups the drive flags to either ATA or
ATAPI (depend on register signature) or OLD (if register signature don't
match, but a working drive appears to be there). Later the ATA code
will look at the OLD flag to decide wether there is no drive, or an old
(ST506) one if WDC_IDENTIFY fails.
- For known PCI chips, don't use the wdcprobe() heuristic to find wether
a channel is enabled/disabled; use the chip's specific registers for this.
- Clear the command queue when we have a polled command. This allows dumps to
work even if a command was being handled by the controller at panic time.
- Increase the number of available DMA segments by one, as user requests may
not be aligned on a page boundary.
|
| 1.1.2.6 | 13-Aug-1998 |
bouyer | - sync with HEAD
- better error detection and report for wd drives
- better wddump() support
- wdcwait() now takes a timout argument, so that we can honnor the
timeout argument of scsipi cmds.
|
| 1.1.2.5 | 23-Jun-1998 |
leo | Add necessary cruft to handle swapped byte-lanes on the data-register. It
basically boils down to the fact that there are 2 new options:
WDC_CAPABILITY_{ATA,ATAPI}_NOSTREAM with which you are able to control
the usage bus_space_read_multi_X() vs. bus_space_read_multi_stream_X().
|
| 1.1.2.4 | 19-Jun-1998 |
leo | Use the stream functions where appropriate.
|
| 1.1.2.3 | 19-Jun-1998 |
leo | Fix hw-locking.
|
| 1.1.2.2 | 11-Jun-1998 |
bouyer | pciide: when setting up DMA mode, also enable fast timings for PIO modes
if possible. For ATAPI devices, commands are always send via PIO.
atapi: matain a drive state (like wd drives) and use it to set up the
timings if needed.
|
| 1.1.2.1 | 04-Jun-1998 |
bouyer | Commit changes to the IDE system in a branch. This allows a better separation
between higth-level and low-level (i.e. registers read/write) and generalize
the queue for all commands. This also add supports for IDE DMA.
|
| 1.20.2.7 | 14-Aug-2000 |
he | Pull up revision 1.38 (requested by bouyer):
If SET_FEATURE for piomode is rejected, fall back to mode 0
instead of giving up. Works around "piomode error (0x4)" problem
with some ATAPI ZIP drives.
|
| 1.20.2.6 | 07-Jul-2000 |
he | Apply patch (requested by bouyer):
Add support for the following PCIIDE controllers:
o AMD 756
o CMD PCI0648 and PCI0649
o Hightpoint HPT366
o OPTi 82c621 (and a few of its derivatives)
o Promise Ultra/33 and Ultra/66
o Intel 82801 (ICH/ICH0)
Also fix PR#10437 (detect more ATAPI devices).
|
| 1.20.2.5 | 23-Jan-2000 |
he | Pull up revision 1.30 (via patch, requested by bouyer):
Improve the downgrade logic to work better in some cases (especially
for Acer Labs M5229 controllers with Ultra-DMA/66 disks).
|
| 1.20.2.4 | 28-Aug-1999 |
he | Pull up revision 1.22:
Retry PIO mode 3 if PIO mode < 3 fails.
|
| 1.20.2.3 | 25-Aug-1999 |
he | Apply patch, pulling up 1.23->1.24 + "return" part of 1.24->1.25:
Add some debug printfs, change timeout handling to be more efficient
and less error-prone, and fix the "lockup after lost interrupt"
problem some users have been reporting. (bouyer)
|
| 1.20.2.2 | 05-May-1999 |
perry | branches: 1.20.2.2.2;
pullup 1.22->1.23 (fvdl)
|
| 1.20.2.1 | 08-Apr-1999 |
bouyer | Pull up atapi_wdc.c rev 1.21 and scsipiconf.h rev 1.30:
Fix botch with atapi capability flags, preventing 16 bytes commands to be ever
sent, and interrupt DRQ capable devices to effectively use interrupts for
the CMD_OUT phase. Didn't break until the code was changed to handle shared
interrupts (and I guess nobody has a device that requires 16 bytes commands
yet). Should fix the hangs reported by some users at first access of an ATAPI
device.
|
| 1.20.2.2.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.27.6.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.27.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.27.2.7 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.27.2.6 | 15-Jan-2001 |
bouyer | Let the mid-layer issue the request sense command.
|
| 1.27.2.5 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.27.2.4 | 20-Oct-1999 |
thorpej | Fix merge botch.
|
| 1.27.2.3 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.27.2.2 | 20-Oct-1999 |
thorpej | Eliminate sleeping in adapter routines. It is not legal as the queue may
be run from interrupt context.
|
| 1.27.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.35.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.36.2.3 | 09-Feb-2002 |
he | Pull up revision 1.48 (requested by bouyer):
If we change the ATA modes, also reset the device, so that we
can retry the command with the new mode. Fixes PR#15331.
|
| 1.36.2.2 | 27-Dec-2001 |
he | Pull up revision 1.39 (requested by mycroft):
Make DVD commands work on drives configured for DMA.
|
| 1.36.2.1 | 04-Aug-2000 |
bouyer | pull up 1.37->1.38 (approved by thorpej):
If the SET_FEATURE for piomode is rejected, fallback to mode 0 instead of
reporting an error. This fixes the "piomode error (0x4)" problem with some
ATAPI ZIP drives reported on port-i386.
|
| 1.39.2.9 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.39.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.39.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.39.2.6 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.39.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.39.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.39.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.39.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.39.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.43.2.5 | 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
|
| 1.43.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.43.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.43.2.2 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.43.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.51.2.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.51.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.52.2.1 | 01-Nov-2002 |
tron | Pull up revision 1.53 (requested by bouyer in ticket #733):
Do PIO transfers for len < 4 bytes. From Matthias Drochner in private mail.
|
| 1.55.6.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.55.6.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.55.6.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.55.6.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.55.6.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.55.6.2 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.55.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.69.2.2 | 11-Aug-2004 |
jmc | branches: 1.69.2.2.2;
Pullup rev 1.71 (requested by bouyer in ticket #733)
Implement asynchronous channel reset.
Use this to reset the channel before doing a dump, instead of the hack in
wdc_exec_xfer() based on C_POLL. This hack was causing problems on
controllers with a shared queue, because we now can have C_POLL set during
concurent channels probes
|
| 1.69.2.1 | 29-May-2004 |
tron | Pull up revision 1.70 (requested by bouyer in ticket #396):
Add a delay(10) after re-enabling interrupts in the control register.
Some controllers/drives (e.g. SataLink 3114 with WD Raptor) require
it. Should fix kern/23808 by Chris Gilbert, patch suplied by Chris Gilbert
on tech-kern, extended to all places enabling interrupts by me.
|
| 1.69.2.2.2.1 | 18-Jul-2005 |
riz | Pull up revision 1.95 (requested by thorpej in ticket #2072):
- When starting an ATA or ATAPI transfer, handle the case where
(*dma_init)()
returns EINVAL, indicating that DMA cannot be done for this
transfer.
Fall back to PIO in this case.
- Add a geodeide_dma_init() routine that checks to make sure that
transfers
start on a 16 byte boundary, returning EINVAL if not. Works
around a chip
bug that causes a hard system hang.
Problem reported and patch tested by Erik Fair.
|
| 1.89.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.89.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.92.2.1 | 06-Jul-2005 |
tron | Pull up revision 1.95 (requested by thorpej in ticket #553):
- When starting an ATA or ATAPI transfer, handle the case where (*dma_init)()
returns EINVAL, indicating that DMA cannot be done for this transfer.
Fall back to PIO in this case.
- Add a geodeide_dma_init() routine that checks to make sure that transfers
start on a 16 byte boundary, returning EINVAL if not. Works around a chip
bug that causes a hard system hang.
Problem reported and patch tested by Erik Fair.
|
| 1.95.2.5 | 24-Mar-2008 |
yamt | sync with head.
|
| 1.95.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.95.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.95.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.95.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.97.6.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.97.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.97.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.97.2.1 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.98.8.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.98.6.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.98.2.3 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.98.2.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.98.2.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.99.2.1 | 13-Jul-2006 |
gdamore | Merge from HEAD.
|
| 1.101.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.101.2.2 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.101.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.103.2.1 | 29-Jan-2007 |
bouyer | Pull up following revision(s) (requested by reinoud in ticket #388):
sys/dev/scsipi/atapi_wdc.c: revision 1.104
Fix for atapi devices that have 32 bit databusses selected. The devices
won't error out but will return garbage data. Implemented by generalising
the odd-check for PIO selection to check for 4 byte alignment in offset and
size.
Note that this check could be relaxed by checking if the ATAPI/ide bus
width is 16 for the device but its unknown to me if that information is
available here.
|
| 1.104.24.1 | 25-Oct-2007 |
bouyer | Sync with HEAD.
|
| 1.104.20.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.104.20.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.104.18.1 | 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.
|
| 1.104.6.1 | 23-Oct-2007 |
ad | Sync with head.
|
| 1.105.16.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.105.16.2 | 05-Oct-2008 |
mjf | Sync with HEAD.
|
| 1.105.16.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.105.12.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.107.10.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.107.10.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.107.6.1 | 10-Oct-2008 |
skrll | Sync with HEAD.
|
| 1.107.4.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.107.4.1 | 11-Mar-2010 |
yamt | sync with head
|
| 1.108.4.1 | 20-Nov-2008 |
snj | Pull up following revision(s) (requested by bouyer in ticket #67):
sys/dev/ic/wdc.c: revision 1.256
sys/dev/ata/ata.c: revision 1.102
sys/dev/scsipi/atapi_wdc.c: revision 1.109
sys/dev/ata/atavar.h: revision 1.78
sys/dev/ata/ata_wdc.c: revision 1.91
cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so
we can't use it here. Rssurect ATACH_TH_RUN, backing out
src/sys/dev/ata/ata.c 1.101
src/sys/dev/ata/ata_wdc.c 1.90
src/sys/dev/ata/atavar.h 1.77
src/sys/dev/ic/wdc.c 1.255
src/sys/dev/scsipi/atapi_wdc.c 1.108
Should fix kern/39927 and kern/39725.
|
| 1.108.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.110.4.2 | 31-May-2011 |
rmind | sync with head
|
| 1.110.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.110.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.111.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.112.8.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.112.4.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.112.4.1 | 23-May-2012 |
yamt | sync with head.
|
| 1.119.16.3 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.119.16.2 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.119.16.1 | 22-Apr-2016 |
skrll | Sync with HEAD
|
| 1.119.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.122.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.123.4.14 | 30-Sep-2017 |
jdolecek | must drop channel lock before calling ata_dmaerr() to avoid 'locking against
myself' in case of consecutive errors from ata_reset_channel() called via
ata_downgrade_mode()
|
| 1.123.4.13 | 10-Sep-2017 |
jdolecek | refactor code so that xfer c_start() hook is called with channel mutex held,
and hence the controller submit code no longer relies on spl
tested all the affected drivers - wdc (via piixide), ahci, mvsata, siisata,
both disk and atapi I/O
|
| 1.123.4.12 | 12-Aug-2017 |
jdolecek | convert the atabus thread to use the channel lock and a condvar, adjust
code which sets the relevant channel flags to take the lock while doing so
|
| 1.123.4.11 | 12-Aug-2017 |
jdolecek | remove all logic around ATACH_IRQ_WAIT and channel-global ch_error/ch_status,
so that there is less hidden state shared by commands; primary intent is
to make the NCQ and non-NCQ paths more similar, and remove possibility
of incorrect handling for the NCQ commands
tested both disk and ATAPI - piixide(4) on QEMU, and siisata(4),
ahcisata(4), mvsata(4) on real hw
|
| 1.123.4.10 | 29-Jul-2017 |
jdolecek | do not use freed xfer in error message in wdc_atapi_get_params()
|
| 1.123.4.9 | 27-Jun-2017 |
jdolecek | attend error paths, more strict asserts and code consistency
- atastart() and ata_kill_pending() now KASSERT() that all xfers on queue
have same channel
- inactive xfers are killed via new reason KILL_GONE_INACTIVE, controller
code must not call any resource deactivation in that case
- c_intr() must call ata_waitdrain_xfer_check() as first thing, and must not
further touch any xfer structures on exit path; any resource cleanup
is supposed to be done in c_kill_xfer()
- c_kill_xfer() should never call atastart()
- ata_waitdrain_check() removed, replaced by ata_waitdrain_xfer_check()
- ATA_DRIVE_WAITDRAIN handling converted to use condvar
- removed unused ata_c callback
|
| 1.123.4.8 | 21-Jun-2017 |
jdolecek | two last forgotten on-stack xfers replaced with using the queue xfer -
ata_get_params() and ata_set_mode()
fix wdc_atapi_get_params() to free the xfer used for soft reset before
calling ata_get_params() - it's now necessary, as ata_get_params() doesn't
invent a private xfer own any more
|
| 1.123.4.7 | 20-Jun-2017 |
jdolecek | remove ata queue downsizing - every device, attached to the same channel,
uses slots according to it's own limits
wdc code changed to expect maximum one active xfer, and not check number
of openings in the channel; this is to facilitate using wdc functions
for e.g. handling of atapi commands for drivers which support both ATAPI
and NCQ
|
| 1.123.4.6 | 19-Jun-2017 |
jdolecek | add ata_channel lock, use it to protect queue manipulation (only that for now);
add ata_channel_detach() to destroy the locks
change ata_get_xfer() so that it can wait for xfer, convert all on-stack
xfer code to use the blocking variant
fix siisata_reset_drive() to use polled reset and not try ata_activate_xfer(),
convert drive probe code also over from slot0 XXX to ata_get_xfer()
drive reset and PMP now works on siisata(4) too; changes tested also
on piixide(4), ahci(4), mvsata(4)
|
| 1.123.4.5 | 16-Jun-2017 |
jdolecek | adjust reset channel and dump paths
- channel reset now always kills active transfer, even on dump path, but
now doesn't touch the queued waiting transfers; also kill_xfer hook is
always called, so that HBA can free any private xfer resources and thus
the dump request has chance to work
- kill_xfer routines now always call ata_deactivate_xfer(); added KASSERT()s
to ata_free_xfer() to expect deactivated xfer
- when called during channel reset before dump, ata_kill_active() drops
any queued waiting transfers without processing
- do not (re)queue any transfers in wddone() when dumping
- kill AT_RST_NOCMD flag
This should also hopefully fix the 'polled command has been queued' panic
as reported in:
PR kern/11811 by John Hawkinson
PR kern/47041 by Taylor R Campbell
PR kern/51979 by Martin Husemann
dump tested working with piixide(4) and ahci(4). mvsata(4) dump times out,
but otherwise tested working, will be fixed separately. siisata(4) mechanically
changed and not tested.
|
| 1.123.4.4 | 19-Apr-2017 |
jdolecek | adjust ata code to support more than one active command, including the
timeout handling, add support for NCQ commands
move probe for NCQ and number of tags to middle layer, negotiate mutual
support between drive and controller
implement NCQ support in ahci(4)
|
| 1.123.4.3 | 15-Apr-2017 |
jdolecek | pass also ata_command via ata_xfer, callers of ata_exec_command() is now
responsible for allocation/disposal of the structure
change code to allocate ata_xfer for commands on stack same way as previously
the ata_command were, using c_slot 0; adjust asserts so that it would allow
several xfers with same c_slot, as long as only one such transfer is active
at a time
|
| 1.123.4.2 | 15-Apr-2017 |
jdolecek | make ata_xfer's allocated as part of ata_queue and make it include ata_bio;
they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the
pool
modify wd(4) to file the bio requests using the xfers and hence
make it possible to have more than one active I/O request in flight;
ata_bio callback doesn't need to allocate any memory any more,
require it to never return ATACMD_TRY_AGAIN
move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive,
not per transfer
drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio
c_hwslot
|
| 1.123.4.1 | 10-Apr-2017 |
jdolecek | ATA infrastructure improvements to eventually support more outstanding
commands
patch by Matt Thomas
|
| 1.129.6.4 | 22-Sep-2018 |
jdolecek | separate ata_xfer slot allocation and the memory allocation, so that
there can be more queued xfers than number of supported slots by controller,
and use a pool instead of custom pre-allocation
primarily to help PR kern/52614
remove no longer needed custom wd(4) logic for flush cache
switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting
for the memory, they are all called from process context and this
avoids spurious failures
|
| 1.129.6.3 | 17-Sep-2018 |
jdolecek | move ATAPI-only members of ata_xfer to an union struct to further save space
|
| 1.129.6.2 | 17-Sep-2018 |
jdolecek | move low-level protocol handlers hooks from ata_xfer to separate struct,
initialized statically
primarily to reduce ata_xfer struct size, but also improves readibility,
and enforces consistency
|
| 1.129.6.1 | 31-Aug-2018 |
jdolecek | refactor ata_xfer to be just dumb structure; move all callouts/condvars out
retry callout to wd(4); reset callout and the active/cmd finish condvars
to channel queue; change code using the condvars so it works if there
are multiple waiters
simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer()
with ata_wait_cmd()
fix the callout_invoking/ack race handling code for timeouts to
actually have chance to work; change mvsata(4) to use generic timeout func
towards resolution of kern/52614
|
| 1.129.4.4 | 21-Apr-2020 |
martin | Sync with HEAD
|
| 1.129.4.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.129.4.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.129.4.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.129.2.2 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.129.2.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.133.4.1 | 30-Dec-2022 |
martin | Pull up following revision(s) (requested by tsutsui in ticket #1557):
sys/dev/ic/ahcisata_core.c: revision 1.83
sys/dev/ic/ahcisata_core.c: revision 1.102
sys/dev/ata/ata.c: revision 1.164
sys/dev/ata/ata_wdc.c: revision 1.115
sys/dev/ata/ata_recovery.c: revision 1.4
sys/dev/ic/siisata.c: revision 1.42
sys/dev/ic/wdc.c: revision 1.308
sys/dev/ic/mvsata.c: revision 1.56
sys/dev/scsipi/atapi_wdc.c: revision 1.138
sys/dev/ic/siisata.c: revision 1.49
sys/dev/ata/atavar.h: revision 1.105
sys/dev/ata/wd.c: revision 1.460
sys/dev/ata/ata.c: revision 1.155
sys/dev/ata/wd.c: revision 1.462
sys/dev/ata/atavar.h: revision 1.109
sys/dev/ata/satapmp_subr.c: revision 1.16
sys/dev/ic/wdc.c: revision 1.299
sys/dev/ic/ahcisata_core.c: revision 1.93
sys/dev/ata/ata_wdc.c: revision 1.120
sys/dev/ic/wdcvar.h: revision 1.100
sys/dev/scsipi/atapi_wdc.c: revision 1.141
sys/dev/ic/mvsata.c: revision 1.61
sys/dev/usb/umass_isdata.c (apply patch)
drop wd lock in wdstart1() before calling the ata_bio hook; when called
from ata thread context, that can still need to sleep for wdc attachments
in wdcwait()
fix use-after-free for ata xfer on bio submission found by KASAN
driver ata_bio hooks read parts of the xfer after ata_exec_xfer()
call in order to determine return value, change so that the hook
doesn't return any value - callers do not care already,
as all I/O requests are asynchronous
this problem was uncovered by recent change for wd(4) to not hold
wd mutex during ata_bio call, the interrupt for the xfer might
thus actually fire immediately
adjust also ata_exec_command driver hooks similarily - remove all
completion and waiting logic from drivers, upper layer ata code
using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself
PR kern/55169 by Nick Hudson
Function declaration formating whitespace consistency. NFCI.
PR kern/56403
Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ:
(1) Change ata_xfer_ops:c_poll from void to int function. When it returns
ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again.
(2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is
achieved.
A similar change has been made for mvsata(4) (see mvsata_bio_poll()),
and no functional changes for other devices.
This is how the drivers worked before jdolecek-ncq branch was merged.
Note that this changes are less likely to cause infinite recursion:
(1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE
in its error handling paths via wdc_ata_bio_done().
(2) Return value from c_start (= wdc_ata_bio_start()) is checked in
ata_xfer_start().
Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines
terminate the recursion for wdc(4). The situation is similar for mvsata(4).
Still, there is a possibility where ata_xfer_start() takes long time to
finish a normal operation. This can result in a delayed response for lower
priority interrupts. But, I've never observed such a situation, even when
heavy thrashing takes place for swap partition in wd(4).
"Go ahead" by jdolecek@.
|
| 1.134.2.1 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.137.2.1 | 20-Apr-2020 |
bouyer | Sync with HEAD
|
| 1.138.4.1 | 02-Apr-2021 |
thorpej | config_found_ia() -> config_found() w/ CFARG_IATTR.
|
| 1.139.8.1 | 04-Aug-2021 |
thorpej | Adapt to CFARGS().
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file atapicd.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
bouyer | Moved atapicd and scsicd to cd_scsi and cd_atapi, to be concistent with
other bus front-end naming conventions (Suggested by Mike Long)
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.95 | 09-Apr-2022 |
riastradh | sys: Use membar_release/acquire around reference drop.
This just goes through my recent reference count membar audit and
changes membar_exit to membar_release and membar_enter to
membar_acquire -- this should make everything cheaper on most CPUs
without hurting correctness, because membar_acquire is generally
cheaper than membar_enter.
|
| 1.94 | 12-Mar-2022 |
riastradh | sys: Membar audit around reference count releases.
If two threads are using an object that is freed when the reference
count goes to zero, we need to ensure that all memory operations
related to the object happen before freeing the object.
Using an atomic_dec_uint_nv(&refcnt) == 0 ensures that only one
thread takes responsibility for freeing, but it's not enough to
ensure that the other thread's memory operations happen before the
freeing.
Consider:
Thread A Thread B
obj->foo = 42; obj->baz = 73;
mumble(&obj->bar); grumble(&obj->quux);
/* membar_exit(); */ /* membar_exit(); */
atomic_dec -- not last atomic_dec -- last
/* membar_enter(); */
KASSERT(invariant(obj->foo,
obj->bar));
free_stuff(obj);
The memory barriers ensure that
obj->foo = 42;
mumble(&obj->bar);
in thread A happens before
KASSERT(invariant(obj->foo, obj->bar));
free_stuff(obj);
in thread B. Without them, this ordering is not guaranteed.
So in general it is necessary to do
membar_exit();
if (atomic_dec_uint_nv(&obj->refcnt) != 0)
return;
membar_enter();
to release a reference, for the `last one out hit the lights' style
of reference counting. (This is in contrast to the style where one
thread blocks new references and then waits under a lock for existing
ones to drain with a condvar -- no membar needed thanks to mutex(9).)
I searched for atomic_dec to find all these. Obviously we ought to
have a better abstraction for this because there's so much copypasta.
This is a stop-gap measure to fix actual bugs until we have that. It
would be nice if an abstraction could gracefully handle the different
styles of reference counting in use -- some years ago I drafted an
API for this, but making it cover everything got a little out of hand
(particularly with struct vnode::v_usecount) and I ended up setting
it aside to work on psref/localcount instead for better scalability.
I got bored of adding #ifdef __HAVE_ATOMIC_AS_MEMBAR everywhere, so I
only put it on things that look performance-critical on 5sec review.
We should really adopt membar_enter_preatomic/membar_exit_postatomic
or something (except they are applicable only to atomic r/m/w, not to
atomic_load/store_*, making the naming annoying) and get rid of all
the ifdefs.
|
| 1.93 | 07-Aug-2021 |
thorpej | Merge thorpej-cfargs2.
|
| 1.92 | 24-Apr-2021 |
thorpej | branches: 1.92.8;
Merge thorpej-cfargs branch:
Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.
Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)
Remove unnecessary or redundant interface attributes where they're not
needed.
There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)
...and a sentinel value CFARG_EOL.
Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.
Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).
|
| 1.91 | 17-Jun-2017 |
mlelstv | branches: 1.91.22;
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.90 | 29-Nov-2016 |
mlelstv | branches: 1.90.8;
reference count adapter mutex possibly shared by multiple channels.
fix error in atapibusdetach, when a child device cannot be detached,
keep atapibus instance alive.
|
| 1.89 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.88 | 13-Mar-2016 |
tsutsui | branches: 1.88.2;
Sync with scsiconf.c. (use aprint_normal(9) for "not configured")
|
| 1.87 | 05-Mar-2014 |
skrll | branches: 1.87.6;
Don't probe beyond chan_ntargets in atapi_probe_bus in the "all" case.
PR/48626: ahci_atapi_probe_device panic with kmemguard
|
| 1.86 | 24-Jun-2012 |
riastradh | branches: 1.86.2; 1.86.4;
Take the kernel lock in atapibusdetach just like atapibuschilddet.
Fixes kassert in scsipi_lookup_periph when I press the power button
on one of my laptops (and maybe another one) to power it off.
ok mrg
|
| 1.85 | 19-Apr-2012 |
bouyer | Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.84 | 06-Apr-2012 |
chs | take the kernel lock during detach of atapibus as well.
|
| 1.83 | 07-Jun-2010 |
pgoyette | branches: 1.83.8; 1.83.12; 1.83.14;
Update scsiverbose module to use module_autoload() rather than module_load().
Load the module right before each attempt to use its features, and let the
module subsystem handle unloading.
|
| 1.82 | 30-May-2010 |
pgoyette | Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.81 | 12-Nov-2009 |
dyoung | Remove superfluous activation hooks.
|
| 1.80 | 19-Oct-2009 |
bouyer | Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !
|
| 1.79 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.78 | 07-Apr-2009 |
dyoung | Detach atapibus(4), scsibus(4), cd(4), and sd(4) during shutdown.
Destroy sd->sc_callout in sddetach(). Delete some dead code in
cddetach().
|
| 1.77 | 24-Mar-2008 |
cube | branches: 1.77.4; 1.77.12; 1.77.18;
Split device_t and softc for atapibus(4).
|
| 1.76 | 08-Mar-2008 |
yamt | fix a double-free bug introduced by the following change. PR/38179.
revision 1.75
date: 2008/01/29 17:26:57; author: dyoung; state: Exp; lines: +34 -15
Use device_t. Add a handler for child detachment. Now I can detach
cd0 at atapibus0 without getting a panic when atapibus0 detaches,
later.
|
| 1.75 | 29-Jan-2008 |
dyoung | branches: 1.75.2; 1.75.6;
Use device_t. Add a handler for child detachment. Now I can detach
cd0 at atapibus0 without getting a panic when atapibus0 detaches,
later.
|
| 1.74 | 09-Dec-2007 |
jmcneill | Merge jmcneill-pm branch.
|
| 1.73 | 01-Dec-2007 |
jmcneill | branches: 1.73.2; 1.73.4;
aprintify
|
| 1.72 | 16-Nov-2006 |
christos | branches: 1.72.22; 1.72.24; 1.72.30;
__unused removal on arguments; approved by core.
|
| 1.71 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.70 | 30-Mar-2006 |
thorpej | branches: 1.70.8; 1.70.10;
Use device_private().
|
| 1.69 | 11-Dec-2005 |
christos | branches: 1.69.4; 1.69.6; 1.69.8; 1.69.10; 1.69.12;
merge ktrace-lwp.
|
| 1.68 | 26-Aug-2005 |
drochner | s/locdesc_t/int/g
|
| 1.67 | 29-May-2005 |
christos | branches: 1.67.2;
- Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.66 | 13-Sep-2004 |
drochner | a round of autoconf cleanup:
-convert submatch() style functions (passed to config_search() or
config_found_sm()) to the locator passing variants
-pass interface attributes in some cases
-make submatch() functions look uniformly as far as possible
-avoid macros which just hide cfdata members, and reduce dependencies
on "locators.h"
|
| 1.65 | 21-Aug-2004 |
thorpej | Const poison scsipi_dtype().
|
| 1.64 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.63 | 17-Oct-2003 |
mycroft | Remove the Fuji quirk from here, too.
|
| 1.62 | 08-Oct-2003 |
bouyer | Make the ATA mid-layer appears as atabus, as proposed in
http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html
This adds a device (atabus) between IDE controllers and wd or atapibus, to
have each ATA channel show up in the device tree. Later there will be atabus
devices in /dev, so that we can do IOCTL on them.
Each atabus has its own kernel thread, to handle operations that needs polling,
e.g. reset and others.
Device probing on each bus it defered to the atabus thread creation.
This allows to do the reset and basic device probes in parallel, which reduce
boot time on systems with several pciide controllers.
|
| 1.61 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.60 | 17-Sep-2003 |
mycroft | Remove PQUIRK_BYTE5_ZERO.
|
| 1.59 | 09-Sep-2003 |
mycroft | Exorcise PQUIRK_NODOORLOCK.
|
| 1.58 | 08-Sep-2003 |
mycroft | Do a START UNIT only if the TEST UNIT READY reports that the device is not
ready. This avoids gratuitously starting the motor on floppy and CD-ROM
drives, and eliminates the need for the audio playing test in cdopen().
Therefore, also remove PQUIRK_NOSTARTUNIT.
|
| 1.57 | 03-Apr-2003 |
erh | branches: 1.57.2;
Add a quirk to allow my NEO Jukebox to work again.
|
| 1.56 | 19-Feb-2003 |
hannken | Add PQUIRK_NO_FLEX_PAGE for Fujitsu MO MCJ3230AP.
This drive returns bogus geometry (0 heads, 0 sectors).
|
| 1.55 | 01-Jan-2003 |
thorpej | Use aprint_normal() in cfprint routines.
|
| 1.54 | 04-Oct-2002 |
soren | As in scsiconf.c, don't print the numberic device type in the attach message.
|
| 1.53 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.52 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.51 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.50 | 27-Sep-2002 |
thorpej | Introduce a new routine, config_match(), which invokes the
cfattach->ca_match function in behalf of the caller. Use it
rather than invoking cfattach->ca_match directly.
|
| 1.49 | 19-Sep-2002 |
jmc | Force the initial probes to happen within the newly forked off kthread.
This eliminates problems where the underlying interrupt handler isn't the
specific layer calling scsipi_complete() for a given scsi transaction.
This avoids deadlocks where the kthread that called the autoconf routines
to configure a scsibus shouldn't be the one put to sleep waiting on a
scsipi_complete (only the scsibus's kthread should be doing that).
To avoid jitter this will force the scsibus's to probe in the order they
run through autoconf (so machines with multiple bus's don't move sd* devices
around on every reboot).
|
| 1.48 | 01-Apr-2002 |
bouyer | Add a chan_name to struct scsipi_channel, holding the channel's name.
Set this to dv_xname for scsibus and atapibus.
Set the name of the kernel thread to chan_name instead of controller's
name:channel number (so that we can use this name for controller-specific
threads).
|
| 1.47 | 22-Dec-2001 |
gehenna | The FujiFilm USB storage drive doesn't have a flex geom page and
doesn't know REQUEST SENSE.
|
| 1.46 | 03-Dec-2001 |
bouyer | Update my copyrigth.
|
| 1.45 | 02-Dec-2001 |
bouyer | No need to include atavar.h here.
|
| 1.44 | 02-Dec-2001 |
bouyer | Clean up attach of wd/atapibus:
kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead
of ata_atapi_attach as attach arch. Create a ata_device, compatible with
scsipi_channel, to attach wd.
|
| 1.43 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.42 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.41 | 13-Sep-2001 |
enami | Pass the correct pointer to atapibusprint().
|
| 1.40 | 14-May-2001 |
bouyer | branches: 1.40.2; 1.40.4;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.39 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.38 | 01-Apr-2001 |
augustss | The ZiO! MMC adapter doesn't have a flex geometry page.
|
| 1.37 | 20-Mar-2001 |
augustss | Add quirks for Panasonic MultiMediaCard adapter (attaches via USB).
|
| 1.36 | 18-Jan-2001 |
jdolecek | branches: 1.36.2;
constify
|
| 1.35 | 02-Apr-2000 |
augustss | Let the device thet gets an atapibus attached specify how to kill pending
transfers by giving a function pointer.
The old method always called wdc specific code.
|
| 1.34 | 01-Apr-2000 |
bouyer | - DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op
if an IRQ was not detected, unless the force flag was given. Use this to
detect if the IRQ was for us (closer to shared IRQ for controllers which
don't have their own IRQ handler in pciide.c) and to poll for DMA xfer.
Also makes the timeout recovery code simpler.
- ATAPI cleanup: don't call controller-specific functions from atapiconf.c
(wdc_*), so that it's possible to attach an atapibus to something else
than a wdc/pciide (Hi Lennart :).
Overload struct scsi_adapter with struct atapi_adapter, defined
as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points
to an scsi_adapter, atapi code casts it to atapi_adapter if needed.
Move atapi_softc to atapiconf.h so that it can be used by the underlying
controller code (e.g. atapi_wdc.c).
Add an atapi-specific callback *atapi_probedev(), which probe a drive
in a controller-specific way, allocate the sc_link and fills in the
ataparams if needed. It then calls atapi_probedev() (from atapiconf.c)
to do the generic initialisations and attach the device.
- While I'm there merge and centralise the state definitions in atavar.h.
It should now be possible to use a common ata/atapi routine to set the
drive's modes (will do later).
|
| 1.33 | 28-Mar-2000 |
augustss | Change a printf() to a panic() since the kernel is going to die on the next
line anyway.
|
| 1.32 | 17-Mar-2000 |
soren | atapiprint() does not exist.
|
| 1.31 | 28-Feb-2000 |
jdolecek | Add ADEV_NOSENSE quirk for HITACHI CDR-7730.
Patch sent by Hume Smith in kern/9489.
|
| 1.30 | 20-Jan-2000 |
mjacob | Nobody said no to adding a pointer to original scsi inquiry data
to the scsibus attach args. Make sure it's nulled for ATAPI. Also,
for scsiconf.c, modify SENA's quirk entry.
|
| 1.29 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.28 | 30-Sep-1999 |
thorpej | branches: 1.28.2; 1.28.4; 1.28.6;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.27 | 23-Sep-1999 |
enami | Allow to detach wdc, atapibus, wd and cd.
|
| 1.26 | 08-Jul-1999 |
bouyer | Add another CD to the quirk table.
|
| 1.25 | 15-Feb-1999 |
bouyer | branches: 1.25.2; 1.25.4; 1.25.6;
Revert to 1.23, this was not supposed to be checked in yet.
|
| 1.24 | 15-Feb-1999 |
bouyer | For CDIOCCLOSE, ignore media changes, as some drive seems to return this
info here.
|
| 1.23 | 26-Jan-1999 |
bouyer | Another CD drive that doesn't support REQUEST SENSE, from Patrick Welche.
|
| 1.22 | 21-Jan-1999 |
bouyer | Add a mitsumi CD that can't handle REQUEST SENSE command. From
Johan Danielsson <joda@pdc.kth.se>.
|
| 1.21 | 19-Jan-1999 |
bouyer | Move test for SDEV_NOSTARTUNIT quirk from sd.c to scsipi_start().
Add a SDEV_NOSTARTUNIT quirk entry for BCD-16X 1997-04-25", "", "VER 2.2"
CD-rom (from Michael Santos).
|
| 1.20 | 19-Jan-1999 |
bouyer | "Memorex CRW-2642" CD-R can't properly handle REQUEST SENSE command.
Add it to the quick table, per PR kern/6844.
|
| 1.19 | 21-Dec-1998 |
leo | Change quirk entry: "FX320S", "", " q01" to: "FX320S", "", "q01". This
space does seem to matter...
|
| 1.18 | 17-Dec-1998 |
bouyer | Add a new quirk flags, "ADEV_NOSENSE", for devices that don't
handle properly the request sense command. Add <FX320S, , q01> as being
sense-unfriendly in the quirk table.
|
| 1.17 | 16-Dec-1998 |
bouyer | Leave some room for the string-terminating char in temp. buffers.
Pointed out by Scott Presnell (thanks !).
|
| 1.16 | 08-Dec-1998 |
thorpej | When allocating a device's scsipi_link, initialize the pending_xfers
queue.
|
| 1.15 | 19-Nov-1998 |
thorpej | Add a reference to the adapter before probing the bus, and delete it
once we are done probing.
|
| 1.14 | 13-Oct-1998 |
enami | Make this file compile again with -DATAPI_DEBUG_PROBE.
|
| 1.13 | 12-Oct-1998 |
bouyer | Merge bouyer-ide
|
| 1.12 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.11 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.10 | 31-Jul-1998 |
thorpej | Use the pool allocator for scsipi_xfer structures.
|
| 1.9 | 15-Jan-1998 |
cgd | branches: 1.9.2; 1.9.4;
add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.8 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.7 | 06-Jan-1998 |
bouyer | Add a few quirk entries from OpenBSD.
|
| 1.6 | 05-Nov-1997 |
bouyer | Add a quirk for a sanyo CD (From OpenBSD).
|
| 1.5 | 31-Oct-1997 |
bouyer | Add NOCAPACITY quirk entry for MATSHITA CR-574, rev 1.02.
|
| 1.4 | 01-Oct-1997 |
enami | branches: 1.4.2;
Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.3 | 28-Aug-1997 |
bouyer | Added a quirk entry for the SANYO CRD-254P (from OpenBSD), and
one for the MATSHITA CR-574 (from Josef Wulf <wulf@ping.net.au> via OpenBSD).
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file atapiconf.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.3 | 17-Jul-1997 |
bouyer | Use "locator.h".
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: spaces -> tabs in a few places, and
a general KNF sweep.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.3 | 01-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file atapiconf.c was added on branch marc-pcmcia on 1997-08-27 23:32:56 +0000
|
| 1.4.2.3 | 05-Feb-1999 |
cgd | pull up rev 1.7 from trunk (bouyer)
|
| 1.4.2.2 | 06-Nov-1997 |
mellon | Pull rev 1.6 up from trunk (bouyer)
|
| 1.4.2.1 | 31-Oct-1997 |
mellon | Pull rev 1.5 up from trunk (bouyer)
|
| 1.9.4.1 | 08-Aug-1998 |
eeh | Revert cdevsw mmap routines to return int.
|
| 1.9.2.4 | 20-Sep-1998 |
bouyer | Remove a stale 'XXX'.
|
| 1.9.2.3 | 11-Sep-1998 |
bouyer | Sync with HEAD.
|
| 1.9.2.2 | 13-Aug-1998 |
bouyer | - sync with HEAD
- better error detection and report for wd drives
- better wddump() support
- wdcwait() now takes a timout argument, so that we can honnor the
timeout argument of scsipi cmds.
|
| 1.9.2.1 | 04-Jun-1998 |
bouyer | Commit changes to the IDE system in a branch. This allows a better separation
between higth-level and low-level (i.e. registers read/write) and generalize
the queue for all commands. This also add supports for IDE DMA.
|
| 1.25.6.1 | 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.
|
| 1.25.4.1 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.25.2.2 | 14-Dec-2000 |
he | Pull up revision 1.31 (requested by bouyer):
Add a quirk entry for Hitachi CDR-7730. Fixes PR#11654.
|
| 1.25.2.1 | 08-Jul-1999 |
perry | pullup 1.25->1.26 (bouyer); last minute but it looked okay
|
| 1.28.6.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.28.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.28.2.10 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.28.2.9 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.28.2.8 | 18-Feb-2001 |
bouyer | Move atapiprint() to ic/wdc.c so that we can compile a kernel with IDE
controller but without atapibus.
|
| 1.28.2.7 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.28.2.6 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.28.2.5 | 04-Feb-2000 |
thorpej | Make sure the channel's completion thread exits when the channel
is detached, and completely encapsulate the way periph's are stored
by the channel.
|
| 1.28.2.4 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.28.2.3 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.28.2.2 | 20-Oct-1999 |
enami | Fix compilation error when ATAPI_DEBUG_PROBE is defined.
|
| 1.28.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.36.2.8 | 03-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.36.2.7 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.36.2.6 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.36.2.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.36.2.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.36.2.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.36.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.36.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.40.4.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.40.2.3 | 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
|
| 1.40.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.40.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.57.2.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.57.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.57.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.57.2.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.57.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.67.2.6 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.67.2.5 | 04-Feb-2008 |
yamt | sync with head.
|
| 1.67.2.4 | 21-Jan-2008 |
yamt | sync with head
|
| 1.67.2.3 | 07-Dec-2007 |
yamt | sync with head
|
| 1.67.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.67.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.69.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.69.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.69.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.69.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.69.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.70.10.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.70.10.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.70.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.72.30.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.72.30.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.72.30.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.72.24.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.72.24.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.72.22.3 | 08-Dec-2007 |
jmcneill | Rename pnp(9) -> pmf(9), as requested by many.
|
| 1.72.22.2 | 01-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.72.22.1 | 12-Nov-2007 |
joerg | Add null handlers for atapibus, it doesn't seem to need any magic.
|
| 1.73.4.1 | 11-Dec-2007 |
yamt | sync with head.
|
| 1.73.2.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.75.6.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.75.2.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.77.18.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.77.12.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.77.4.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.77.4.2 | 16-May-2009 |
yamt | sync with head
|
| 1.77.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.83.14.3 | 02-Jul-2012 |
jdc | Pull up revision 1.86 (requested by riastradh in ticket #373).
Take the kernel lock in atapibusdetach just like atapibuschilddet.
Fixes kassert in scsipi_lookup_periph when I press the power button
on one of my laptops (and maybe another one) to power it off.
ok mrg
|
| 1.83.14.2 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.83.14.1 | 09-Apr-2012 |
riz | Pull up following revision(s) (requested by chs in ticket #166):
sys/dev/scsipi/atapiconf.c: revision 1.84
take the kernel lock during detach of atapibus as well.
|
| 1.83.12.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.83.8.4 | 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")
|
| 1.83.8.3 | 30-Oct-2012 |
yamt | sync with head
|
| 1.83.8.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.83.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.86.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.86.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.86.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.87.6.3 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.87.6.2 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.87.6.1 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.88.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.90.8.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.91.22.3 | 03-Apr-2021 |
thorpej | Give config_attach() the tagged variadic argument treatment and
mechanically convert all call sites.
|
| 1.91.22.2 | 21-Mar-2021 |
thorpej | CFARG_IATTR usage audit:
If a device carries only one interface attribute, there is no need
to specify it when calling config_search(); that specification is
meant only to disambiguate which interface attribute (which is a
proxy for "what kind of attach args are being used") is having
children attached. cfparent_match() will take care of ensuring that
any potential children can attach to one of the parent's iterface
attributes, and if the parent only carries one, no disambiguation is
necessary.
|
| 1.91.22.1 | 20-Mar-2021 |
thorpej | The proliferation if config_search_*() and config_found_*() combinations
is a little absurd, so begin to tidy this up:
- Introduce a new cfarg_t enumerated type, that defines the types of
tag-value variadic arguments that can be passed to the various
config_*() functions (CFARG_SUBMATCH, CFARG_IATTR, and CFARG_LOCATORS,
for now, plus a CFARG_EOL sentinel).
- Collapse config_search_*() into config_search() that takes these
variadic arguments.
- Convert all call sites of config_search_*() to the new signature.
Noticed several incorrect usages along the way, which will be
audited in a future commit.
|
| 1.92.8.1 | 04-Aug-2021 |
thorpej | Adapt to CFARGS().
|
| 1.23 | 19-Oct-2009 |
bouyer | Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !
|
| 1.22 | 24-Mar-2008 |
cube | branches: 1.22.4;
Split device_t and softc for atapibus(4).
|
| 1.21 | 23-Nov-2006 |
he | branches: 1.21.48;
This file needs to know the size of "struct device", so include
<sys/device.h> here. This should allow the sparc GENERIC_SUN4U and
sparc64 kernels to build sata_subr.c again.
|
| 1.20 | 11-Dec-2005 |
christos | branches: 1.20.20; 1.20.22;
merge ktrace-lwp.
|
| 1.19 | 17-Sep-2004 |
mycroft | branches: 1.19.12;
Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.18 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.17 | 21-Aug-2004 |
thorpej | De-__P.
|
| 1.16 | 12-Aug-2004 |
thorpej | Protect against multiple inclusion.
|
| 1.15 | 04-Aug-2004 |
bouyer | Remove an unused prototype.
|
| 1.14 | 03-Dec-2001 |
bouyer | branches: 1.14.16; 1.14.18;
Update my copyrigth.
|
| 1.13 | 02-Dec-2001 |
bouyer | Clean up attach of wd/atapibus:
kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead
of ata_atapi_attach as attach arch. Create a ata_device, compatible with
scsipi_channel, to attach wd.
|
| 1.12 | 14-May-2001 |
bouyer | branches: 1.12.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.11 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.10 | 02-Apr-2000 |
augustss | branches: 1.10.6;
Let the device thet gets an atapibus attached specify how to kill pending
transfers by giving a function pointer.
The old method always called wdc specific code.
|
| 1.9 | 01-Apr-2000 |
bouyer | - DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op
if an IRQ was not detected, unless the force flag was given. Use this to
detect if the IRQ was for us (closer to shared IRQ for controllers which
don't have their own IRQ handler in pciide.c) and to poll for DMA xfer.
Also makes the timeout recovery code simpler.
- ATAPI cleanup: don't call controller-specific functions from atapiconf.c
(wdc_*), so that it's possible to attach an atapibus to something else
than a wdc/pciide (Hi Lennart :).
Overload struct scsi_adapter with struct atapi_adapter, defined
as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points
to an scsi_adapter, atapi code casts it to atapi_adapter if needed.
Move atapi_softc to atapiconf.h so that it can be used by the underlying
controller code (e.g. atapi_wdc.c).
Add an atapi-specific callback *atapi_probedev(), which probe a drive
in a controller-specific way, allocate the sc_link and fills in the
ataparams if needed. It then calls atapi_probedev() (from atapiconf.c)
to do the generic initialisations and attach the device.
- While I'm there merge and centralise the state definitions in atavar.h.
It should now be possible to use a common ata/atapi routine to set the
drive's modes (will do later).
|
| 1.8 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.7 | 12-Oct-1998 |
bouyer | branches: 1.7.6; 1.7.12; 1.7.14; 1.7.16;
Merge bouyer-ide
|
| 1.6 | 13-Feb-1998 |
enami | branches: 1.6.2;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.5 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.4 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file atapiconf.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: some spaces -> tabs, indenting
consistency in prototypes.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file atapiconf.h was added on branch marc-pcmcia on 1997-08-27 23:32:57 +0000
|
| 1.6.2.2 | 11-Jun-1998 |
bouyer | pciide: when setting up DMA mode, also enable fast timings for PIO modes
if possible. For ATAPI devices, commands are always send via PIO.
atapi: matain a drive state (like wd drives) and use it to set up the
timings if needed.
|
| 1.6.2.1 | 04-Jun-1998 |
bouyer | Commit changes to the IDE system in a branch. This allows a better separation
between higth-level and low-level (i.e. registers read/write) and generalize
the queue for all commands. This also add supports for IDE DMA.
|
| 1.7.16.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.7.14.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.7.12.4 | 21-Apr-2001 |
bouyer | Get rid of references to sc_link/scsipi_link in debug or comments.
|
| 1.7.12.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.7.12.2 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.7.12.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.7.6.1 | 07-Jul-2000 |
he | Apply patch (requested by bouyer):
Add support for the following PCIIDE controllers:
o AMD 756
o CMD PCI0648 and PCI0649
o Hightpoint HPT366
o OPTi 82c621 (and a few of its derivatives)
o Promise Ultra/33 and Ultra/66
o Intel 82801 (ICH/ICH0)
Also fix PR#10437 (detect more ATAPI devices).
|
| 1.10.6.2 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.10.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.12.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.14.18.1 | 11-Sep-2004 |
he | Pull up revision 1.18 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.14.16.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.14.16.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.14.16.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.14.16.1 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.19.12.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.20.22.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.20.20.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.21.48.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.22.4.1 | 11-Mar-2010 |
yamt | sync with head
|
| 1.356 | 01-Dec-2024 |
andvar | s/assised/assisted/ in comment.
|
| 1.355 | 22-Jun-2024 |
palle | branches: 1.355.2;
Add quirk for devices that does not handle READ_DISCINFO
|
| 1.354 | 26-Jun-2022 |
andvar | s/Ramdom/Random/ in comments.
|
| 1.353 | 12-Oct-2021 |
andvar | fix various typos, mainly in comments.
|
| 1.352 | 21-Aug-2021 |
andvar | fix mainly same typos as in my previous commit but outside sys/dev/dm.
|
| 1.351 | 16-Apr-2021 |
reinoud | Limit buffer size for device capabilities requests as a work-around for PR
kern/56109.
|
| 1.350 | 10-Feb-2021 |
christos | branches: 1.350.2;
PR/55986: Ryo Onodera: DK_BUSY must have mask as second argument. Make cd.c
consistent by also using __BIT()
|
| 1.349 | 26-Oct-2020 |
mlelstv | branches: 1.349.2;
Avoid buffer overflow when copying from bounce buffer.
Fixes PR 54810
Don't use uninitialized pointer in split bounce buffer case and
free a partially allocated bounce buffer on error.
|
| 1.348 | 29-Sep-2020 |
msaitoh | s/settng/setting/
|
| 1.347 | 29-Sep-2020 |
msaitoh | s/implicitely/implicitly/
|
| 1.346 | 29-Sep-2020 |
msaitoh | s/occurence/occurrence/
|
| 1.345 | 29-Sep-2020 |
msaitoh | s/parition/partition/
|
| 1.344 | 13-Apr-2020 |
maxv | constify
|
| 1.343 | 27-Mar-2020 |
mlelstv | branches: 1.343.2;
Allow open of RAWPART even when no medium is loaded.
Keep errors silent if no medium is loaded.
Fixes PR kern/55104
|
| 1.342 | 03-Sep-2018 |
riastradh | branches: 1.342.4;
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!)
|
| 1.341 | 17-Jun-2017 |
mlelstv | branches: 1.341.4; 1.341.6;
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.340 | 08-Apr-2017 |
mlelstv | branches: 1.340.4; 1.340.6;
ignore a failed scspi_prevent when accessing the RAW_PART as before.
|
| 1.339 | 05-Jan-2017 |
mlelstv | branches: 1.339.2;
Fix memory leak.
Found by maxv@
|
| 1.338 | 21-Dec-2016 |
mlelstv | really keep request queued after resource shortage
|
| 1.337 | 21-Dec-2016 |
mlelstv | cdrestart takes a softc parameter, not a periph.
|
| 1.336 | 16-Dec-2016 |
mlelstv | add comment about "missing" dk_start.
|
| 1.335 | 10-Dec-2016 |
mlelstv | appease gcc.
|
| 1.334 | 10-Dec-2016 |
mlelstv | Refactored sd and cd to use common disk subroutines.
|
| 1.333 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.332 | 20-Nov-2016 |
pgoyette | Avoid calling bufq_free() from critical code sections.
|
| 1.331 | 15-May-2016 |
reinoud | branches: 1.331.2;
Use _align(2) workaround for PR kern/51141. This fixes odd CD length
reporting.
|
| 1.330 | 26-Apr-2015 |
mlelstv | Use C99-style initializers for struct dkdriver.
|
| 1.329 | 13-Apr-2015 |
riastradh | Convert sys/dev to use <sys/rndsource.h>.
|
| 1.328 | 02-Jan-2015 |
christos | We have three sets of DTYPE_ constants in the kernel:
altq Drop Type
disklabel Disk Type
file Descriptor Type
(not to mention constants that contain the string DTYPE).
Let's make them two, by changing the disklabel one to be DisK TYPE since the
other disklabel constants seem to do that. Not many userland programs use
these constants (and the ones that they do are mostly in ifdefs). They will
be fixed shortly.
|
| 1.327 | 31-Dec-2014 |
christos | make more drivers use disk_ioctl, and add a dev parameter to it so that
we can merge the "easy" disklabel ioctls to it. Ultimately all this will
go do dk_ioctl once all the drivers have been converted.
|
| 1.326 | 31-Dec-2014 |
mlelstv | forgot this one. disk_set_info includes the functionality of disk_blocksize.
|
| 1.325 | 18-Oct-2014 |
snj | branches: 1.325.2;
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.324 | 03-Oct-2014 |
justin | Use uint32_t for blocksize as per other uses and to fix warnings
|
| 1.323 | 10-Aug-2014 |
tls | Merge tls-earlyentropy branch into HEAD.
|
| 1.322 | 25-Jul-2014 |
dholland | Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.321 | 25-Jul-2014 |
dholland | Add d_discard to all struct bdevsw instances I could find.
I've set them all to nodiscard. Some of them (wd, dk, vnd, ld,
raidframe, maybe cgd) should be implemented for real.
|
| 1.320 | 12-Jul-2014 |
hannken | Prevent detaching an open device unless forced.
|
| 1.319 | 18-Apr-2014 |
martin | Fix PR kern/48550 by aligning the single instance of scsipi_read_cd_cap_data
that we found misaligned in the wild so far properly for the ahcisata
driver. Also point at PR kern/48754 for the real issue.
|
| 1.318 | 19-Mar-2014 |
martin | branches: 1.318.2;
PR kern/48550: additional initialization and sanity checking on the reported
blocksize of the medium.
|
| 1.317 | 16-Mar-2014 |
dholland | 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.
|
| 1.316 | 25-Oct-2013 |
martin | Turn a few __unused into __diagused
|
| 1.315 | 19-Oct-2013 |
martin | Mark a potentially unused variable
|
| 1.314 | 28-Sep-2013 |
skrll | Don't return EPASSTHROUGH when succeeding in ODIOCGDINFO or ODIOCGDEFLABEL
|
| 1.313 | 12-Sep-2013 |
martin | Remove unused variables
|
| 1.312 | 02-Jul-2013 |
reinoud | Small patch to clarify last_lba and now using size confusingly. No functional
change.
|
| 1.311 | 29-May-2013 |
christos | branches: 1.311.2;
phase 1 of disk geometry cleanup:
- centralize the geometry -> plist code so that we don't have
n useless copies of it.
|
| 1.310 | 15-Mar-2013 |
martin | Prompted by PR kern/47646, zero the value buffer before running the
GET_CONFIGURATION command and check for an unexpectedly large feature
length answer afterwards.
|
| 1.309 | 06-May-2012 |
martin | branches: 1.309.2;
When ejecting a medium, invalidate the in core disklabel - it is not
meaningfull anymore. This makes the following cdclose() use silent
mode and finally fixes PR kern/43785.
|
| 1.308 | 06-May-2012 |
martin | In cdopen: make the test for media presence always silent. In all cases
requiring a message, we will most likely get that from the spinup attempt
anyway.
This avoids the spurious "Check Condition on CDB, Not Ready, Medium Not
Present, Tray Closed" messages at boot/shutdown time.
|
| 1.307 | 19-Apr-2012 |
bouyer | Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.306 | 25-Feb-2012 |
shattered | Fix typos in comments.
OK by wiz@
|
| 1.305 | 02-Feb-2012 |
tls | branches: 1.305.2;
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.
|
| 1.304 | 25-Nov-2011 |
joerg | branches: 1.304.2;
Replace memset with incorrect size argument with M_ZERO.
|
| 1.303 | 20-May-2011 |
reinoud | branches: 1.303.4;
Fix size reporting for DVD+R/DL and BluRay's; the value was trunced due to a
32 bit trunc due to a lacking type cast. The number of sectors for such media
can be more than 1<<32-1.
|
| 1.302 | 04-Apr-2010 |
martin | branches: 1.302.2;
cd_size: if we fake a size (and I realy have no idea why this would be a
good idea), at least set up all values to the fake values, as the caller
expects.
Should fix PR kern/39904, though if noone can find out why the fake value
would be needed, we should change it to just return 0 as suggested in the PR.
|
| 1.301 | 23-Mar-2010 |
martin | If we find a device in non-2048 byte/block mode and manage to switch
it over, we need to adjust the capacity - so just read it again.
|
| 1.300 | 23-Mar-2010 |
martin | Make cdclose() silent, if we only have the raw partition open (e.g. when
probing for media and not finding any)
|
| 1.299 | 22-Mar-2010 |
martin | Add a DIOCTUR (test unit ready) ioctl, to allow userland to easily
(and silently) check for presence of media. While this would be easy
to implement via SCIOCCOMMAND, our kernel has quirks for this
command, which otherwise would not be readily available to userland
applications.
|
| 1.298 | 11-Mar-2010 |
mrg | branches: 1.298.2;
various aprint_* fixes.
|
| 1.297 | 06-Jan-2010 |
martin | branches: 1.297.2;
Some usb devices come with an internal emulated umass CD drive (containing
windows drivers). I have such a device that has 0 features - avoid wrappig
features_len to a very big unsiged 32bit number in this case.
|
| 1.296 | 06-Dec-2009 |
dyoung | Delete do-nothing device-activation hooks.
|
| 1.295 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.294 | 26-Jul-2009 |
reinoud | Significantly bump time for SCSI track/disc closure. DVD-RW when used
sequentially can significantly take more time.
|
| 1.293 | 05-Jun-2009 |
haad | Add support for DIOCGDISKINFO to disk like device drivers. Change
partutil.c::getdiskinfo to use it to get disk geometry info.
Use DIOCGWEDGEINFO ioctl to get information about partition size, if disk
driver doesn't support it use old DIOCGDINFO. This patch adds support for
wedge like devices(lvm logical volumes, ZFS zvol partitions) to newfs and
other tools.
No objections on tech-userlevel@.
|
| 1.292 | 07-Apr-2009 |
dyoung | Detach atapibus(4), scsibus(4), cd(4), and sd(4) during shutdown.
Destroy sd->sc_callout in sddetach(). Delete some dead code in
cddetach().
|
| 1.291 | 01-Apr-2009 |
reinoud | Implement MMC_TRACKINFO_DATA and MMC_TRACKINFO_AUDIO detection for cdrom and
dvdrom devices. These two compatibility codes were forgotten when those were
introduced.
|
| 1.290 | 18-Mar-2009 |
cegger | bzero -> memset
|
| 1.289 | 17-Mar-2009 |
reinoud | Fix long standing problem with CD/DVD writing with `dd'. If done this way the
device will remain in a wait-for-more-stuff-to-be-written state delaying other
SCSI calls, esp. some read-size or read-layout calls, until its told there is
no more comming.
To fix this, on last close explicitly flush the caches releasing the device
from this mode and allowing other SCSI commands to reach it.
This fixes PR 40528
|
| 1.288 | 14-Mar-2009 |
ad | 'boot -z' bogons
|
| 1.287 | 21-Jan-2009 |
cegger | branches: 1.287.2;
buildfix: re-adapt to major()/minor() returning a 32bit value.
|
| 1.286 | 13-Jan-2009 |
yamt | g/c BUFQ_FOO() macros and use bufq_foo() directly.
|
| 1.285 | 11-Jan-2009 |
cegger | make this compile
|
| 1.284 | 30-Dec-2008 |
reinoud | Add ISO partition detection enabling auto-detection of iso9660 and UDF
partitions on optical media like CD/DVD/BD but also on all other media if
there is no NetBSD disklabel or MBR label.
Also fix cd's readdisklabel arguments so the ioctl's arrive at the right
device (!) and update its default label to make more sense.
|
| 1.283 | 05-Sep-2008 |
gmcgarry | branches: 1.283.2; 1.283.4;
Remove unreachable code introduced by rev 1.224. Fixes PR#38973.
|
| 1.282 | 12-Jun-2008 |
cegger | branches: 1.282.2;
use device_lookup_private to get softc
|
| 1.281 | 12-May-2008 |
jnemeth | branches: 1.281.2;
add support for drvctl properties
|
| 1.280 | 12-May-2008 |
tron | Fix SSP kernel builds.
|
| 1.279 | 08-May-2008 |
reinoud | Implement write support for the MMC framework. This consists of a setting
up of write parameters call and a call to perform a series of operations on
these devices.
Note that the MMC framework interface is still not exposed to userland
unless the expose_mmc flag is set. This is to prevent applications to build
on a possibly still moving target. This flag will eventually be removed.
|
| 1.278 | 06-May-2008 |
yamt | branches: 1.278.2;
getiobuf(false, NULL) -> getiobuf(NULL, false)
|
| 1.277 | 02-May-2008 |
reinoud | Extend the MMC framework to also create sensible reports on audio discs.
|
| 1.276 | 02-May-2008 |
reinoud | Fixup CD and DVD drives that appear to be on crack when reporting features.
They state f.e. that a recordable CD-R is rewritable or completely forget
that the fixed packet size formatted CD-RW is strict overwrite rewritable
and not randomly rewritable.
|
| 1.275 | 02-May-2008 |
reinoud | Make the SCSI sense code "Logical Unit Not Ready, Operation In Progress"
(Sense code 0x04, 0x07) not a fatal error anymore but retry command later.
This sense code can be returned on a CD-MRW write that encounters a bad
block resulting in the drive being busy relocating it. During that time it
can return this sense code to indicate its busy for a while.
Also reduce waiting time for the other "Logical Unix Not Ready, Long Write
In Progress" to half a second.
|
| 1.274 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.273 | 24-Mar-2008 |
cube | branches: 1.273.2; 1.273.4;
Split device_t for cd(4).
|
| 1.272 | 02-Jan-2008 |
ad | branches: 1.272.6;
Merge vmlocking2 to head.
|
| 1.271 | 09-Dec-2007 |
jmcneill | branches: 1.271.2;
Merge jmcneill-pm branch.
|
| 1.270 | 27-Nov-2007 |
reinoud | branches: 1.270.2; 1.270.4;
Pullup fixes from UDF write development project. This patch fixes a
conceptional bug in mmc_discinfo, improves DVD+R feature detection and
prevents corruption of mmc_trackinfo readout on some devices that return
short trackinfo structures.
|
| 1.269 | 08-Oct-2007 |
ad | branches: 1.269.4;
Merge disk init changes from the vmlocking branch. These seperate init /
destroy of 'struct disk' from attach / detach.
|
| 1.268 | 04-Aug-2007 |
rumble | branches: 1.268.2; 1.268.4; 1.268.6;
Be sure not to make requests over MAXPHYS when employing bounce buffers.
It was previously possible to round up one sector length if the start block
were not properly aligned.
Resolves PR kern/36716. This may also fix the following reported issues
on port-sgimips:
http://mail-index.netbsd.org/port-sgimips/2005/09/12/0000.html
http://mail-index.netbsd.org/port-sgimips/2005/09/14/0000.html
While here, be sure to put back allocated bufs, which previously appear
to have been leaked. Further, remove the dead write bounce code which never
got executed anyway.
No complaints on tech-kern.
|
| 1.267 | 29-Jul-2007 |
ad | branches: 1.267.4;
It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.266 | 21-Jul-2007 |
ad | Replace some uses of lockmgr().
|
| 1.265 | 09-Jul-2007 |
ad | branches: 1.265.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.264 | 30-Jun-2007 |
dsl | The cd ioctl definitions are not condusive to being called from within
the kernel (eg from withthe linux compat code).
Add CDIOCREADSUBCHANNEL_BUF and CDIOREADTOCENTRIES_BUF which are like the
existing ioctls withoutthe _BUF, except that the data area immediately
follows the control structure (and is fixed length).
Modify scsipi/cd.c so that it only ever has one 'struct cd_formatted_toc'
allocated on the stack (maybe this info ought to be in the softc)
structure since even one 800 byte on-stack buffer isn't overly friendly.
cdplay(1) still seems able to read the track-list of an ATAPI cd.
|
| 1.263 | 30-Jun-2007 |
dsl | Remove a load of unnecessary casts now that 'addr' is 'void *'.
|
| 1.262 | 04-Mar-2007 |
christos | branches: 1.262.2; 1.262.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.261 | 14-Jan-2007 |
martin | branches: 1.261.2;
The previous fix for PR 34202 was wrong (some sizeof used the union, not
the command block) - but instead of fixing this, we can now revert it
completley: the real fix has been applied to ../ic/wdc.c in rev. 1.244,
removing the alignement constraints.
|
| 1.260 | 01-Dec-2006 |
martin | branches: 1.260.2;
Fix PR kern/34202 differently, by aligning the variables "the traditional
way".
|
| 1.259 | 25-Nov-2006 |
scw | - Call disk_blocksize(9) when we determine the physical block size of the
media. (Will this ever be anything other than 2048?)
- bounds_check_with_label() has been fixed, so put the correct partition
size in the default disklabel.
|
| 1.258 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.257 | 14-Nov-2006 |
reinoud | Implement/add bufq strategy setting by dkctl for the CD class devices.
Though its hardly advisable to change the CD strategy setting to anything
other than `disksort' or `prioscan' it ought to be possible.
|
| 1.256 | 28-Oct-2006 |
reinoud | Implement another class of `unit not ready' sense handling that is not
fatal.. A `long write in progress' is a retry again later command that is
issued when a device returns immediately after a write request but needs
some time before it can handle read requests.
|
| 1.255 | 27-Oct-2006 |
christos | - don't allocate huge arrays on the stack
- no bogus ; after }; in block statements!
|
| 1.254 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.253 | 10-Oct-2006 |
riz | Also print blksize with %u - this is what I get for eyeballing
a patch instead of applying it directly. PR#33966.
|
| 1.252 | 10-Oct-2006 |
riz | Print an unsigned parameter with %lu, not %ld. From David A. Holland
in PR#33966.
|
| 1.251 | 08-Oct-2006 |
mlelstv | Use last track info only if it gives a sane value. Fixes PR#34688.
|
| 1.250 | 08-Sep-2006 |
reinoud | branches: 1.250.2;
Oeps! I overlooked the size argument of the SCSI call. It ought to be equal
to the number of bytes we are reading: READ_DISCINFO_BIGSIZE
sorry.
|
| 1.249 | 07-Sep-2006 |
reinoud | The size of struct scsipi_read_discinfo_data is not even due to the data[1]
pseudo array in the structure. The scsi command issued used
sizeof(scsipi_read_discinfo_data) but included thus the last byte that
wasn't going to be inspected anyway. Using the constant
READ_DISCINFO_BIGSIZE fixes the bug.
Even though SCSI adapters *should* accept odd lengths, the Sun U10
(sparc64's) crashes in bus_space_read_multi_stream_2(). That explains the
crash.
NOTE: somewhere in NetBSD/sparc64's atapibus support there is the
assumption on the even size; this needs to be fixed! (Hi Sparc64 portmaster
:-) )
|
| 1.248 | 01-Sep-2006 |
matt | branches: 1.248.2;
Use an unsigned int for the blksize on read_cd_capacity.
|
| 1.247 | 31-Aug-2006 |
reinoud | Fix panic on reading/writing to a (raw) CD device and encountering an
error. It now correctly sets bp->b_resid to the full size of the buffer.
The failed SCSI read/write command allways contains the complete buffer.
When encountering a read (or write) error, the scsipi stack sets the
xs->resid to zero since the command has been accepted by the drive. The
scsi command set does not allow a partial read or write to be performed and
will signal a success or an error.
|
| 1.246 | 31-Aug-2006 |
reinoud | Remove two spurious empty lines in functions
|
| 1.245 | 31-Aug-2006 |
reinoud | Fix fencing bug on CD disk size; the comment was wrong, its the *length* of
the device and since we count from zero its equal to the number of sectors
and thus one higher than the last sector.
This fixes the read/write problems on the (raw) device where the last
sector was not readable/writable.
|
| 1.244 | 28-Aug-2006 |
christos | Add missing initializer.
|
| 1.243 | 10-Aug-2006 |
reinoud | Fix incorrect `loast possible lba' reporting, remove some empty lines and
add mmc classification for some still rare HD DVD device types.
|
| 1.242 | 30-Mar-2006 |
thorpej | branches: 1.242.6;
Use device_private().
|
| 1.241 | 28-Mar-2006 |
thorpej | Use device_unit().
|
| 1.240 | 08-Feb-2006 |
reinoud | branches: 1.240.2; 1.240.4; 1.240.6;
Add name of the processed feature like the others have.
|
| 1.239 | 04-Feb-2006 |
reinoud | Add support for SCSI MMC feature 0x0024: hardware assisted defect
management to be passed in mmc_discinfo's device capabilities.
When a device reports support for this the hardware will automatically
remap sectors on read- or write-errors. CD-MRW, DVD+MRW, DVD-RAM and BR-RE
support these.
|
| 1.238 | 02-Feb-2006 |
reinoud | branches: 1.238.2;
Adding two Ioctl's to scsipi's cd.c to abstract SCSI MMC devices. The
ioctl's and their structures are currenly hidden from inclusion for normal
userland applications to allow the MMC abstraction interface to mature
first.
Its useage is mainly for applications dedicated to handling and processing
SCSI MMC compliant devices and their media. Examples of these are CDs, DVDs
and comparable optical devices but also some magnetic swapable devices that
present themselves as a SCSI MMC compliant device.
Its current use is currently the to be committed in-kernel UDF filingsystem
and its userland tools.
|
| 1.237 | 07-Jan-2006 |
christos | branches: 1.237.2;
fix typo.
|
| 1.236 | 07-Jan-2006 |
yamt | remove B_EINTR as it isn't used anymore.
|
| 1.235 | 04-Jan-2006 |
yamt | - add simple functions to allocate/free a buffer for i/o.
- make bufpool static.
|
| 1.234 | 21-Dec-2005 |
reinoud | branches: 1.234.2;
Improve disc size reporting. The current discsize thats reported is the
current recorded extent and not the complete maximum extent of the disc.
An empty CD/DVD disc is seen as having an invalid length and given 800 Mb
size. This makes it possible to write to a CD-R using `dd' as some tend to
do.
However when an extent is recorded the initial recorded length is seen as
valid and is thus set as absolute upperbound to read and write actions
preventing extention of the disc.
Committed after consultation on tech-kern with a positive advice by Manuel
Bouyer.
|
| 1.233 | 16-Dec-2005 |
christos | delete extraneous verbiage.
|
| 1.232 | 15-Dec-2005 |
reinoud | Clean up cdgetdisklabel. It was jumping over a return...
|
| 1.231 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.230 | 11-Dec-2005 |
reinoud | Fix illogical use of extra variable. It was mainly to get a better
readability in KNF. This solution gives better readability without the
extra variable.
|
| 1.229 | 10-Dec-2005 |
reinoud | Second stage of read TOC modifications. Introducing the various forms of
TOC reading. Note that the external interfaces haven't changed. Only the
formatted toc is requestable.
The read msinfo command hasn't been changed in this patch though it could
become more clever using the CD_TOC_MSINFO format.
|
| 1.228 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.227 | 06-Sep-2005 |
reinoud | Cleanup and fix cd_size() function; it was returning wrong results.
Also factor out read_cd_capacity()
|
| 1.226 | 05-Sep-2005 |
reinoud | Implement DIOCCACHESYNC for SCSI/ATAPI cd/mmc recordable devices found by
scsipi/cd.c by issueing the SYNCHRONISE CACHES scsi/atapi call as defined
per MMC standard.
|
| 1.225 | 28-Aug-2005 |
reinoud | Document executed SCSI commands with the IOCTL's to prevent confusion.
|
| 1.224 | 06-Jul-2005 |
bouyer | Always allow open() on the raw partition to succeed, even if any of the
SCSI command in cdopen() fails. Fix PR kern/30288 by Pavel Cahyna.
|
| 1.223 | 29-May-2005 |
christos | branches: 1.223.2;
- Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.222 | 25-Apr-2005 |
drochner | fix more SSD_RCODE_VALID misuse introduced in cleanup
|
| 1.221 | 31-Mar-2005 |
yamt | introduce a function to drain bufq and use it where appropriate.
|
| 1.220 | 27-Feb-2005 |
perry | branches: 1.220.2;
nuke trailing whitespace
|
| 1.219 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.218 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.217 | 31-Jan-2005 |
reinoud | Fix LP64 problems introduced by my u_long->uint32_t conversion.
|
| 1.216 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.215 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.214 | 30-Jan-2005 |
reinoud | As in revision 1.213, not all information is returned in read-TOC format 0. In order to add a new generic read-toc ioctl, i renamed the origional to _f0 to indicate its only format 0. The routines can then be changed one at a time to the new generic calling. This patch fixes two calls i'd overseen.... silly me forgot to recompile :(
|
| 1.213 | 30-Jan-2005 |
reinoud | The code asumes all programs want to know is TOC response format 0. Support
for TOC response format 1 and 2 are mandatory on CD/DVD too and provide
more information.
Next an IOCTL needs to be implemented that can read all TOC formats in a
generic way. This is pending.
|
| 1.212 | 07-Dec-2004 |
thorpej | branches: 1.212.2; 1.212.4;
USe more appropriate macro/struct names for READ/WRITE (6) and
READ/WRITE (10).
|
| 1.211 | 28-Oct-2004 |
yamt | move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.210 | 26-Sep-2004 |
dogcow | Fix debug message output args to match yamt's src/sys/sys/buf.h changes.
|
| 1.209 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.208 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.207 | 17-Sep-2004 |
mycroft | In places where we've already called scsipi_make_xs(), call scsipi_execute_xs()
directly rather than going through scsipi_command().
|
| 1.206 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.205 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.204 | 06-Sep-2004 |
bouyer | Fix comment: xxstart() can also be called from xxrestart()
|
| 1.203 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.202 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.201 | 24-Apr-2004 |
pk | Some older devices do not understand the `disable block descriptor' bit in
the mode sense request. So fall back on mode sense data including a block
descriptor section.
See also sd.c rev. 1.214. Again: should we bother trying DBD at all?
|
| 1.200 | 22-Feb-2004 |
enami | branches: 1.200.2;
Backout an obvious bug introduced in rev. 1.145.
|
| 1.199 | 10-Jan-2004 |
yamt | store a i/o priority hint in struct buf for buffer queue discipline.
|
| 1.198 | 10-Nov-2003 |
wiz | Spell address with two d's. Inspired by similar changes in OpenBSD,
originating from Jonathon Gray and forwarded by jmc@openbsd.
|
| 1.197 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.196 | 13-Sep-2003 |
mycroft | Don't be silent if there is no media present.
|
| 1.195 | 13-Sep-2003 |
mycroft | Make sure the "raw partition" can always be opened again.
|
| 1.194 | 09-Sep-2003 |
mycroft | In the test for whether to start the unit, I used the wrong error code.
ENODEV is only returned when we get "medium not present," which we can fail
immediately on. All other "not ready" cases return EIO.
|
| 1.193 | 08-Sep-2003 |
mycroft | Oops. Fix a typo in cd_mode_select().
|
| 1.192 | 08-Sep-2003 |
mycroft | Do a START UNIT only if the TEST UNIT READY reports that the device is not
ready. This avoids gratuitously starting the motor on floppy and CD-ROM
drives, and eliminates the need for the audio playing test in cdopen().
Therefore, also remove PQUIRK_NOSTARTUNIT.
|
| 1.191 | 08-Sep-2003 |
mycroft | Attempt to deal with Martin's weirdass Sun drive by setting the mode sense
allocation length a little more precisely -- add the space for the header in
cd_mode_sense(). Also delete the XS_CTL_SILENT, since we really do want to
see errors.
Lastly, add a similar wrapper for mode select, simplifying the callers
slightly.
|
| 1.190 | 08-Sep-2003 |
mycroft | Eliminate the separate ATAPI and SCSI attachments for "cd".
|
| 1.189 | 08-Sep-2003 |
mycroft | Nuke some printf()s.
|
| 1.188 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.187 | 18-Jul-2003 |
wiz | Add a cast to avoid an integer overflow.
Fixes playing (at least some) DVDs.
Patch from Tsubai Masanari (tsubai at iri.co.jp) -- thanks a lot!
|
| 1.186 | 10-Jul-2003 |
martin | If a CD drive reports != 2048 byte block size, try to change it into
2048 byte mode. Fixes PR kern/22090.
Reviewed by Manuel Bouyer.
|
| 1.185 | 29-Jun-2003 |
fvdl | branches: 1.185.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.184 | 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
|
| 1.183 | 10-May-2003 |
thorpej | Change bounds_check_with_label() to take a pointer to the disk structure,
rather than the label itself. This paves the way for some future changes.
|
| 1.182 | 02-May-2003 |
dsl | Change return type of readdisklabel() to const char *
I hope I've found all the correct places!
|
| 1.181 | 03-Apr-2003 |
fvdl | Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.180 | 20-Mar-2003 |
dbj | use PRId64 to printf bp->b_blkno, which is of type daddr_t
|
| 1.179 | 25-Feb-2003 |
thorpej | Add a new BUF_INIT() macro which initializes b_dep and b_interlock, and
use it. This fixes a few places where either b_dep or b_interlock were
not properly initialized.
|
| 1.178 | 05-Feb-2003 |
pk | Make the buffer cache code MP-safe.
|
| 1.177 | 03-Feb-2003 |
thorpej | Test callout_pending(), not callout_active(), and eliminate now-unnecessary
callout_deactivate() calls.
|
| 1.176 | 23-Jan-2003 |
bad | Being able to make image backups of your whole disk, not to mention not
causing certain ata disks to lock up by reading beyond the last block,
beats catering to broken devices.
bounds_check_with_label() RAW_PART too.
|
| 1.175 | 20-Jan-2003 |
simonb | The Double-Semi-Colon Police.
|
| 1.174 | 13-Jan-2003 |
toshii | Use the correct byte positions to read data returned by DVD_LU_SEND_RPC_STATE.
|
| 1.173 | 15-Dec-2002 |
jmcneill | Another sizeof -> 4 + 2048 fix for dvd_read_manufact after malloc changes
|
| 1.172 | 15-Dec-2002 |
jmcneill | In dvd_read_disckey, sizeof(buf) is no longer '4 + 2048' after the
malloc change. Replace 'sizeof(buf)' with '4 + 2048' -- this makes
dvd_read_disckey work properly again.
|
| 1.171 | 15-Dec-2002 |
fvdl | Don't waste too much kernel stack space on (infrequent) ioctl operations,
use malloc instead for temp space.
|
| 1.170 | 09-Nov-2002 |
thorpej | Fix signed/unsigned comparison warnings.
|
| 1.169 | 01-Nov-2002 |
mrg | implement separate read/write disk statistics:
- disk_unbusy() gets a new parameter to tell the IO direction.
- struct disk_sysctl gets 4 new members for read/write bytes/transfers.
when processing hw.diskstats, add the read&write bytes/transfers for
the old combined stats to attempt to keep backwards compatibility.
unfortunately, due to multiple bugs, this will cause new kernels and old
vmstat/iostat/systat programs to fail. however, the next time this is
change it will not fail again.
this is just the kernel portion.
|
| 1.168 | 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
|
| 1.167 | 18-Sep-2002 |
chs | remove all vesitages of dk_establish().
|
| 1.166 | 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.
|
| 1.165 | 30-Aug-2002 |
hannken | Remove the old device buffer queue interface.
Approved by: Jason R. Thorpe <thorpej@wasabisystems.com>
|
| 1.164 | 22-Jul-2002 |
hannken | Convert to new device buffer queue interface.
|
| 1.163 | 27-May-2002 |
drochner | put multisession offset code into a separate function and initialize
the p_cdsession field of partition 'a' in the default disklabel
|
| 1.162 | 05-May-2002 |
bouyer | branches: 1.162.2;
If periph->periph_callout is already active, don't freeze the periph again:
scispi_periph_timed_thaw() will be called only one time anyway.
|
| 1.161 | 09-Dec-2001 |
veego | Support for dvd region code (RPC).
|
| 1.160 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.159 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.158 | 22-Oct-2001 |
bouyer | We can't have XS_CTL_DATA_OUT and XS_CTL_DATA_IN at the same time.
Pointed out by Bernd Ernesti.
|
| 1.157 | 02-Sep-2001 |
tsutsui | branches: 1.157.2;
Use be16toh(), be32toh(), le16toh() and le32toh() rather than
ntohs(), ntohl() or homegrown bswap() functions.
XXX Does anyone use drives with PQUIRK_LITTLETOC on big endian machines?
|
| 1.156 | 20-Aug-2001 |
ad | Don't special case the retry for Media Changed... More later.
|
| 1.155 | 20-Aug-2001 |
ad | Two changes submitted by Sergey Svishchev <svs@ropnet.ru>:
- Don't send START UNIT upon open if the drive is playing audio (PR 11768).
- If we receive sense indicating that the media changed, retry (PR 8326).
|
| 1.154 | 15-Aug-2001 |
eeh | Add support for smaller sector sizes so we can mount ffs filesystems.
|
| 1.153 | 18-Jul-2001 |
thorpej | bcopy -> memcpy
|
| 1.152 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.151 | 26-Jun-2001 |
bouyer | branches: 1.151.2;
Add a XS_CTL_SILENT_NODEV flag: if the sense info is "not ready, medium
not present" don't print any message but still return ENODEV.
Use this in cd driver to allow open of character raw partition even if
the drive is empty (older drives fails at LOAD_UNLOAD time, newer
ones fail at TEST_UNIT_READY time).
|
| 1.150 | 20-May-2001 |
christos | PR/12991: Dave Huang: DVD ioctls don't work because we send the wrong size
CDB (16 instead of 12).
|
| 1.149 | 17-May-2001 |
bouyer | SDEV_DB -> SCSIPI_DB
|
| 1.148 | 06-May-2001 |
drochner | Since __HAVE_OLD_DISKLABEL, DISKMINOR(u, p) isn't continuous over "p"
anymore. So we can't pass a range to vdevgone() but have to loop over the
partitions explicitely.
(otherwise the next device unit will be killed)
|
| 1.147 | 28-Apr-2001 |
tsutsui | Don't forget to set XS_CTL_DATA_OUT (that was removed in previous).
|
| 1.146 | 28-Apr-2001 |
thorpej | Use simple tags for read and write, unless B_ORDERED is set, in
which case we use an ordered tag.
|
| 1.145 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.144 | 19-Jan-2001 |
kenh | branches: 1.144.2;
Retry requests if the CD-ROM drive returns the sense code "Device in
Process of Becoming Ready".
|
| 1.143 | 08-Jan-2001 |
fvdl | Return error in the case of using ODIOCGDINFO or ODIOCGDEFLABEL when
the number of partitions is > OLDMAXPARTITIONS. This is better
than silently truncating the label (don't want to silently throw
away partitions when using an old disklabel binary on a label with
> 8 partitions). From Enami Tsugutomo.
|
| 1.142 | 07-Jan-2001 |
fvdl | Adapt all disk devices in MI directories to handle ODIOC* calls
for ports that have bumped MAXPARTITIONS (and thus define
__HAVE_OLD_DISKLABEL).
|
| 1.141 | 09-Jun-2000 |
enami | branches: 1.141.2;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.140 | 30-May-2000 |
augustss | Add a quirk, SDEV_ONLYBIG, which implies that the device cannot handle
the 6 byte versions of READ, WRITE, and MODE_SENSE.
This greatly simplifies the UFI (USB Floppy) handling.
|
| 1.139 | 16-May-2000 |
thorpej | branches: 1.139.2;
Nuke dk_establish() from orbit except from those ports which still use
it to determine the boot device: mvme68k, pc532, macppc, ofppc. Those
platforms should be changed to use device_register(). In the mean time,
those ports defined __BROKEN_DK_ESTABLISH.
|
| 1.138 | 30-Mar-2000 |
augustss | Get rid of register declarations.
|
| 1.137 | 13-Mar-2000 |
soren | Fix doubled 'the's in comments.
|
| 1.136 | 07-Feb-2000 |
thorpej | Fix a bug in disksort_*() which caused non-optimal ordering when multiple
active partitions were on a single spindle. Add a b_rawblkno member to
struct buf which contains the non-partition-relative block number to sort
by.
|
| 1.135 | 21-Jan-2000 |
thorpej | Update for sys/buf.h/disksort_*() changes.
|
| 1.134 | 03-Nov-1999 |
matt | use __vax__ and __i386__ instead of vax and i386
|
| 1.133 | 31-Oct-1999 |
mycroft | Fix the size of start_sector, end_sector and end_sector_l0. (I'm just going to
assume the Linux people will also fix this, since I reported it to them.)
|
| 1.132 | 29-Oct-1999 |
mycroft | Emulate the Linux DVD_* ioctls(2). This gets us 90% of the way to running the
LiViD DVD player. (See forthcoming mail to current-users.)
XXX NOTE: We should do something to probe capabilities, rather than allowing
these ioctls on any device.
|
| 1.131 | 17-Oct-1999 |
ragge | branches: 1.131.2; 1.131.4;
Don't call dk_establish() on vax either.
|
| 1.130 | 30-Sep-1999 |
thorpej | branches: 1.130.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.129 | 23-Sep-1999 |
enami | Allow to detach wdc, atapibus, wd and cd.
|
| 1.128 | 07-Aug-1999 |
mycroft | Increase several timeouts to 30s.
|
| 1.127 | 27-May-1999 |
bouyer | in cdclose(), ignore "media change" when calling scsipi_prevent(), after
a 'eject -l' some ATAPI drives report this instead of illegal request.
|
| 1.126 | 08-Apr-1999 |
bouyer | Add SCSI_IGNORE_NOT_READY to the flags for CDIOCCLOSE: some devices will
attempt to spin up, and report an error if there's no CD. Fixes port-i386/7345.
|
| 1.125 | 05-Apr-1999 |
mycroft | If scsipi_command() fails, always print out the error code.
|
| 1.124 | 28-Feb-1999 |
explorer | branches: 1.124.2;
Update to slightly altered rnd_attach_source() api
|
| 1.123 | 15-Feb-1999 |
bouyer | For CDIOCCLOSE, ignore media changes, as some drive seems to return this
info here.
|
| 1.122 | 10-Feb-1999 |
bouyer | Abort transfer if b_blkno is negative. Closes PR kern/5553 by Johan Danielsson.
|
| 1.121 | 08-Feb-1999 |
bouyer | Change DIOCEJECT to do what's needed to eject a device before the eject
command (unlock for sd and cd) if no other partitions are open, return
EBUSY otherwise. DIOCEJECT will have the old semantic if its argument is not
0. The old ioctl has been renamed to ODIOCEJECT for binary compatibility.
|
| 1.120 | 29-Jan-1999 |
bouyer | Return ENODEV instead of EIO when we are trying to open a device without media
in the drive. restrict "opening of empty drive" to character devices only
(reading a block device returns a short read instead of ENODEV, which can lead
to confusion).
|
| 1.119 | 26-Jan-1999 |
bouyer | Allows the raw partition to be open()'ed, even when scsipi_start() fails
(no media or other ...) so that we can always send ioctl's to the device.
|
| 1.118 | 04-Jan-1999 |
is | Multisession-CD-Support, by Torsten Duwe <duwe@ns.lst.de>.
|
| 1.117 | 08-Dec-1998 |
thorpej | When closing, wait for pending xfers to drain before unlocking the door,
and wait again before deleting the reference to the adapter.
|
| 1.116 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.115 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.114 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.113 | 13-Jul-1998 |
hpeyerl | Add support for ATA CD changer devices like the NEC CDR-251.
Evidence of managerial coding removed by Victor T. Cleaner (thorpej)
|
| 1.112 | 16-Mar-1998 |
mycroft | Implement CDIOCCLOSE.
|
| 1.111 | 15-Jan-1998 |
cgd | move the 'cd_cd' declaration back up to where it was before
the config changes were made. Though the nature of the declaration
had to change, there wasn't a reason to change its location.
|
| 1.110 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.109 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.108 | 02-Dec-1997 |
mikel | fix typo; from Dave Sainty in PR kern/4602
|
| 1.107 | 18-Oct-1997 |
thorpej | branches: 1.107.2;
Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.106 | 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.
|
| 1.105 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.104 | 08-Oct-1997 |
thorpej | Implement DIOCGDEFLABEL.
|
| 1.103 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.102 | 09-Sep-1997 |
bouyer | Uses PLAY_MSF instead of PLAY for the CDIOCPLAYTRACKS ioctl. This avoid a int16
overflow in the PLAY cbd when the resquested track(s) are more than 65535
blocks long. Submitted by t-nkyma@tcp-ip.or.jp in PR kern/4092.
|
| 1.101 | 27-Aug-1997 |
bouyer | branches: 1.101.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.100 | 02-Apr-1997 |
mycroft | Push the buffer cleanup code into scsi_done(), and split it so that biodone()
is called *after* the driver `done' routine. This fixes disk I/O statistics
on SCSI devices.
Also, calling the `done' routine with a `complete' argument of 0 and actually
having it do anything meaningful loses in at least 3 ways, so just nuke the
argument altogether and don't call it this way. If the driver needs to do
some error handling, that's what `err_handler' is for.
|
| 1.99 | 29-Mar-1997 |
christos | PR/3409: Koji Imada: cdsize() timeout too short for cd changers.
|
| 1.98 | 21-Feb-1997 |
thorpej | If posting an error condition because the media has been unloaded,
make sure to set the residual count to reflect that no data was
transfered.
From Naofumi HONDA / MINOURA Makoto, PR #3007.
|
| 1.97 | 05-Dec-1996 |
cgd | branches: 1.97.6;
update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.96 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.95 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.94 | 13-Aug-1996 |
explorer | be quiet about read_subchannel. Some CD players use this to poll, and having
a screen filled with kernel messages isn't nice. Closes pr kern/817
|
| 1.93 | 13-Aug-1996 |
explorer | Fix condition where disk_unbusy may not get called
|
| 1.92 | 05-May-1996 |
christos | Cleanup the rest of the SCSIDEBUG printfs. From Bernd Ernesti.
|
| 1.91 | 22-Apr-1996 |
christos | remove include of <sys/cpu.h>
|
| 1.90 | 30-Mar-1996 |
christos | Eliminate scsi_conf.h.
|
| 1.89 | 29-Mar-1996 |
mrg | eliminate unused variables.
|
| 1.88 | 27-Mar-1996 |
cgd | fix pasto: sdminphys -> cdminphys in comment.
|
| 1.87 | 27-Mar-1996 |
mycroft | SDUNIT -> CDUNIT in last.
|
| 1.86 | 26-Mar-1996 |
mycroft | Put back dk_establish() for now.
|
| 1.85 | 26-Mar-1996 |
mycroft | Increase *OUTSTANDING.
Remove old dk_establish() garbage.
Support the `ancient' bit in both drivers.
|
| 1.84 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.83 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.82 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.81 | 30-Jan-1996 |
thorpej | Accept DIOCEJECT as a synonym for CDIOCEJECT. Implement DIOCLOCK
separately from CDIOCALLOW and CDIOCPREVENT, even though they perform
basically the same function (with a different interface XXX).
|
| 1.80 | 12-Jan-1996 |
thorpej | Handle cases like the following:
- controller calls scsi_done() with error XS_TIMEOUT
- scsi_done() calls sddone()
- sddone() calls disk_unbusy()
- scsi_done() calls controller to retry command (missing the
call to disk_busy())
- controller calls scsi_done()
- scsi_done() calls sddone()
- sddone() calls disk_busy(), which panics because of the imbalance.
Bug noticed by Leo Weppleman, who also suggested this fix; pass an additional
boolean argument ("complete") to the device's "done" routine, with a
value of `0' passed from the previous call to "done", and add an additional
call to "done" when the xfer resources are freed.
|
| 1.79 | 07-Jan-1996 |
thorpej | New generic disk framework. Highlights:
- New metrics handling. Metrics are now kept in the new
`struct disk'. Busy time is now stored as a timeval, and
transfer count in bytes.
- Storage for disklabels is now dynamically allocated, so that
the size of the disk structure is not machine-dependent.
- Several new functions for attaching and detaching disks, and
handling metrics calculation.
Old-style instrumentation is still supported in drivers that did it before.
However, old-style instrumentation is being deprecated, and will go away
once the userland utilities are updated for the new framework.
For usage and architectural details, see the forthcoming disk(9) manual
page.
|
| 1.78 | 07-Dec-1995 |
thorpej | In both cd.c and sd.c:
If the read or write request can fit into a 6-byte cdb, then
use a 6-byte cdb, otherwise use the 10-byte as before.
In sd.c:
In sdattach(), make a note if the device is "ancient" (i.e.
inqbuf.version & SID_ANSII == 0).
Implement sdminphys(): if the device is "ancient", shorten the
transfer so it will fit into a 6-byte cdb.
These changes have eliminated the rejected read/write requests on my
Sun 3/60 with 2 ESDI disks behind an Emulex MD21.
|
| 1.77 | 11-Nov-1995 |
mycroft | Remove the gratuitous code to check for a disc at boot time.
|
| 1.76 | 10-Oct-1995 |
mycroft | branches: 1.76.2;
Return EINVAL if something other than a whole number of blocks is requested.
|
| 1.75 | 26-Sep-1995 |
thorpej | Don't declare Debugger(). It's handled in <sys/systm.h>.
#include <sys/systm.h> where necessary, as suggested by Jonathan Stone.
Fixes PR #1511.
|
| 1.74 | 12-Aug-1995 |
mycroft | Fix oversight in previous.
|
| 1.73 | 12-Aug-1995 |
mycroft | minphys() functions really should return void.
|
| 1.72 | 05-Aug-1995 |
mycroft | Use an intermediate variable to shorten label initialization code.
|
| 1.71 | 24-Jul-1995 |
cgd | update SCSI minphys routines' definitions to match standard minphys()
definition and usage.
|
| 1.70 | 04-Jul-1995 |
mycroft | Make each disk and tape driver define its own read and write functions.
Deprecate rawread() and rawwrite() completely. Remove d_strategy from cdevsw to
force the abstraction barrier.
|
| 1.69 | 26-Jun-1995 |
cgd | make dump stubs consistent
|
| 1.68 | 03-May-1995 |
mycroft | Make the byte-shifting code consistent.
|
| 1.67 | 15-Apr-1995 |
mycroft | Don't boundary check I/O to the `raw' partition.
|
| 1.66 | 01-Apr-1995 |
mycroft | Add a missing unlock.
|
| 1.65 | 29-Mar-1995 |
mycroft | Revamp the locking mechanism slightly.
|
| 1.64 | 25-Mar-1995 |
mycroft | Remove the write protect check altogether, and rely on the drive to do it.
|
| 1.63 | 23-Mar-1995 |
mycroft | Don't bother with DIOCWLABEL.
|
| 1.62 | 23-Mar-1995 |
mycroft | Check for read-only media in open(), not write().
|
| 1.61 | 23-Mar-1995 |
mycroft | Revert cdsize() to do nothing.
|
| 1.60 | 23-Mar-1995 |
mycroft | Rearrange DIOCWDINFO a little; closer to wd.c.
|
| 1.59 | 23-Mar-1995 |
mycroft | Fix typo.
|
| 1.58 | 23-Mar-1995 |
mycroft | Allow DIOCWDINFO even if label is not `writable'.
|
| 1.57 | 07-Mar-1995 |
mycroft | Correct spelling of `fictitious'. Add patterns for optical memory devices.
From Alistair Crooks.
|
| 1.56 | 30-Jan-1995 |
mycroft | Finish last change.
|
| 1.55 | 30-Jan-1995 |
mycroft | Some drives don't grok START with LoEj=1, either.
|
| 1.54 | 30-Jan-1995 |
mycroft | Set the LoadEject bit when issuing a START.
|
| 1.53 | 30-Jan-1995 |
mycroft | Some devices really do require the START before the PREVENT. This makes no
sense.
|
| 1.52 | 26-Jan-1995 |
mycroft | Update copyrights.
|
| 1.51 | 26-Jan-1995 |
mycroft | Redo the `wait for spin up' code. The Mach 3 method clearly doesn't work.
Also, rearrange the first open sequences a bit; SDEV_OPEN is no longer magic.
|
| 1.50 | 23-Jan-1995 |
mycroft | Do the PREVENT before the START.
|
| 1.49 | 16-Jan-1995 |
mycroft | Remove unused macros.
|
| 1.48 | 13-Jan-1995 |
mycroft | Always boundary check I/O.
|
| 1.47 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.46 | 16-Dec-1994 |
mycroft | Remove DIOCSBAD handling. It's not actually handled, so why recognize it?
|
| 1.45 | 14-Dec-1994 |
mycroft | Remove dkbad.h.
|
| 1.44 | 23-Nov-1994 |
mycroft | There's no point in bothering to unlock what we didn't lock.
|
| 1.43 | 22-Nov-1994 |
mycroft | Add still more locking.
|
| 1.42 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.41 | 20-Nov-1994 |
mycroft | Add some missing locking, and some general cleanup.
|
| 1.40 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.39 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.38 | 20-Oct-1994 |
mycroft | Fix minor omission.
|
| 1.37 | 20-Oct-1994 |
mycroft | openpart -> openmask
|
| 1.36 | 20-Oct-1994 |
mycroft | Similar changes to sd.c.
|
| 1.35 | 14-Oct-1994 |
cgd | move arch dependence to archs. clean up a bit. deal with weird MAXPARTITONS.
|
| 1.34 | 11-Aug-1994 |
mycroft | On probe or open, retry the READ CAPACITY once if the drive reports NOT READY.
|
| 1.33 | 05-Aug-1994 |
mycroft | Fix spelling of `STEREO'.
|
| 1.32 | 04-Aug-1994 |
mycroft | Obey the command queue size for the controller.
|
| 1.31 | 04-Jul-1994 |
chopps | branches: 1.31.2;
change to support other MAXPARTITIONS vals aside from 8
|
| 1.30 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.29 | 16-Jun-1994 |
chopps | resolve confusion over who owns the buf after calling scsi_scsi_cmd()
|
| 1.28 | 16-Jun-1994 |
mycroft | b_un.b_addr -> b_data
|
| 1.27 | 11-May-1994 |
mycroft | Add dummy *dump() routines.
|
| 1.26 | 09-May-1994 |
chopps | remove union's from sense_data struct, conditionaly define RAW_PART
|
| 1.25 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.24 | 11-Apr-1994 |
mycroft | Combine scsi_start_unit() and scsi_stop_unit(), and increase the start
timeout (again).
|
| 1.23 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.22 | 06-Feb-1994 |
mycroft | Use b_actf, not av_forw.
|
| 1.21 | 11-Jan-1994 |
mycroft | *strategy functions return void.
|
| 1.20 | 23-Dec-1993 |
cgd | fix from Nick Cuccia (cuccia@remarque.berkeley.edu), apparently
originating from Gary Grebus <glg@k8lt.ampr.org>: partition size
must be in units of DEV_BSIZE.
|
| 1.19 | 17-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.18 | 04-Aug-1993 |
brezak | branches: 1.18.2;
Normalize play_msf changes with 386bsd patch.
|
| 1.17 | 04-Aug-1993 |
brezak | Add play_msf command and MSF variants. This is SCSI2 stuff and is needed for many flavors of player utilities.
|
| 1.16 | 01-Aug-1993 |
mycroft | Add RCS identifiers (this time on the correct side of the branch), and
incorporate recent changes in netbsd-0-9 branch.
|
| 1.15 | 19-Jul-1993 |
cgd | branches: 1.15.2;
patches from allen briggs to fix a minor bug in *attach()
|
| 1.14 | 27-Jun-1993 |
andrew | ANSIfications.
|
| 1.13 | 17-Jun-1993 |
brezak | Silence size complaint at probe.
|
| 1.12 | 16-Jun-1993 |
deraadt | whoops. typo.
|
| 1.11 | 16-Jun-1993 |
deraadt | fix to intuit the maximum number of scsi units available on a device
driver. this piece at least, should be safe from changing sizeof(dev_t)
|
| 1.10 | 25-May-1993 |
deraadt | patch00149 by Julian Elischer <julian@jules.dialix.oz.au> & Rodney Grimes.
When an error was encountered, the sd/cd drivers printed blockno&ff0000
rather that blockno.
|
| 1.9 | 20-May-1993 |
deraadt | First cut at cpu independent disklabels.
There will be niggly little details no doubt..
|
| 1.8 | 20-May-1993 |
cgd | add rcsids and clean up file headers
|
| 1.7 | 11-May-1993 |
cgd | fix stupid line transposition (from John Brezak <brezak@osf.org>)
|
| 1.6 | 04-May-1993 |
deraadt | support for making dev->id_alive be set, this is for iostat to
find disk devices. wee bit of a kludge. sub-device attach()
routines must now return 1 for successful attach(), 0 otherwise.
Other bsd's do this too..
|
| 1.5 | 20-Apr-1993 |
mycroft | Display more meaningful message on SCSI `unit attention'.
|
| 1.4 | 12-Apr-1993 |
deraadt | fixed various bugs like cdattach() returning garbage.
|
| 1.3 | 12-Apr-1993 |
deraadt | new scsi subsystem.
changes also in config/mkioconf.c
i386/isa/wd.c, fd.c, and all scsi drivers.
|
| 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 | after 0.2.2 "stable" patches applied
|
| 1.15.2.1 | 31-Jul-1993 |
cgd | give names, err, wmesg's, to my "pain" -- i.e. convert sleep() to tsleep()
|
| 1.18.2.10 | 16-Feb-1994 |
mycroft | More KNF-like.
|
| 1.18.2.9 | 16-Feb-1994 |
mycroft | Don't return the toc header in CDIOREADTOCENTRYS.
|
| 1.18.2.8 | 06-Feb-1994 |
mycroft | Use b_actf, not av_forw.
|
| 1.18.2.7 | 01-Feb-1994 |
mycroft | Another change from the Mac code...
|
| 1.18.2.6 | 01-Feb-1994 |
mycroft | Add scsi_stop_unit(), from Mac code. Other minor nits.
|
| 1.18.2.5 | 29-Nov-1993 |
mycroft | Use dk_establish().
|
| 1.18.2.4 | 25-Nov-1993 |
mycroft | Don't display extraneous errors if there is no CD present during probe.
|
| 1.18.2.3 | 25-Nov-1993 |
mycroft | (Hopefully) make cd work. Make cd and sd more similar. Implement variable
block sizes in sd.
|
| 1.18.2.2 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.18.2.1 | 24-Sep-1993 |
mycroft | cd.c, sd.c, st.c: strategy functions return void.
|
| 1.31.2.2 | 05-Aug-1994 |
mycroft | update from trunk
|
| 1.31.2.1 | 04-Aug-1994 |
mycroft | update from trunk
|
| 1.76.2.1 | 04-Feb-1996 |
gwr | Pull in rev 1.78 (Jason's fix to use 6-byte command blocks).
|
| 1.97.6.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.101.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.101.2.3 | 16-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.101.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.101.2.1 | 27-Aug-1997 |
thorpej | file cd.c was added on branch marc-pcmcia on 1997-08-27 23:33:00 +0000
|
| 1.107.2.1 | 18-Jan-1999 |
cgd | pull up rev 1.118 from trunk (PR#6093). (is)
|
| 1.124.2.2 | 24-Jun-1999 |
perry | pullup 1.126->1.127 (bouyer)
|
| 1.124.2.1 | 08-Apr-1999 |
bouyer | branches: 1.124.2.1.2; 1.124.2.1.4;
Pull up 1.126 from trunk: add SCSI_IGNORE_NOT_READY to the flags for
CDIOCCLOSE, so if the drive tries to spin up after the clause and fails
because there's no media, eject(1) doesn't get ENODEV. Closes port-i386/7345.
|
| 1.124.2.1.4.1 | 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.
|
| 1.124.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.130.2.2 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.130.2.1 | 21-Dec-1999 |
wrstuden | Initial commit of recent changes to make DEV_BSIZE go away.
Runs on i386, needs work on other arch's. Main kernel routines should be
fine, but a number of the stand programs need help.
cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512
byte block devices. vnd, raidframe, and lfs need work.
Non 2**n block support is automatic for LKM's and conditional for kernels
on "options NON_PO2_BLOCKS".
|
| 1.131.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.131.2.7 | 21-Apr-2001 |
bouyer | Get rid of references to sc_link/scsipi_link in debug or comments.
|
| 1.131.2.6 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.131.2.5 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.131.2.4 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.131.2.3 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.131.2.2 | 20-Oct-1999 |
thorpej | Remove an unnecessary comment regarding XS_CTL_NOSLEEP.
|
| 1.131.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.139.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.141.2.3 | 26-Jun-2002 |
he | Pull up revision 1.150 (via patch, requested by dillo):
Make DVD ioctls use the right size CDB (16 instead of 12). Fixes
PR#12991.
|
| 1.141.2.2 | 06-May-2001 |
he | Pull up revision 1.148 (requested by fvdl):
Compensate for the fact that after __HAVE_OLD_DISKLABEL,
DISKMINOR(u,p) is no longer contiguous over ``p'', so ranges
cannot be passed to vdevgone().
|
| 1.141.2.1 | 01-May-2001 |
he | Pull up revisions 1.142-1.143 (requested by fvdl):
Increase the number of BSD disklabel partitions on i386 to 16.
|
| 1.144.2.13 | 15-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.144.2.12 | 19-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.144.2.11 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.144.2.10 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.144.2.9 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.144.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.144.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.144.2.6 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.144.2.5 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.144.2.4 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.144.2.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.144.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.144.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.151.2.7 | 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
|
| 1.151.2.6 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.151.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.151.2.4 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.151.2.3 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.151.2.2 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.151.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.157.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.157.2.1 | 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).
|
| 1.162.2.5 | 31-Aug-2002 |
gehenna | catch up with -current.
|
| 1.162.2.4 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.162.2.3 | 31-May-2002 |
gehenna | fix conflicts.
|
| 1.162.2.2 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.162.2.1 | 16-May-2002 |
gehenna | Add the block/character device switches.
Replace the direct-access to devsw table with calling devsw API.
|
| 1.185.2.14 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.185.2.13 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.185.2.12 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.185.2.11 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.185.2.10 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.185.2.9 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.185.2.8 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.185.2.7 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.185.2.6 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.185.2.5 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.185.2.4 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.185.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.185.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.185.2.1 | 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.
|
| 1.200.2.2 | 11-Sep-2004 |
he | Pull up revisions 1.203-1.205 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.200.2.1 | 27-Apr-2004 |
jdc | Pull up revision 1.201 (requested by pk in ticket #186)
Some older devices do not understand the `disable block descriptor' bit in
the mode sense request. So fall back on mode sense data including a block
descriptor section.
See also sd.c rev. 1.214. Again: should we bother trying DBD at all?
|
| 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
|
| 1.220.2.2 | 27-Apr-2006 |
tron | Pull up following revision(s) (requested by pavel in ticket #1290):
sys/dev/scsipi/cd.c: revision 1.224
Always allow open() on the raw partition to succeed, even if any of the
SCSI command in cdopen() fails. Fix PR kern/30288 by Pavel Cahyna.
|
| 1.220.2.1 | 06-Apr-2005 |
tron | Pull up revision 1.221 (requested by yamt in ticket #112):
introduce a function to drain bufq and use it where appropriate.
|
| 1.223.2.7 | 21-Jan-2008 |
yamt | sync with head
|
| 1.223.2.6 | 07-Dec-2007 |
yamt | sync with head
|
| 1.223.2.5 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.223.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.223.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.223.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.223.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.234.2.2 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.234.2.1 | 15-Jan-2006 |
yamt | sync with head.
|
| 1.237.2.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.238.2.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.240.6.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.240.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.240.2.4 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.240.2.3 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.240.2.2 | 11-Aug-2006 |
yamt | sync with head
|
| 1.240.2.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.242.6.4 | 11-Sep-2006 |
tron | Pull up following revision(s) (requested by reinoud in ticket #147):
sys/dev/scsipi/cd.c: revision 1.250
Oeps! I overlooked the size argument of the SCSI call. It ought to be equal
to the number of bytes we are reading: READ_DISCINFO_BIGSIZE
sorry.
|
| 1.242.6.3 | 11-Sep-2006 |
tron | Pull up following revision(s) (requested by reinoud in ticket #147):
sys/dev/scsipi/cd.c: revision 1.249
The size of struct scsipi_read_discinfo_data is not even due to the data[1]
pseudo array in the structure. The scsi command issued used
sizeof(scsipi_read_discinfo_data) but included thus the last byte that
wasn't going to be inspected anyway. Using the constant
READ_DISCINFO_BIGSIZE fixes the bug.
Even though SCSI adapters *should* accept odd lengths, the Sun U10
(sparc64's) crashes in bus_space_read_multi_stream_2(). That explains the
crash.
NOTE: somewhere in NetBSD/sparc64's atapibus support there is the
assumption on the even size; this needs to be fixed! (Hi Sparc64 portmaster
-) )
|
| 1.242.6.2 | 08-Sep-2006 |
rpaulo | Pull up following revision(s) (requested by reinoud in ticket #124):
sys/dev/scsipi/cd.c: revision 1.245
sys/dev/scsipi/cd.c: revision 1.246
sys/dev/scsipi/cd.c: revision 1.247
Fix fencing bug on CD disk size; the comment was wrong, its the *length* of
the device and since we count from zero its equal to the number of sectors
and thus one higher than the last sector.
This fixes the read/write problems on the (raw) device where the last
sector was not readable/writable.
Remove two spurious empty lines in functions
Fix panic on reading/writing to a (raw) CD device and encountering an
error. It now correctly sets bp->b_resid to the full size of the buffer.
The failed SCSI read/write command allways contains the complete buffer.
When encountering a read (or write) error, the scsipi stack sets the
xs->resid to zero since the command has been accepted by the drive. The
scsi command set does not allow a partial read or write to be performed and
will signal a success or an error.
|
| 1.242.6.1 | 12-Aug-2006 |
riz | Pull up following revision(s) (requested by reinoud in ticket #12):
sys/fs/udf/udf_vfsops.c: revision 1.9
sys/dev/scsipi/cd.c: revision 1.243
Fix panic on badly or curruptly formatted discs due to an oddity in the
error handling chain.
If NetBSD 4.0 is allready branched at this time this ought to be pulled up.
Fix incorrect `last possible lba' reporting, remove some empty lines and
add mmc classification for some still rare HD DVD device types.
|
| 1.248.2.3 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.248.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.248.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.250.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.250.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.260.2.2 | 01-Feb-2008 |
riz | Pull up following revision(s) (requested by reinoud in ticket #1004):
sys/dev/scsipi/cd.c: revision 1.270
sys/sys/cdio.h: revision 1.29
sys/fs/udf/udf_subr.c: revision 1.42
Pullup fixes from UDF write development project. This patch fixes a
conceptional bug in mmc_discinfo, improves DVD+R feature detection and
prevents corruption of mmc_trackinfo readout on some devices that return
short trackinfo structures.
Pullup fixes from UDF RW support project. This commit fixes the
mmc_discinfo conceptional bug and also vastly improves VAT searching.
|
| 1.260.2.1 | 15-Jan-2007 |
pavel | branches: 1.260.2.1.2;
Pull up following revision(s) (requested by martin in ticket #371):
sys/dev/scsipi/cd.c: revision 1.261
sys/dev/ic/wdc.c: revision 1.244
If __NO_STRICT_ALIGNMENT is not defined, take special care to do PIO
transfers with proper alignement - this removes alignement restrictions
of the buffer pointer passed.
Thanks to bouyer, reinoud, jdc for helping with the code, and special
thanks to Dave Huang for testing.
The previous fix for PR 34202 was wrong (some sizeof used the union, not
the command block) - but instead of fixing this, we can now revert it
completley: the real fix has been applied to ../ic/wdc.c in rev. 1.244,
removing the alignement constraints.
|
| 1.260.2.1.2.1 | 03-Jun-2008 |
skrll | Sync with netbsd-4.
|
| 1.261.2.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.262.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.262.2.9 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.262.2.8 | 24-Aug-2007 |
ad | Sync with buffer cache locking changes. See buf.h/vfs_bio.c for details.
Some minor portions are incomplete and needs to be verified as a whole.
|
| 1.262.2.7 | 20-Aug-2007 |
ad | Sync with head.
|
| 1.262.2.6 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.262.2.5 | 20-Aug-2007 |
ad | - Alter disk attach/detach to fix a panic when closing a vnd device.
- Sync with HEAD.
|
| 1.262.2.4 | 19-Aug-2007 |
ad | - Back out the biodone() changes.
- Eliminate B_ERROR (from HEAD).
|
| 1.262.2.3 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.262.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.262.2.1 | 13-May-2007 |
ad | - Pass the error number and residual count to biodone(), and let it handle
setting error indicators. Prepare to eliminate B_ERROR.
- Add a flag argument to brelse() to be set into the buf's flags, instead
of doing it directly. Typically used to set B_INVAL.
- Add a "struct cpu_info *" argument to kthread_create(), to be used to
create bound threads. Change "bool mpsafe" to "int flags".
- Allow exit of LWPs in the IDL state when (l != curlwp).
- More locking fixes & conversion to the new API.
|
| 1.265.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.267.4.6 | 08-Dec-2007 |
jmcneill | Rename pnp(9) -> pmf(9), as requested by many.
|
| 1.267.4.5 | 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.267.4.4 | 01-Dec-2007 |
jmcneill | Since I'm not convinced that cd_suspend does anything useful other than
to complain that a disc is not present in my drive, nuke it.
|
| 1.267.4.3 | 06-Nov-2007 |
joerg | Flush caches on suspend as discussed with reinoud@
|
| 1.267.4.2 | 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.
|
| 1.267.4.1 | 04-Aug-2007 |
jmcneill | Sync with HEAD.
|
| 1.268.6.2 | 04-Aug-2007 |
rumble | Be sure not to make requests over MAXPHYS when employing bounce buffers.
It was previously possible to round up one sector length if the start block
were not properly aligned.
Resolves PR kern/36716. This may also fix the following reported issues
on port-sgimips:
http://mail-index.netbsd.org/port-sgimips/2005/09/12/0000.html
http://mail-index.netbsd.org/port-sgimips/2005/09/14/0000.html
While here, be sure to put back allocated bufs, which previously appear
to have been leaked. Further, remove the dead write bounce code which never
got executed anyway.
No complaints on tech-kern.
|
| 1.268.6.1 | 04-Aug-2007 |
rumble | file cd.c was added on branch matt-mips64 on 2007-08-04 02:59:00 +0000
|
| 1.268.4.1 | 14-Oct-2007 |
yamt | sync with head.
|
| 1.268.2.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.268.2.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.269.4.3 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.269.4.2 | 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.269.4.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.270.4.1 | 11-Dec-2007 |
yamt | sync with head.
|
| 1.270.2.2 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.270.2.1 | 04-Dec-2007 |
ad | Pull the vmlocking changes into a new branch.
|
| 1.271.2.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.272.6.7 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.272.6.6 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.272.6.5 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.272.6.4 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.272.6.3 | 06-Apr-2008 |
mjf | - after some discussion with agc@ i agreed it would be a good idea to move
device_unregister_* to device_deregister_* to be more like the pmf(9)
functions, especially since a lot of the time the function calls are next
to each other.
- add device_register_name() support for dk(4).
|
| 1.272.6.2 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.272.6.1 | 29-Mar-2008 |
mjf | - etc/devfsd.conf: Add some rules to give nodes like /dev/tty and
/dev/null better default modes, i.e. 0666.
- sbin/init: Run devfsd -s before going to multiuser.
- sys/arch: Provide arm32, i386, sparc with a mem_init() function to request
device nodes for /dev/null, /dev/zero, etc.
- sys/dev: Convert rnd, wd, agp, raid, cd, sd, wsdisplay, wskbd, wsmouse,
wsmux, tty, bpf, swap to devfs New World Order.
- sys/fs/devfs: Make the visibility attribute of device nodes configurable.
Also provide a function to mount a devfs on boot.
- sys/kern: Add a new boot flag, -n. This disables devfs support. Unless
the -n flag is specified the kernel will mount a devfs file
system on boot.
|
| 1.273.4.6 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.273.4.5 | 11-Mar-2010 |
yamt | sync with head
|
| 1.273.4.4 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.273.4.3 | 20-Jun-2009 |
yamt | sync with head
|
| 1.273.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.273.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.273.2.2 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.273.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.278.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.278.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.281.2.1 | 18-Jun-2008 |
simonb | Sync with head.
|
| 1.282.2.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.283.4.3 | 20-Nov-2010 |
riz | Pull up following revision(s) (requested by hubertf in ticket #1386):
sys/dev/raidframe/rf_netbsdkintf.c: revision 1.258
sys/dev/pad/pad.c: revision 1.11
sys/dev/scsipi/cd.c: revision 1.288
sys/dev/isa/pcppi.c: revision 1.33
sys/kern/subr_autoconf.c: revision 1.169
'boot -z' bogons
|
| 1.283.4.2 | 08-Apr-2009 |
snj | Pull up following revision(s) (requested by reinoud in ticket #633):
sys/dev/scsipi/cd.c: revision 1.291
sys/dev/scsipi/scsipi_cd.h: revision 1.21
Implement MMC_TRACKINFO_DATA and MMC_TRACKINFO_AUDIO detection for cdrom and
dvdrom devices. These two compatibility codes were forgotten when those were
introduced.
|
| 1.283.4.1 | 24-Mar-2009 |
snj | Pull up following revision(s) (requested by reinoud in ticket #586):
sys/dev/scsipi/cd.c: revision 1.289
Fix long standing problem with CD/DVD writing with `dd'. If done this way the
device will remain in a wait-for-more-stuff-to-be-written state delaying other
SCSI calls, esp. some read-size or read-layout calls, until its told there is
no more comming.
To fix this, on last close explicitly flush the caches releasing the device
from this mode and allowing other SCSI commands to reach it.
This fixes PR 40528
|
| 1.283.2.3 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.283.2.2 | 03-Mar-2009 |
skrll | Sync with HEAD.
|
| 1.283.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.287.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.287.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.297.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.298.2.2 | 31-May-2011 |
rmind | sync with head
|
| 1.298.2.1 | 30-May-2010 |
rmind | sync with head
|
| 1.302.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.303.4.3 | 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")
|
| 1.303.4.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.303.4.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.304.2.4 | 02-Jun-2012 |
mrg | sync to latest -current.
|
| 1.304.2.3 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.304.2.2 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.304.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.305.2.1 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.309.2.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.309.2.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.309.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.309.2.1 | 12-Sep-2012 |
tls | Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for
physio (I/O to raw devices); needs more doing to get it going with the
filesystems, but it shouldn't damage data.
All work's been done on amd64 so far. Not hard to add support to other
ports. If others want to pitch in, one very helpful thing would be to
sort out when and how IDE disks can do 128K or larger transfers, and
adjust the various PCI IDE (or at least ahcisata) drivers and wd.c
accordingly -- it would make testing much easier. Another very helpful
thing would be to implement a smart minphys() for RAIDframe along the
lines detailed in the MAXPHYS-NOTES file.
|
| 1.311.2.2 | 18-May-2014 |
rmind | sync with head
|
| 1.311.2.1 | 28-Aug-2013 |
rmind | sync with head
|
| 1.318.2.2 | 10-Aug-2014 |
tls | Rebase.
|
| 1.318.2.1 | 07-Apr-2014 |
tls | Be a little more clear and consistent about harvesting entropy from devices:
1) deprecate RND_FLAG_NO_ESTIMATE
2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE
3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE
4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME|
RND_FLAG_COLLECT_VALUE|RND_FLAG_ESTIMATE_TIME
5) Make entropy harvesting from environmental sensors a little more generic
and remove it from individual sensor drivers.
6) Remove individual open-coded delta-estimators for values from a few
places in the tree (uvm, environmental drivers).
7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers
that had stubbed out code, other minor cleanups.
|
| 1.325.2.6 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.325.2.5 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.325.2.4 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.325.2.3 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.325.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.325.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.331.2.5 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.331.2.4 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.331.2.3 | 26-Jul-2016 |
pgoyette | Rename LOCALCOUNT_INITIALIZER to DEVSW_MODULE_INIT. This better describes
what we're doing, and why.
|
| 1.331.2.2 | 19-Jul-2016 |
pgoyette | Instead of repeatedly typing the conditional initialization of the
.d_localcount members in the various {b,c}devsw, define an initializer
macro and use it. This also removes the need for defining new symbols
for each 'struct localcount'.
As suggested by riastradh@
|
| 1.331.2.1 | 18-Jul-2016 |
pgoyette | Rump drivers are always installed via devsw_attach() so we need to
always allocate a 'struct localcount' for these drivers whenever they
are built as modules.
|
| 1.339.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.340.6.3 | 11-Feb-2021 |
martin | Pull up following revision(s) (requested by ryoon in ticket #1652):
sys/dev/scsipi/cd.c: revision 1.350
sys/dev/scsipi/sd.c: revision 1.331
PR 55986: Ryo Onodera: DK_BUSY must have mask as second argument. Make cd.c
consistent by also using __BIT()
|
| 1.340.6.2 | 29-Mar-2020 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1527):
sys/dev/scsipi/cd.c: revision 1.343
sys/kern/subr_disk.c: revision 1.130
Avoid division by zero if label isn't valid.
Allow open of RAWPART even when no medium is loaded.
Keep errors silent if no medium is loaded.
Fixes PR kern/55104
|
| 1.340.6.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.340.4.2 | 29-Apr-2017 |
pgoyette | Remove more unnecessary #include for sys/localcount.h
|
| 1.340.4.1 | 27-Apr-2017 |
pgoyette | Restore all work from the former pgoyette-localcount branch (which is
now abandoned doe to cvs merge botch).
The branch now builds, and installs via anita. There are still some
problems (cgd is non-functional and all atf tests time-out) but they
will get resolved soon.
|
| 1.341.6.3 | 21-Apr-2020 |
martin | Sync with HEAD
|
| 1.341.6.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.341.6.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.341.4.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.342.4.3 | 01-Feb-2022 |
martin | Pull up following revision(s) (requested by reinoud in ticket #1421):
sys/dev/scsipi/cd.c: revision 1.351
Limit buffer size for device capabilities requests as a work-around for PR
kern/56109.
|
| 1.342.4.2 | 11-Feb-2021 |
martin | Pull up following revision(s) (requested by ryoon in ticket #1203):
sys/dev/scsipi/cd.c: revision 1.350
sys/dev/scsipi/sd.c: revision 1.331
PR 55986: Ryo Onodera: DK_BUSY must have mask as second argument. Make cd.c
consistent by also using __BIT()
|
| 1.342.4.1 | 02-Apr-2020 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #814):
sys/dev/scsipi/cd.c: revision 1.343
sys/kern/subr_disk.c: revision 1.130
Avoid division by zero if label isn't valid.
Allow open of RAWPART even when no medium is loaded.
Keep errors silent if no medium is loaded.
Fixes PR kern/55104
|
| 1.343.2.1 | 20-Apr-2020 |
bouyer | Sync with HEAD
|
| 1.349.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.350.2.1 | 17-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.355.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.28 | 08-Sep-2003 |
mycroft | Eliminate the separate ATAPI and SCSI attachments for "cd".
|
| 1.27 | 08-Sep-2003 |
mycroft | Use the const shaker.
|
| 1.26 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.25 | 05-Sep-2003 |
mycroft | Use SMS_DBD in a several places to avoid even thinking about block descriptors.
|
| 1.24 | 10-Jul-2003 |
martin | If a CD drive reports != 2048 byte block size, try to change it into
2048 byte mode. Fixes PR kern/22090.
Reviewed by Manuel Bouyer.
|
| 1.23 | 02-Oct-2002 |
thorpej | branches: 1.23.6;
Add trailing ; to CFATTACH_DECL.
|
| 1.22 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.21 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.20 | 23-Apr-2002 |
bouyer | More copyright fixes, pointed out by Thomas. Thanks !
|
| 1.19 | 09-Jan-2002 |
drochner | -don't apply scsipi_strvis() twice, it tries to octal-print the string
terminating zero and uninitialized dada following
-use use the "vendor" field for the name which is used in the default
disklabel, the "product" field is meaningless or not existing for ATAPI
(serial number)
|
| 1.18 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.17 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.16 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.15 | 14-May-2001 |
bouyer | branches: 1.15.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.14 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.13 | 09-Jun-2000 |
enami | branches: 1.13.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.12 | 29-Mar-2000 |
simonb | branches: 1.12.2;
Don't need to include <sys/conf.h> here.
|
| 1.11 | 23-Sep-1999 |
enami | branches: 1.11.2;
Allow to detach wdc, atapibus, wd and cd.
|
| 1.10 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.9 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.8 | 13-Jul-1998 |
hpeyerl | Add support for ATA CD changer devices like the NEC CDR-251.
Evidence of managerial coding removed by Victor T. Cleaner (thorpej)
|
| 1.7 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.6 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.5 | 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.
|
| 1.4 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file atapicd.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.7 | 04-Jul-1997 |
bouyer | the "NEC CD-ROM DRIVE:260" seems definitively brocken, don't
keep the firmware revision number in the inquiry_match entry.
|
| 1.1.2.6 | 01-Jul-1997 |
thorpej | Remove extra newline printf.
|
| 1.1.2.5 | 01-Jul-1997 |
thorpej | "cd" instances attach to "scsibus" and "atapibus", so name the cfattach's
in a way what reflects that.
|
| 1.1.2.4 | 01-Jul-1997 |
thorpej | Use a naming convention consistent with all other drivers in the source
tree that use multiple attachments... i.e. cd_scsibus_* and cd_atapibus_*.
|
| 1.1.2.3 | 01-Jul-1997 |
thorpej | Use an operations vector set up at attach time rather than checking
for device type each time, for setvol/getvol/setchan operations.
Also, some KNF.
|
| 1.1.2.2 | 01-Jul-1997 |
bouyer | Get this compile on the i386.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file cd_atapi.c was added on branch marc-pcmcia on 1997-08-27 23:33:01 +0000
|
| 1.11.2.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.11.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.11.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.12.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.13.4.8 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.13.4.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.13.4.6 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.13.4.5 | 11-Jan-2002 |
nathanw | More catchup.
|
| 1.13.4.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.13.4.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.13.4.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.13.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.15.2.4 | 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
|
| 1.15.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.15.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.15.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.23.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.5 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.4 | 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.
|
| 1.3 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file cd_link.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Use an operations vector set up at attach time rather than checking
for device type each time, for setvol/getvol/setchan operations.
Also, some KNF.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file cd_link.h was added on branch marc-pcmcia on 1997-08-27 23:33:02 +0000
|
| 1.32 | 08-Sep-2003 |
mycroft | Eliminate the separate ATAPI and SCSI attachments for "cd".
|
| 1.31 | 08-Sep-2003 |
mycroft | Use the const shaker.
|
| 1.30 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.29 | 05-Sep-2003 |
mycroft | Use SMS_DBD in a several places to avoid even thinking about block descriptors.
|
| 1.28 | 10-Jul-2003 |
martin | If a CD drive reports != 2048 byte block size, try to change it into
2048 byte mode. Fixes PR kern/22090.
Reviewed by Manuel Bouyer.
|
| 1.27 | 03-Nov-2002 |
jdc | branches: 1.27.6;
Apply patch from Takeshi Nakayama on current-users which fixes problems
when trying to play audio CD's on some drives. The original message was :
> In my analysis, rev 1.19 of cd_scsi.c used cd_scsibus_set_mode()
> instead of scsipi_mode_select(). cd_scsibus_set_mode() had the code
> which clear a data.header.data_length, but scsipi_mode_select()
> does not have. So we need `data.header.data_length = 0;' before
> calling scsipi_mode_select().
Fixes PR 7299 and probably PR 15155 and PR 16883 too.
|
| 1.26 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.25 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.24 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.23 | 15-Nov-2001 |
lukem | branches: 1.23.10;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.22 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.21 | 02-Sep-2001 |
thorpej | Use the AUDIO_PAGE definition common to SCSI and ATAPI.
|
| 1.20 | 14-May-2001 |
bouyer | branches: 1.20.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.19 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.18 | 09-Jun-2000 |
enami | branches: 1.18.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.17 | 29-Mar-2000 |
simonb | branches: 1.17.2;
Don't need to include <sys/conf.h> here.
|
| 1.16 | 30-Sep-1999 |
thorpej | branches: 1.16.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.15 | 23-Sep-1999 |
enami | Allow to detach wdc, atapibus, wd and cd.
|
| 1.14 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.13 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.12 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.11 | 13-Jul-1998 |
hpeyerl | Add support for ATA CD changer devices like the NEC CDR-251.
Evidence of managerial coding removed by Victor T. Cleaner (thorpej)
|
| 1.10 | 05-Feb-1998 |
mrg | initial import of the new virtual memory system, UVM, into -current.
UVM was written by chuck cranor <chuck@maria.wustl.edu>, with some
minor portions derived from the old Mach code. i provided some help
getting swap and paging working, and other bug fixes/ideas. chuck
silvers <chuq@chuq.com> also provided some other fixes.
this is the rest of the MI portion changes.
this will be KNF'd shortly. :-)
|
| 1.9 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.8 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.7 | 17-Oct-1997 |
bouyer | kern/4267 (Takeshi Nakayama): Add a 'len' parameter to cd_scsibus_get_mode and
cd_scsibus_set_mode, to specifie the real len of "data" we need. These
functions were broken when scsi_cd_write_params_page was added to
scsi_cd_pages, thus changing the len of scsi_cd_mode_data.
|
| 1.6 | 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.
|
| 1.5 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.4 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.3 | 05-Sep-1997 |
bouyer | Don't attach worm devices to the CD driver. This is a private hack, and should
never have been there !
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsicd.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.6 | 01-Jul-1997 |
thorpej | Remove extra newline printf.
|
| 1.1.2.5 | 01-Jul-1997 |
thorpej | "cd" instances attach to "scsibus" and "atapibus", so name the cfattach's
in a way what reflects that.
|
| 1.1.2.4 | 01-Jul-1997 |
thorpej | Use a naming convention consistent with all other drivers in the source
tree that use multiple attachments... i.e. cd_scsibus_* and cd_atapibus_*.
|
| 1.1.2.3 | 01-Jul-1997 |
thorpej | Use an operations vector set up at attach time rather than checking
for device type each time, for setvol/getvol/setchan operations.
Also, some KNF.
|
| 1.1.2.2 | 01-Jul-1997 |
bouyer | Get this compile on the i386.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.3 | 06-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file cd_scsi.c was added on branch marc-pcmcia on 1997-08-27 23:33:04 +0000
|
| 1.16.2.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.16.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.16.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.17.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.18.4.6 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.18.4.5 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.18.4.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.18.4.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.18.4.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.18.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.20.2.3 | 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
|
| 1.20.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.20.2.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.23.10.1 | 13-Dec-2002 |
jmc | Pull up revisions 1.26-1.27 (requested by jdc in ticket #969)
Apply patch from Takeshi Nakayama on current-users which fixes problems
when trying to play audio CD's on some drives.
Fixes PR 7299 and probably PR 15155 and PR 16883 too.
|
| 1.27.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.33 | 10-Dec-2016 |
mlelstv | Refactored sd and cd to use common disk subroutines.
|
| 1.32 | 14-Apr-2015 |
riastradh | branches: 1.32.2;
Include <sys/rndsource.h> where it is actually used.
I had removed <sys/rnd.h> from files that didn't mention anything of
the rnd(9) API. But they included other files which assumed
<sys/rnd.h> had already been included.
|
| 1.31 | 02-Feb-2012 |
tls | branches: 1.31.6; 1.31.24;
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.
|
| 1.30 | 19-Nov-2011 |
tls | branches: 1.30.2;
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.
|
| 1.29 | 19-Oct-2009 |
bouyer | branches: 1.29.12;
Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !
|
| 1.28 | 24-Mar-2008 |
cube | branches: 1.28.4;
Split device_t for cd(4).
|
| 1.27 | 21-Jul-2007 |
ad | branches: 1.27.26; 1.27.28;
Replace some uses of lockmgr().
|
| 1.26 | 11-Dec-2005 |
christos | branches: 1.26.30; 1.26.40;
merge ktrace-lwp.
|
| 1.25 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.24 | 01-Feb-2005 |
reinoud | branches: 1.24.6;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.23 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.22 | 17-Sep-2004 |
mycroft | branches: 1.22.4;
Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.21 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.20 | 08-Sep-2003 |
mycroft | branches: 1.20.2;
Eliminate the separate ATAPI and SCSI attachments for "cd".
|
| 1.19 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.18 | 10-Jul-2003 |
martin | If a CD drive reports != 2048 byte block size, try to change it into
2048 byte mode. Fixes PR kern/22090.
Reviewed by Manuel Bouyer.
|
| 1.17 | 10-May-2003 |
thorpej | branches: 1.17.2;
Change bounds_check_with_label() to take a pointer to the disk structure,
rather than the label itself. This paves the way for some future changes.
|
| 1.16 | 03-Apr-2003 |
fvdl | Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.15 | 22-Jul-2002 |
hannken | Convert to new device buffer queue interface.
|
| 1.14 | 23-Apr-2002 |
bouyer | branches: 1.14.2;
More copyright fixes, pointed out by Thomas. Thanks !
|
| 1.13 | 25-Apr-2001 |
bouyer | branches: 1.13.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.12 | 21-Jan-2000 |
thorpej | branches: 1.12.6;
Update for sys/buf.h/disksort_*() changes.
|
| 1.11 | 23-Sep-1999 |
enami | branches: 1.11.2;
Allow to detach wdc, atapibus, wd and cd.
|
| 1.10 | 02-Feb-1999 |
bouyer | Bump CDRETRIES to 4. Now if a DMA error occurs, the driver should downgrade
to PIO mode and the command be retried properly.
|
| 1.9 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.8 | 13-Jul-1998 |
hpeyerl | Add support for ATA CD changer devices like the NEC CDR-251.
Evidence of managerial coding removed by Victor T. Cleaner (thorpej)
|
| 1.7 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.6 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.5 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.4 | 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.
|
| 1.3 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file cd_link.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Use an operations vector set up at attach time rather than checking
for device type each time, for setvol/getvol/setchan operations.
Also, some KNF.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.2 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.11.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.11.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.12.6.3 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.12.6.2 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.12.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.13.2.2 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.13.2.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.14.2.1 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.17.2.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.17.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.17.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.17.2.2 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.17.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.20.2.1 | 11-Sep-2004 |
he | Pull up revision 1.21 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.22.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.24.6.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.24.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.26.40.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.26.30.1 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.27.28.2 | 21-Jul-2007 |
ad | Replace some uses of lockmgr().
|
| 1.27.28.1 | 21-Jul-2007 |
ad | file cdvar.h was added on branch matt-mips64 on 2007-07-21 19:51:49 +0000
|
| 1.27.26.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.28.4.1 | 11-Mar-2010 |
yamt | sync with head
|
| 1.29.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.30.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.31.24.2 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.31.24.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.31.6.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.32.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.96 | 18-Aug-2025 |
andvar | Fix various typos, mainly in comments:
s/invaid/invalid/
s/instad/instead/
s/wich/with/
s/tranform/transform/
s/tranmist/transmit/
s/tranceiver/transceiver/
s/Tranparent/Transparent/
s/tranlated/translated/
s/tranfer/transfer/
s/tranmissions/transmissions/
s/condtions/conditions/
s/Recient/Recent/
|
| 1.95 | 26-Sep-2021 |
thorpej | Use seltrue_filtops rather than rolling our own with filt_seltrue.
|
| 1.94 | 26-Sep-2021 |
thorpej | Change the kqueue filterops::f_isfd field to filterops::f_flags, and
define a flag FILTEROP_ISFD that has the meaning of the prior f_isfd.
Field and flag name aligned with OpenBSD.
This does not constitute a functional or ABI change, as the field location
and size, and the value placed in that field, are the same as the previous
code, but we're bumping __NetBSD_Version__ so 3rd-party module source code
can adapt, as needed.
NetBSD 9.99.89
|
| 1.93 | 18-Dec-2020 |
thorpej | Use sel{record,remove}_knote().
|
| 1.92 | 25-Oct-2017 |
maya | branches: 1.92.18;
Use C99 initializer for filterops
Mostly done with spatch with touchups for indentation
@@
expression a;
identifier b,c,d;
identifier p;
@@
const struct filterops p =
- { a, b, c, d
+ {
+ .f_isfd = a,
+ .f_attach = b,
+ .f_detach = c,
+ .f_event = d,
};
|
| 1.91 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.90 | 25-Jul-2014 |
dholland | branches: 1.90.4; 1.90.8;
Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.89 | 16-Mar-2014 |
dholland | branches: 1.89.2;
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.
|
| 1.88 | 09-Aug-2013 |
kardel | bump command timeout to 5 minutes. several
types of changers (Overland PowerLoader, Dell
PowerVault) have been exceeding the 100 sec
limit aborting a perfectly (slowly) progressing
operation.
|
| 1.87 | 27-Oct-2012 |
chs | branches: 1.87.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.
|
| 1.86 | 23-Nov-2009 |
rmind | branches: 1.86.12; 1.86.22;
Remove some unecessary includes sys/user.h header.
|
| 1.85 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.84 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.83 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.82 | 08-Jun-2008 |
tsutsui | branches: 1.82.12;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.81 | 28-Apr-2008 |
martin | branches: 1.81.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.80 | 05-Apr-2008 |
cegger | branches: 1.80.2; 1.80.4;
use aprint_*_dev and device_xname
|
| 1.79 | 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>
|
| 1.78 | 05-Dec-2007 |
pooka | branches: 1.78.8; 1.78.12;
Do not "return 1" from kqfilter for errors. That value is passed
directly to the userland caller and results in a mysterious EPERM.
Instead, return EINVAL or something else sensible depending on the
case.
|
| 1.77 | 04-Mar-2007 |
christos | branches: 1.77.14; 1.77.16; 1.77.22; 1.77.24;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.76 | 16-Nov-2006 |
christos | branches: 1.76.4;
__unused removal on arguments; approved by core.
|
| 1.75 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.74 | 28-Aug-2006 |
christos | branches: 1.74.2; 1.74.4;
add missing initializer
|
| 1.73 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.72 | 11-Dec-2005 |
christos | branches: 1.72.4; 1.72.6; 1.72.8; 1.72.10; 1.72.12;
merge ktrace-lwp.
|
| 1.71 | 30-May-2005 |
christos | branches: 1.71.2;
- remove bogus casts
- add more const
|
| 1.70 | 25-Apr-2005 |
drochner | fix more SSD_RCODE_VALID misuse introduced in cleanup
|
| 1.69 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.68 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.67 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.66 | 31-Jan-2005 |
reinoud | Fix LP64 problems introduced by my u_long->uint32_t conversion.
|
| 1.65 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.64 | 18-Sep-2004 |
mycroft | branches: 1.64.4; 1.64.6;
Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.63 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.62 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.61 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.60 | 23-Jun-2004 |
seb | Remove special handling of ASC/ASCQ 0x29/0x00 ("Power On, Reset, or Bus
Device Reset Occurred") when the device is not opened. So let the default
error handler deal with that.
This fix the retrieval of the sense page during autoconfig.
Provided by mycroft@.
Approved by bouyer@.
|
| 1.59 | 23-Apr-2004 |
itojun | use bounded string ops
|
| 1.58 | 08-Sep-2003 |
mycroft | branches: 1.58.2;
Use the const shaker.
|
| 1.57 | 29-Jun-2003 |
fvdl | branches: 1.57.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.56 | 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
|
| 1.55 | 26-Nov-2002 |
christos | si_ -> sel_
|
| 1.54 | 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
|
| 1.53 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.52 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.51 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.50 | 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.
|
| 1.49 | 06-May-2002 |
bouyer | branches: 1.49.2;
Don't return an error for "Not Ready To Ready Transition (Medium May Have
Changed)" and "Power On, Reset, or Bus Device Reset Occurred" senses, if
we're trying to open the device.
|
| 1.48 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.47 | 18-Jul-2001 |
thorpej | branches: 1.47.2;
bzero -> memset
|
| 1.46 | 19-Jun-2001 |
lukem | branches: 1.46.2;
delint
|
| 1.45 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.44 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.43 | 09-Jun-2000 |
enami | branches: 1.43.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.42 | 04-Jan-2000 |
mjacob | branches: 1.42.2;
oops, bad patch
|
| 1.41 | 04-Jan-2000 |
mjacob | - If the user application issues an INITIALIZE ELEMENT STATUS which
succeeds, note that we now are valid.
- Don't attempt to try and run initialize element status from interrupt level-
we don't really support that yet. Also, key more correctly off of ASC/ASCQ
instead of just the sense key.
- Make the practice of doing an INITIALIZE ELEMENT STATUS automatically when
we get params (from chopen time even) a policy decision that is not the
default for now- this can be a dangerous practice as well as time consuming.
It's dangerous in that you can have a hung open when all you really want
to do is do a read of parameters- and parameters, including slot status,
are perfectly fine to read even before an INITIALIZE ELEMENT STATUS is
done- all the elements whos status your read are going to be marked with
an exception- so leave it up to the application to decide how important
this is.
|
| 1.40 | 30-Sep-1999 |
thorpej | branches: 1.40.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.39 | 09-Sep-1999 |
thorpej | Rework the changer driver a bit:
- Take note of magazine changes, and enqueue "Element Status Changed" events
that user processes can read or select on.
- Normalize some structure names.
- Report back more status about changer elements:
* Volume tags (e.g. barcode labels on the backs of your tapes)
* External device names (for drive units in a changer)
* Last element a unit of media was moved from
* Sense information for SCSI changer elements in EXCEPT condition
* Vendor-specific data if the user requests it.
- Add support for setting volume tags.
|
| 1.38 | 04-Apr-1999 |
bouyer | Adjust timeout for 'ielem' to match the comment: 5 mn (not 5s) per element +
10mn initial.
|
| 1.37 | 17-Dec-1998 |
gibbs | branches: 1.37.2;
Return ENODEV when the device capability check fails for MOVE and EXCHANGE
operations.
|
| 1.36 | 08-Dec-1998 |
thorpej | When closing, wait for pending xfers to drain before deleting the reference
to the adapter.
|
| 1.35 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.34 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.33 | 03-Jul-1998 |
mjacob | kern/5514: take the submitter's advice, at least partially. The time it
takes to do IELEM can be proportional to the number of elements, but is
also affected by wierd things like how readable the barcodes on the
media are. There are worst case scenarios I've seen where there are
white labels on the back of tapes with pencilled in labels which is
*just* close enough to being a bar code that an Exabyte 120 would
peer at them myopically and long enough for a *really* long time to
pass in inventorying the jukebox.
I've upped the limit to be proportional to 5 minutes per element. That
is long enough that someone I'm sure will complain about "you wait
to long and should time out" for broken h/w.
As is also noted in the PR, there are a lot of other issues here. It's
really also a question as to whether to update this driver or go
with CAM's driver. This one doesn't have switching between block
descriptors and not, doesn't support volume tag setting, and so on.
Time is limited. This PR should have been closed and fixed right away,
tho.
|
| 1.32 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.31 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.30 | 01-Oct-1997 |
thorpej | Add missing SCSI_DATA_IN when issuing a READ_ELEMENT_STATUS command.
Noticed by Dave Huang <khym@bga.com>.
|
| 1.29 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.28 | 29-Sep-1997 |
mjacob | Add 'initialize element status' command; note (for future fixing in ch.c)
that a number of jukeboxes can't cope with DBD (disable block descriptors)-
we'll have to fix that some day.
|
| 1.27 | 27-Aug-1997 |
bouyer | branches: 1.27.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.26 | 21-Feb-1997 |
thorpej | - Tidy up some printf code in chattach().
- Implement a simple quirks framework for changers.
- Add a quirk for the Spectra 9000 8mm tape library; it requires a
slightly-longer-than-1-minute delay to take tape inventory. Many
thanks to David Webster at Cygnus Solutions for testing this for me.
|
| 1.25 | 05-Dec-1996 |
cgd | branches: 1.25.6;
update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.24 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.23 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.22 | 18-Sep-1996 |
thorpej | For ioctl commands which may change the device's state, ensure that
the caller has the device open for writing.
|
| 1.21 | 19-Apr-1996 |
christos | branches: 1.21.4;
gcc -Wall delinting: unused variables, assignments inside if's and missing
arguments in printf statements.
|
| 1.20 | 03-Apr-1996 |
thorpej | New SCSI media changer driver, written from scratch by me. Some
influence from the old driver (written by Stefan Grefen) and from
an autochanger driver written by the Systems Programming Group at the
University of Utah Computer Science Department (currently residing
in src/sys/arch/hp300/dev/ac.c).
|
| 1.19 | 30-Mar-1996 |
christos | Eliminate scsi_conf.h.
|
| 1.18 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.17 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.16 | 05-Mar-1996 |
thorpej | Normalize SCSI autoconfiguration output.
From Chris Demetriou <cgd@NetBSD.ORG>. Fixes PR #1958.
|
| 1.15 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.14 | 16-Jan-1995 |
mycroft | Remove unused macros.
|
| 1.13 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.12 | 21-Nov-1994 |
mycroft | Clean up open and close routines somewhat.
|
| 1.11 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.10 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.9 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.8 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.7 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.6 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.5 | 17-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.4 | 01-Aug-1993 |
mycroft | branches: 1.4.2;
Add RCS identifiers (this time on the correct side of the branch), and
incorporate recent changes in netbsd-0-9 branch.
|
| 1.3 | 20-May-1993 |
cgd | branches: 1.3.2;
add rcsids and clean up file headers
|
| 1.2 | 20-Apr-1993 |
mycroft | Display more meaningful message on SCSI `unit attention'.
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.3.2.1 | 31-Jul-1993 |
cgd | give names, err, wmesg's, to my "pain" -- i.e. convert sleep() to tsleep()
|
| 1.4.2.3 | 16-Feb-1994 |
mycroft | More KNF-like.
|
| 1.4.2.2 | 29-Nov-1993 |
mycroft | Converted. Needs testing.
|
| 1.4.2.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.21.4.1 | 14-Nov-1996 |
thorpej | Pull up from trunk:
> For ioctl commands which may change the device's state, ensure that
> the caller has the device open for writing.
|
| 1.25.6.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.27.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.27.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.27.2.1 | 27-Aug-1997 |
thorpej | file ch.c was added on branch marc-pcmcia on 1997-08-27 23:33:05 +0000
|
| 1.37.2.1 | 05-Apr-1999 |
bouyer | branches: 1.37.2.1.2;
Pull up rev 1.38 (correct 'ielem' timeout computation); approved by perry.
|
| 1.37.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.40.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.40.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.42.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.43.4.8 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.43.4.7 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.43.4.6 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.43.4.5 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.43.4.4 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.43.4.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.43.4.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.43.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.46.2.7 | 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
|
| 1.46.2.6 | 02-Oct-2002 |
jdolecek | do not need the (void *) cast for kn_hook anymore
|
| 1.46.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.46.2.4 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.46.2.3 | 08-Sep-2001 |
thorpej | Use the seltrue filter as appropriate (or, rather, as the "poll"
entry points of these drivers indicate).
|
| 1.46.2.2 | 08-Sep-2001 |
thorpej | Add kqueue support.
|
| 1.46.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.47.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.47.2.1 | 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).
|
| 1.49.2.1 | 16-May-2002 |
gehenna | Add the character device switch.
|
| 1.57.2.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.57.2.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.57.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.57.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.57.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.57.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.57.2.1 | 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.
|
| 1.58.2.2 | 11-Sep-2004 |
he | Pull up revision 1.62 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.58.2.1 | 24-Jun-2004 |
he | Pull up revision 1.60 (requested by seb in ticket #545):
Remove special handling of ASC/ASCQ 0x29/0x00 ("Power On,
Reset, or Bus Device Reset Occurred") when the device is
not opened. Let the default error handler deal with that
instead. This fixes the retreival of the sense page during
autoconfig.
|
| 1.64.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.64.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.71.2.5 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.71.2.4 | 07-Dec-2007 |
yamt | sync with head
|
| 1.71.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.71.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.71.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.72.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.72.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.72.8.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.72.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.72.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.72.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.74.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.74.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.74.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.76.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.77.24.1 | 08-Dec-2007 |
ad | Sync with head.
|
| 1.77.22.1 | 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.77.16.2 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.77.16.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.77.14.1 | 09-Dec-2007 |
jmcneill | Sync with HEAD.
|
| 1.78.12.4 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.78.12.3 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.78.12.2 | 05-Apr-2008 |
mjf | - add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing
as these are always needed.
- convert many, many drivers over to the New Devfs World Order. For a
list of device drivers yet to be converted see,
http://www.netbsd.org/~mjf/devfs-todo.html.
- add a new device_unregister_all(device_t) function to remove all device
names associated with a device_t, which saves us having to construct
device names when the driver is detached.
- add a DEV_AUDIO type for devices.
|
| 1.78.12.1 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.78.8.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.80.4.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.80.4.3 | 16-May-2009 |
yamt | sync with head
|
| 1.80.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.80.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.80.2.2 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.80.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.81.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.82.12.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.86.22.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.86.22.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.86.22.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 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
|
| 1.87.2.2 | 18-May-2014 |
rmind | sync with head
|
| 1.87.2.1 | 28-Aug-2013 |
rmind | sync with head
|
| 1.89.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.90.8.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.90.4.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.92.18.1 | 03-Jan-2021 |
thorpej | Sync w/ HEAD.
|
| 1.44 | 22-Dec-2022 |
nat | Driver for DaynaPORT SCSI/Link (dse.4).
Written by Hiroshi Noguchi, of which an updated version was posted to
port-mac68k in 2001.
Attachments were added to kernel configs for platforms that already had
the Cabletron (se.4) driver added, although other platorms may benefit.
Reviewed on tech-net by Izumi Tsutsui.
|
| 1.43 | 14-Apr-2022 |
pgoyette | Split some common stuff into scsi_subr module. This enables loading
of the iscsi module whether or not there are any scsi things built
into the kernel.
Addresses the iscsi portion of kern/56772
|
| 1.42 | 12-Feb-2019 |
kardel | Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.41 | 11-Dec-2005 |
christos | branches: 1.41.138; 1.41.142; 1.41.150; 1.41.156; 1.41.164;
merge ktrace-lwp.
|
| 1.40 | 15-Apr-2005 |
drochner | make the scsibus target/lun and the atapibus drive locators optional
(they already had default values; we could kill the [...=N] syntax
completely or at least have config(8) assume that a locator with
a default is optional)
|
| 1.39 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.38 | 12-Aug-2004 |
thorpej | branches: 1.38.4; 1.38.6;
Don't assume wdc-specific ATA / ATAPI code should be brought in if atabus
or atapibus are configured; use a separate wdc_common attribute to indicate
that the shared wdc code is also present.
|
| 1.37 | 08-Oct-2003 |
bouyer | Make the ATA mid-layer appears as atabus, as proposed in
http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html
This adds a device (atabus) between IDE controllers and wd or atapibus, to
have each ATA channel show up in the device tree. Later there will be atabus
devices in /dev, so that we can do IOCTL on them.
Each atabus has its own kernel thread, to handle operations that needs polling,
e.g. reset and others.
Device probing on each bus it defered to the atabus thread creation.
This allows to do the reset and basic device probes in parallel, which reduce
boot time on systems with several pciide controllers.
|
| 1.36 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.35 | 08-Sep-2003 |
mycroft | Eliminate the separate ATAPI and SCSI attachments for "cd".
|
| 1.34 | 08-Sep-2003 |
mycroft | We only need a single attachment for uk at scsibus and atapibus.
|
| 1.33 | 09-Oct-2002 |
thorpej | branches: 1.33.6;
Define a "scsi_core" attribute. Make the "scsi" attribute depend on
"scsi_core". Make all the files previously selected by the "scsi"
attribute selected by the "scsi_core" attribute. Give the "scsibus"
device the "scsi_core" attribute.
|
| 1.32 | 16-May-2002 |
thorpej | Garbage-collect the empty-for-years "su" device.
|
| 1.31 | 28-Nov-2001 |
lukem | branches: 1.31.8;
- convert usage of "defopt" to "defflag" where the relevant option does
not support a value (e.g., it's to be used as "options FOO" instead of
"options FOO=xxx"). options that take a value were converted to
defparam recently.
- minor whitespace & formatting cleanups
|
| 1.30 | 28-Nov-2001 |
fredette | Added the new defopt SCSI_OLD_NOINQUIRY; this is used instead
of sun2 to wrap code that conjures up INQUIRY responses for
certain specific old devices.
|
| 1.29 | 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
|
| 1.28 | 22-May-2001 |
mjacob | branches: 1.28.2;
Add SDRETRIES and SD_IO_TIMEOUT option for sd.
|
| 1.27 | 04-May-2001 |
bouyer | Add an ATAPI front-end to the st driver. Completely untested for now,
but st at scsi should still work :)
|
| 1.26 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.25 | 24-Feb-2001 |
lukem | branches: 1.25.2;
whitespace police
|
| 1.24 | 22-Feb-2000 |
tls | revert previous change
|
| 1.23 | 22-Feb-2000 |
mjacob | remove SES driver
|
| 1.22 | 20-Jan-2000 |
mjacob | add SES device
|
| 1.21 | 06-Jan-2000 |
mjacob | add ST_MOUNT_DELAY options for st
|
| 1.20 | 07-Feb-1999 |
sommerfe | branches: 1.20.8;
Don't bring scsi/atapi common files in under the "atapi" attribute.
Bring them in under atapibus since they're not needed unless something
attached to an atapibus is also attached.
Tested on i386 and arm32.
|
| 1.19 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.18 | 19-Oct-1998 |
bouyer | Allow uk to attach at atapibus.
|
| 1.17 | 12-Oct-1998 |
bouyer | Merge bouyer-ide
|
| 1.16 | 30-Jul-1998 |
mjacob | branches: 1.16.2;
opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.15 | 28-Mar-1998 |
christos | branches: 1.15.2;
Split out the verbose printing of scsi messages, so that user programs
can use the printing routines.
|
| 1.14 | 16-Feb-1998 |
thorpej | Remove "class" declarations, and add "devclass" declarations where
appropriate. Fix several inconsistencies between device class and
attributes. Mostly from Chris Demetriou.
|
| 1.13 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.12 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.11 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.10 | 13-Sep-1997 |
enami | Declare SCSIVERBOSE by defopt in files.scsipi. Include opt_scsiverbose.h
in scsi_base.c and st.c.
|
| 1.9 | 27-Aug-1997 |
bouyer | branches: 1.9.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.8 | 20-Jul-1997 |
pk | config.new => config
|
| 1.7 | 21-Mar-1997 |
matthias | Add missing arp tag for se driver.
|
| 1.6 | 18-Mar-1997 |
thorpej | Driver for the Cabletron EA41x SCSI Ethernet Adaptor, written by
Ian Dall <ian.dall@dsto.defence.gov.au>. Converted to "new arp"
and some other (very) minor changes by me.
|
| 1.5 | 21-Feb-1997 |
jeremy | Fix typo in comment.
|
| 1.4 | 16-May-1996 |
mycroft | branches: 1.4.8;
Slight adjustments for new version of ahc driver.
|
| 1.3 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.2 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.1 | 17-Apr-1995 |
cgd | config.new description of SCSI mi files.
|
| 1.4.8.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.9.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.3 | 16-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.1 | 27-Aug-1997 |
thorpej | file files.scsipi was added on branch marc-pcmcia on 1997-08-27 23:33:06 +0000
|
| 1.15.2.2 | 13-Aug-1998 |
bouyer | - sync with HEAD
- better error detection and report for wd drives
- better wddump() support
- wdcwait() now takes a timout argument, so that we can honnor the
timeout argument of scsipi cmds.
|
| 1.15.2.1 | 04-Jun-1998 |
bouyer | Commit changes to the IDE system in a branch. This allows a better separation
between higth-level and low-level (i.e. registers read/write) and generalize
the queue for all commands. This also add supports for IDE DMA.
|
| 1.16.2.2 | 30-Jul-1998 |
mjacob | opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.16.2.1 | 30-Jul-1998 |
mjacob | file files.scsipi was added on branch eeh-paddr_t on 1998-07-30 00:49:20 +0000
|
| 1.20.8.3 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.20.8.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.20.8.1 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.25.2.4 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.25.2.3 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.25.2.2 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.25.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.28.2.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.28.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.31.8.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.33.6.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.33.6.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.33.6.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.33.6.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.33.6.2 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.33.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.38.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.38.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.41.164.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.41.156.1 | 01-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1197):
sys/dev/scsipi/st.c: revision 1.236
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
-
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.41.150.1 | 07-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1682):
sys/dev/scsipi/st.c: revision 1.236 (patch)
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.41.142.1 | 07-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1682):
sys/dev/scsipi/st.c: revision 1.236 (patch)
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.41.138.1 | 07-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1682):
sys/dev/scsipi/st.c: revision 1.236 (patch)
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.9 | 23-May-2025 |
nat | Schedule callback immediately if there is a hint that there is more data.
This increased the data rate by 300%
|
| 1.8 | 08-Sep-2024 |
nat | Only input needs to be polled.
Tested with PDMA on mac68k (on emulated hw).
|
| 1.7 | 08-Sep-2024 |
nat | Use aprint_normal_dev for ethernet address.
NFC.
|
| 1.6 | 06-Jul-2024 |
andvar | s/occurence/occurrence/ in comment.
|
| 1.5 | 01-Jan-2024 |
gutteridge | branches: 1.5.2;
if_dse.c: s/addredses/addresses/ in comment
|
| 1.4 | 20-Dec-2023 |
skrll | Remove unnecssary #include
|
| 1.3 | 22-Dec-2022 |
nat | Fix condition for ending the pacet read loop.
len is unsigned 16 bit so testing for less than zero is not valid.
|
| 1.2 | 22-Dec-2022 |
nat | Remove unused commented out code.
Remove unintentional stray debug printfs.
Fix DSE_DEBUG build.
|
| 1.1 | 22-Dec-2022 |
nat | Driver for DaynaPORT SCSI/Link (dse.4).
Written by Hiroshi Noguchi, of which an updated version was posted to
port-mac68k in 2001.
Attachments were added to kernel configs for platforms that already had
the Cabletron (se.4) driver added, although other platorms may benefit.
Reviewed on tech-net by Izumi Tsutsui.
|
| 1.5.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.119 | 20-Dec-2023 |
skrll | Remove unnecssary #include
|
| 1.118 | 29-Aug-2022 |
skrll | Make this build again. Sorry about that.
|
| 1.117 | 28-Aug-2022 |
skrll | se(4): don't set if_watchdog as it's not used.
if_timer is never set in this driver and so if_watchdog will never be
called.
|
| 1.116 | 07-Jul-2022 |
skrll | Trailing whitespace
|
| 1.115 | 01-Jan-2022 |
msaitoh | s/sytem/system/
|
| 1.114 | 19-Sep-2021 |
andvar | fix various typos in comments, messages and documentation.
|
| 1.113 | 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.
|
| 1.112 | 29-Sep-2020 |
msaitoh | branches: 1.112.6;
s/occurence/occurrence/
|
| 1.111 | 27-Jul-2020 |
jdc | Improve the workqueue and callout handling. Prompted by riastradh@.
|
| 1.110 | 22-Jul-2020 |
riastradh | Sort #includes. Nix trailing whitespace.
No functional change intended.
|
| 1.109 | 22-Jul-2020 |
riastradh | Omit needless <sys/pcq.h>.
|
| 1.108 | 29-Jun-2020 |
riastradh | Nix trailing whitespace.
|
| 1.107 | 22-Jun-2020 |
jdc | Add sedetach() and also use it to remove duplicate code from seattach().
Correct a comment about sedone().
|
| 1.106 | 22-Jun-2020 |
jdc | Use workqueues so that we don't call into the scsipi subsystem via
a softint from the network stack.
Don't recurse through scsipi_command() when we have multiple packets
in the send queue - use a loop instead. This means that we no longer
need sestart(), as we can now handle everything in sedone().
Fix a couple of XXX's.
Rework the locking logic slightly from the previous revision.
Now this works with DIAGNOSTIC+LOCKDEBUG.
|
| 1.105 | 19-Jun-2020 |
jdc | First pass at making this work again.
Remove spl and add some locking around network access (needs more work).
Make sure that we consistently use the channel lock for scsipi commands.
Remove the preference for send over receive, as this can lead to deadlocks
- we only advertise 1 opening, but we can try to send before the receive is
complete in this case.
Don't use XS_CTL_ASYNC because we don't provide a buffer.
Tested on UP sparc and compile-tested on atari.
Tested with LOCKDEBUG.
Still fails with DIAGNOSTIC because we can call into the scsipi routines
from a softint.
|
| 1.104 | 29-Jan-2020 |
thorpej | Adopt <net/if_stats.h>.
|
| 1.103 | 10-Nov-2019 |
chs | branches: 1.103.2;
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.
|
| 1.102 | 28-May-2019 |
msaitoh | Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
|
| 1.101 | 23-May-2019 |
msaitoh | Whitespace fix (mainly tabify).
|
| 1.100 | 23-May-2019 |
msaitoh | No functional change:
- Change ac(was arpcom) to ec(ethercom)
- Simplify MII structure initialization.
- u_int*_t -> uint*_t.
- KNF
|
| 1.99 | 05-Feb-2019 |
msaitoh | Remove very old IFF_NOTRAILERS flag.
|
| 1.98 | 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!)
|
| 1.97 | 26-Jun-2018 |
msaitoh | branches: 1.97.2;
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.
|
| 1.96 | 22-Jun-2018 |
msaitoh | It's not required to include net/bpfdesc.h. Remove it.
|
| 1.95 | 23-Oct-2017 |
msaitoh | branches: 1.95.2;
If if_initialize() failed in the attach function, free resources and return.
|
| 1.94 | 15-Dec-2016 |
ozaki-r | branches: 1.94.8;
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
|
| 1.93 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.92 | 02-Oct-2016 |
christos | MFREE -> m_free
|
| 1.91 | 07-Sep-2016 |
jakllsch | '#if 0' a static inline function used only by an already #if 0 function.
|
| 1.90 | 10-Jun-2016 |
ozaki-r | branches: 1.90.2;
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.
|
| 1.89 | 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!
|
| 1.88 | 24-Aug-2015 |
pooka | would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.87 | 25-Jul-2014 |
dholland | branches: 1.87.4;
Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.86 | 16-Mar-2014 |
dholland | branches: 1.86.2;
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.
|
| 1.85 | 27-Oct-2012 |
chs | branches: 1.85.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.
|
| 1.84 | 03-Feb-2012 |
christos | branches: 1.84.6;
ansi prototypes
|
| 1.83 | 02-Apr-2011 |
mbalmer | branches: 1.83.4; 1.83.8;
Fix misplaced parenthesis. From henning.petersen@t-online.de, thanks.
|
| 1.82 | 27-Jul-2010 |
jakllsch | branches: 1.82.2;
Use %zu instead of %d for size_t.
Just in case anyone ever debugs se(4) on LP64.
|
| 1.81 | 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.
|
| 1.80 | 19-Jan-2010 |
pooka | branches: 1.80.2; 1.80.4;
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.
|
| 1.79 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.78 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.77 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.76 | 14-Mar-2009 |
dsl | ANSIfy another 1261 function definitions.
The only ones left in sys are beyond by sed script!
(or in sys/dist or sys/external)
Mostly they have function pointer parameters.
|
| 1.75 | 14-Mar-2009 |
dsl | Change about 4500 of the K&R function definitions to ANSI ones.
There are still about 1600 left, but they have ',' or /* ... */
in the actual variable definitions - which my awk script doesn't handle.
There are also many that need () -> (void).
(The script does handle misordered arguments.)
|
| 1.74 | 11-Jan-2009 |
cegger | branches: 1.74.2;
make this compile
|
| 1.73 | 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.
|
| 1.72 | 08-Jun-2008 |
tsutsui | branches: 1.72.4; 1.72.6;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.71 | 05-Apr-2008 |
cegger | branches: 1.71.2; 1.71.4; 1.71.6;
use aprint_*_dev and device_xname
|
| 1.70 | 20-Dec-2007 |
dyoung | branches: 1.70.6;
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.
|
| 1.69 | 01-Sep-2007 |
dyoung | branches: 1.69.8; 1.69.12;
Remark that the data-length argument passed to se_scsipi_cmd() is
questionable.
|
| 1.68 | 01-Sep-2007 |
dyoung | Create a temporary, non-const copy of a sockaddr. Pass that to
se_set_multi() or se_remove_multi(), because neither is easily
constified. Thanks jmmv@ for reporting the issue.
|
| 1.67 | 01-Sep-2007 |
dyoung | Change a bazillion occurrences of code resembling this,
error = (cmd == SIOCADDMULTI) ?
ether_addmulti(ifr, &sc->sc_ec) :
ether_delmulti(ifr, &sc->sc_ec);
if (error == ENETRESET) {
to this,
if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) {
which does the same thing.
(A bazillion is a very large number. This seems to make the i386
ALL kernel smaller by 3kB to 4kB.)
Use ifreq_getaddr() twice in es(4).
Whitespace nits.
|
| 1.66 | 09-Jul-2007 |
ad | branches: 1.66.2; 1.66.6; 1.66.8;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.65 | 04-Mar-2007 |
yamt | branches: 1.65.2; 1.65.4;
fix fallout from caddr_t changes.
|
| 1.64 | 04-Mar-2007 |
christos | Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.63 | 13-Jan-2007 |
cube | branches: 1.63.2;
Complete initializers so those files compile.
|
| 1.62 | 07-Sep-2006 |
dogcow | branches: 1.62.2;
remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.
|
| 1.61 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.60 | 24-Dec-2005 |
perry | branches: 1.60.4; 1.60.6; 1.60.8; 1.60.10; 1.60.12;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.59 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.58 | 03-Jun-2005 |
jdc | branches: 1.58.2;
Remove cast to non-const and make this compile again.
|
| 1.57 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.56 | 04-Feb-2005 |
perry | de-__P
|
| 1.55 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.54 | 31-Jan-2005 |
simonb | Revert incorrect u_long->uint32_t conversion.
|
| 1.53 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.52 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.51 | 30-Oct-2004 |
thorpej | branches: 1.51.4; 1.51.6;
When adding/deleting multicast addresses, only whack the address
filter if the interface is marked RUNNING.
Fixes kern/27678.
|
| 1.50 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.49 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.48 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.47 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.46 | 23-Apr-2004 |
itojun | use bounded string ops
|
| 1.45 | 08-Sep-2003 |
mycroft | branches: 1.45.2;
Use the const shaker.
|
| 1.44 | 23-Oct-2002 |
jdolecek | branches: 1.44.6;
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
|
| 1.43 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.42 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.41 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.40 | 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.
|
| 1.39 | 15-Nov-2001 |
lukem | branches: 1.39.8;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.38 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.37 | 18-Jul-2001 |
thorpej | branches: 1.37.2;
bcmp -> memcmp
|
| 1.36 | 18-Jul-2001 |
thorpej | bcopy -> memcpy
|
| 1.35 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.34 | 25-Apr-2001 |
bouyer | branches: 1.34.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.33 | 14-Dec-2000 |
thorpej | branches: 1.33.2;
ALTQ'ify.
|
| 1.32 | 15-Nov-2000 |
thorpej | Move bpfattach()/bpfdetach() calls into ether_ifattach()/ether_ifdetach().
|
| 1.31 | 01-Oct-2000 |
thorpej | Move the check for "promisc + unicast + not for us" into ether_input(),
and change Ethernet drivers to always pass all received frames to
ether_input() (with a few exceptions, which are documented in the
code).
|
| 1.30 | 09-Jun-2000 |
enami | Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.29 | 30-Mar-2000 |
augustss | branches: 1.29.2;
Get rid of register declarations.
|
| 1.28 | 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.
|
| 1.27 | 13-Mar-2000 |
soren | Fix doubled 'the's in comments.
|
| 1.26 | 13-Nov-1999 |
matthias | Make this work again by adding XS_CTL_ASYNC to the SCSI requests in
se_ifstart and se_recv.
|
| 1.25 | 30-Sep-1999 |
thorpej | branches: 1.25.2; 1.25.4; 1.25.8;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.24 | 18-May-1999 |
thorpej | Rework layer 2 protocol input routines. Instead of calling e.g. ether_input()
directly, call the function pointer (*if_input)(ifp, m). The input routine
expects the packet header to be at the head of the packet, and will adjust
as necessary. Privatize the layer 2 input and output routines, allowing
*_ifattach() to set them up as appropriate.
|
| 1.23 | 12-Dec-1998 |
mycroft | branches: 1.23.4;
Simplify the copy loops a bit.
|
| 1.22 | 08-Dec-1998 |
thorpej | When closing, wait for pending xfers to drain before deleting the reference
to the adapter.
|
| 1.21 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.20 | 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.
|
| 1.19 | 25-Sep-1998 |
is | correct obvious typo
|
| 1.18 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.17 | 05-Jul-1998 |
jonathan | defopt NS, NSIP.
|
| 1.16 | 05-Jul-1998 |
jonathan | defopt LLC
|
| 1.15 | 05-Jul-1998 |
jonathan | defopt CCITT.
|
| 1.14 | 05-Jul-1998 |
jonathan | defopt INET, NETATALK.
|
| 1.13 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.12 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.11 | 14-Oct-1997 |
matthias | Apply Ian Dall's patch from kern/4004. Ian's comment:
The following patch employs a smarter adaptive polling
scheme. It also improves the comments, in particular giving
due credit to Phil Budne for his efforts in nunderstanding
the device. It also relaxes the pattern for auto config so
more devices match.
|
| 1.10 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.9 | 27-Aug-1997 |
bouyer | branches: 1.9.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.8 | 28-Apr-1997 |
mycroft | Oops; forgot to GC the last mbuf allocated when out of clusters.
|
| 1.7 | 24-Apr-1997 |
mycroft | Fix typo in previous.
|
| 1.6 | 24-Apr-1997 |
mycroft | If we fail to allocate a cluster to hold a large packet, simply
drop it rather than using a chain of tiny mbufs.
Do not force the initial part of a packet into a separate mbuf.
|
| 1.5 | 04-Apr-1997 |
matthias | * Do some KNFing.
* Make all local functions static.
* Add some untested netatalk support.
|
| 1.4 | 02-Apr-1997 |
mycroft | Push the buffer cleanup code into scsi_done(), and split it so that biodone()
is called *after* the driver `done' routine. This fixes disk I/O statistics
on SCSI devices.
Also, calling the `done' routine with a `complete' argument of 0 and actually
having it do anything meaningful loses in at least 3 ways, so just nuke the
argument altogether and don't call it this way. If the driver needs to do
some error handling, that's what `err_handler' is for.
|
| 1.3 | 24-Mar-1997 |
thorpej | Fix a screwup (my fault, oops) that caused the stack to get corrupted.
From Ian Dall <Ian.Dall@dsto.defence.gov.au> on port-pc532.
|
| 1.2 | 18-Mar-1997 |
cgd | add appropriate includes so that this deals properly with NS and CCITT+LLC.
(Includes stolen from dev/ic/am7990.c, because it's a (the?) canonical
networking hardware driver.)
|
| 1.1 | 18-Mar-1997 |
thorpej | Driver for the Cabletron EA41x SCSI Ethernet Adaptor, written by
Ian Dall <ian.dall@dsto.defence.gov.au>. Converted to "new arp"
and some other (very) minor changes by me.
|
| 1.9.2.4 | 15-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.9.2.1 | 27-Aug-1997 |
thorpej | file if_se.c was added on branch marc-pcmcia on 1997-08-27 23:33:07 +0000
|
| 1.23.4.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.25.8.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.25.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.25.2.5 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.25.2.4 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.25.2.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.25.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.25.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.29.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.33.2.7 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.33.2.6 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.5 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.33.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.33.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.34.2.3 | 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
|
| 1.34.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.34.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.37.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.37.2.1 | 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).
|
| 1.39.8.1 | 16-May-2002 |
gehenna | Add the character device switch.
|
| 1.44.6.8 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.44.6.7 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.44.6.6 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.44.6.5 | 21-Nov-2004 |
skrll | Adapt to branch. Alpha kernels now compile.
|
| 1.44.6.4 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.44.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.44.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.44.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.45.2.1 | 11-Sep-2004 |
he | branches: 1.45.2.1.2;
Pull up revision 1.47 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.45.2.1.2.1 | 24-Jan-2005 |
he | Pull up revision 1.51 (requested by thorpej in ticket #939):
When adding or deleting multicast addresses, only change
the address filter if the interface is marked RUNNING.
Fixes PR#27678.
|
| 1.51.6.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.51.6.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.51.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.58.2.5 | 21-Jan-2008 |
yamt | sync with head
|
| 1.58.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.58.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.58.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.58.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.60.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.60.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.60.8.2 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.60.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.60.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.60.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.62.2.1 | 01-Feb-2007 |
ad | Sync with head.
|
| 1.63.2.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.65.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.65.2.2 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.65.2.1 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.66.8.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.66.8.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.66.6.1 | 03-Sep-2007 |
jmcneill | Sync with HEAD.
|
| 1.66.2.1 | 03-Sep-2007 |
skrll | Sync with HEAD.
|
| 1.69.12.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.69.8.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.70.6.3 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.70.6.2 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.70.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.71.6.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.71.4.4 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.71.4.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.71.4.2 | 16-May-2009 |
yamt | sync with head
|
| 1.71.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.71.2.1 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.72.6.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.72.6.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.72.4.1 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.74.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.80.4.3 | 21-Apr-2011 |
rmind | sync with head
|
| 1.80.4.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.80.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.80.2.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.80.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.82.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.83.8.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.83.4.3 | 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")
|
| 1.83.4.2 | 30-Oct-2012 |
yamt | sync with head
|
| 1.83.4.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.84.6.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.84.6.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.84.6.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.85.2.1 | 18-May-2014 |
rmind | sync with head
|
| 1.86.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.87.4.6 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.87.4.5 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.87.4.4 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.87.4.3 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.87.4.2 | 19-Mar-2016 |
skrll | Sync with HEAD
|
| 1.87.4.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.90.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.90.2.1 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.94.8.2 | 26-Jul-2018 |
snj | Pull up following revision(s) (requested by msaitoh in ticket #938):
sys/arch/acorn32/podulebus/if_ie.c: revision 1.41
sys/arch/amiga/dev/if_es.c: revision 1.58
sys/arch/amiga/dev/if_qn.c: revision 1.45
sys/arch/arm/at91/at91emac.c: revision 1.20
sys/arch/arm/ep93xx/epe.c: revision 1.37
sys/arch/emips/ebus/if_le_ebus.c: revision 1.14
sys/arch/emips/ebus/if_le_ebus.c: revision 1.15
sys/arch/mac68k/dev/if_mc.c: revision 1.46
sys/arch/macppc/dev/am79c950.c: revision 1.39
sys/arch/newsmips/apbus/if_sn.c: revision 1.40
sys/arch/next68k/dev/mb8795.c: revision 1.59
sys/arch/playstation2/dev/if_smap.c: revision 1.25
sys/arch/playstation2/dev/if_smap.c: revision 1.26
sys/arch/sun2/dev/if_ec.c: revision 1.28
sys/arch/sun3/dev/if_ie.c: revision 1.63
sys/arch/x68k/dev/if_ne_intio.c: revision 1.19
sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75
sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63
sys/dev/bi/if_ni.c: revision 1.45
sys/dev/cadence/if_cemac.c: revision 1.12
sys/dev/ic/am7990.c: revision 1.78
sys/dev/ic/am79900.c: revision 1.27
sys/dev/ic/an.c: revision 1.67
sys/dev/ic/cs89x0.c: revision 1.40
sys/dev/ic/dm9000.c: revision 1.13
sys/dev/ic/dm9000.c: revision 1.14
sys/dev/ic/dp8390.c: revision 1.88
sys/dev/ic/elink3.c: revision 1.141
sys/dev/ic/elinkxl.c: revision 1.122
sys/dev/ic/hme.c: revision 1.98
sys/dev/ic/i82586.c: revision 1.77
sys/dev/ic/lance.c: revision 1.53
sys/dev/ic/mb86950.c: revision 1.27
sys/dev/ic/mb86960.c: revision 1.86
sys/dev/ic/mtd803.c: revision 1.34
sys/dev/ic/pdq_ifsubr.c: revision 1.59
sys/dev/ic/rrunner.c: revision 1.86
sys/dev/ic/seeq8005.c: revision 1.58
sys/dev/ic/sgec.c: revision 1.47
sys/dev/ic/smc90cx6.c: revision 1.72
sys/dev/ic/smc91cxx.c: revision 1.96
sys/dev/ic/tropic.c: revision 1.49
sys/dev/ic/wi.c: revision 1.245
sys/dev/isa/if_eg.c: revision 1.93
sys/dev/isa/if_el.c: revision 1.95
sys/dev/isa/if_iy.c: revision 1.101
sys/dev/ofw/ofnet.c: revision 1.58
sys/dev/pci/if_alc.c: revision 1.27
sys/dev/pci/if_de.c: revision 1.152
sys/dev/pci/if_fpa.c: revision 1.61
sys/dev/pci/if_jme.c: revision 1.34
sys/dev/pci/if_tl.c: revision 1.108
sys/dev/pci/if_vte.c: revision 1.19
sys/dev/pci/ixgbe/ixgbe.h: revision 1.50
sys/dev/pcmcia/if_cnw.c: revision 1.62
sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17
sys/dev/pcmcia/if_ray.c: revision 1.89
sys/dev/pcmcia/if_xi.c: revision 1.81
sys/dev/pcmcia/mhzc.c: revision 1.51
sys/dev/pcmcia/xirc.c: revision 1.34
sys/dev/qbus/if_de.c: revision 1.33
sys/dev/qbus/if_qe.c: revision 1.78
sys/dev/qbus/if_qt.c: revision 1.22
sys/dev/sbus/be.c: revision 1.87
sys/dev/sbus/qe.c: revision 1.68
sys/dev/scsipi/if_se.c: revision 1.96
sys/dev/usb/if_atu.c: revision 1.59
sys/net/if_l2tp.c: revision 1.28 via patch
sys/net/if_ppp.c: revision 1.160
It's not required to include net/bpfdesc.h. Remove it.
--
Simplify like other drivers. NULL check of ifp->if_bpf is done in
bpf_mtap(), so it's not required to do it here.
--
Remove duplicated inclusion of net/bpf.h.
--
Remove duplicated inclusion of net/bpf.h.
--
Simplify bpf_mtap() call. No functional change.
|
| 1.94.8.1 | 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
|
| 1.95.2.3 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.95.2.2 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.95.2.1 | 25-Jun-2018 |
pgoyette | Sync with HEAD
|
| 1.97.2.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.97.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.97.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.103.2.1 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.112.6.1 | 17-Jun-2021 |
thorpej | Sync w/ HEAD.
|
| 1.25 | 25-Feb-2012 |
shattered | Fix typos in comments.
OK by wiz@
|
| 1.24 | 11-Dec-2005 |
christos | branches: 1.24.110; 1.24.114;
merge ktrace-lwp.
|
| 1.23 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.22 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.21 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.20 | 07-Dec-2004 |
thorpej | branches: 1.20.2; 1.20.4;
Nothing references "struct scsi_generic" anymore; delete it.
|
| 1.19 | 06-Jan-2003 |
matt | branches: 1.19.2;
Add multiple inclusion protection.
|
| 1.18 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.17 | 19-Apr-2001 |
augustss | Correct the opcode for SCSI_MODE_SENSE_BIG.
|
| 1.16 | 29-Nov-2000 |
wiz | branches: 1.16.2;
Fix typo in comment reported by Tetsuya Isaki in kern/11587.
|
| 1.15 | 05-Oct-2000 |
ad | Add some more completion status codes.
|
| 1.14 | 13-Feb-1998 |
enami | branches: 1.14.14;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.13 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.12 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.11 | 27-Aug-1997 |
bouyer | branches: 1.11.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.10 | 12-Sep-1996 |
thorpej | Remove a couple of opcodes that are already (and more appropriately)
defined in scsi_changer.h
|
| 1.9 | 03-Sep-1996 |
thorpej | Store the SCSI version in the scsi_link, and set the LUN in the CDB
if the version is <= SCSI-2. This should help some older SCSI
devices that previously needed the "NOLUNS" quirk. While this is
not strictly necessary on SCSI-2 devices, the spec allows it,
so we set it for SCSI-2 devices "just in case". See section 7.2.2 of
Draft X3T9.2 Rev 10L for details.
|
| 1.8 | 10-Jul-1996 |
explorer | Add SCSI_QUEUE_FULL ; part of pr port-i386/2600
|
| 1.7 | 19-Mar-1996 |
mycroft | branches: 1.7.4;
Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.6 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 09-May-1994 |
chopps | remove union's from sense_data struct, conditionaly define RAW_PART
|
| 1.3 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.2 | 20-May-1993 |
cgd | branches: 1.2.3;
add rcsids and clean up file headers
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.2.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.7.4.1 | 18-Jul-1996 |
jtc | Pulled up from rev 1.8, ahc fixes
|
| 1.11.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.11.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.11.2.1 | 27-Aug-1997 |
thorpej | file scsi_all.h was added on branch marc-pcmcia on 1997-08-27 23:33:09 +0000
|
| 1.14.14.3 | 21-Apr-2001 |
bouyer | Sync with HEAD
|
| 1.14.14.2 | 08-Dec-2000 |
bouyer | Sync with HEAD.
|
| 1.14.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.16.2.2 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.16.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.19.2.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.19.2.1 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.20.4.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.20.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.24.114.1 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.24.110.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.93 | 03-May-2019 |
mlelstv | Avoid null pointer deref in printing xfer mode when no target driver
is attached. Fixes kern/54157.
|
| 1.92 | 17-Jun-2017 |
mlelstv | branches: 1.92.6;
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.91 | 20-Nov-2016 |
mlelstv | branches: 1.91.8;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.90 | 20-Apr-2012 |
bouyer | branches: 1.90.2; 1.90.16; 1.90.20;
Add a bustype_async_event_xfer_mode() callback to scsipi_bustype (which can
be NULL), so that transport-specific details of transfer mode setting/printing
can be handled more easily.
Move scsipi_async_event_xfer_mode() and scsipi_print_xfer_mode() to
scsi_base.c and split in parallel scsi and FC/SAS parts.
size of struct scsipi_bustype has changed, welcome to 6.99.5
|
| 1.89 | 28-Apr-2008 |
martin | branches: 1.89.34; 1.89.38;
Remove clause 3 and 4 from TNF licenses
|
| 1.88 | 05-Apr-2008 |
cegger | branches: 1.88.2; 1.88.4;
use aprint_*_dev and device_xname
|
| 1.87 | 16-Nov-2006 |
christos | branches: 1.87.48;
__unused removal on arguments; approved by core.
|
| 1.86 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.85 | 11-Dec-2005 |
christos | branches: 1.85.20; 1.85.22;
merge ktrace-lwp.
|
| 1.84 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.83 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.82 | 17-Sep-2004 |
mycroft | Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.81 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.80 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.79 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.78 | 05-Aug-2004 |
bouyer | scsi_kill_pending(): don't kill the commands in periph_xferq here.
The controller is handling them, calling scsipi_done() here will end up in the
xfer being scsipi_done()'ed a second time when it completes in the controller
code. In addition, the way the loop was done here would end up in an infinite
loop, because the channel kernel thread needs to run to remove a command from
this queue.
scsibusdetach(): scsipi_done() all commands from periph_xferq. The controller
is already gone, and these commands will never complete.
Shut down the channel (which will cause the kenrel thread to exit) after
detaching the childs, as they will need the kernel thread for
scsipi_wait_drain().
Fix kernel hang or deadlock when detaching devices (either by scsictl detach
or unplug) with active commands.
|
| 1.77 | 15-Nov-2001 |
lukem | branches: 1.77.16; 1.77.18;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.76 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.75 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.74 | 25-Apr-2001 |
bouyer | branches: 1.74.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.73 | 03-Apr-2000 |
enami | branches: 1.73.6;
When killing pending xfers on device detach, we can't expect scsipi_done
to remove all xfers from the pending queue. It removes only xfers for
asynchronous transactions. So, simply loop over all pending xfers
with calling scsipi_done and wait xfers to drain. Addresses PR#9703.
|
| 1.72 | 17-Mar-2000 |
soren | scsiprint() is needed even without scsibus'es, so move it to scsi_base.c .
|
| 1.71 | 17-Jan-2000 |
bouyer | Use SCSIPIRETRIES instead of hard-coded value '2' for number of retries in
common routines. Define SCSIPIRETRIES as 4, so that the command will finally
succeed after several consecutive downgrades from UDMA2 to PIO4.
|
| 1.70 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.69 | 30-Sep-1999 |
thorpej | branches: 1.69.2; 1.69.4; 1.69.6;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.68 | 28-Aug-1999 |
thorpej | If scsipi_make_xs() fails, make sure to flag the buffer as error'd,
if provided, and biodone() it. Callers of scsipi_command() expect
this to be done.
|
| 1.67 | 12-Jun-1999 |
pk | scsipi_free_xs() wants to be called at splbio; do it.
|
| 1.66 | 17-Nov-1998 |
bouyer | branches: 1.66.4; 1.66.6; 1.66.8;
Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.65 | 15-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.64 | 30-Jul-1998 |
mjacob | branches: 1.64.2;
opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.63 | 15-Jul-1998 |
mjacob | part of fix for kern/3835: use of enumerated returns from target sense handlers
|
| 1.62 | 11-Jul-1998 |
mjacob | Clarify that Sense Key 1 is a "RECOVERED ERROR"- not just an error.
|
| 1.61 | 01-Jul-1998 |
mjacob | Use new Sense Key defines (for clarity)
|
| 1.60 | 28-Mar-1998 |
christos | Split out the verbose printing of scsi messages, so that user programs
can use the printing routines.
|
| 1.59 | 05-Feb-1998 |
mrg | initial import of the new virtual memory system, UVM, into -current.
UVM was written by chuck cranor <chuck@maria.wustl.edu>, with some
minor portions derived from the old Mach code. i provided some help
getting swap and paging working, and other bug fixes/ideas. chuck
silvers <chuq@chuq.com> also provided some other fixes.
this is the rest of the MI portion changes.
this will be KNF'd shortly. :-)
|
| 1.58 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.57 | 30-Dec-1997 |
is | EACCESS->EROFS, for the write on readonly medium error.
|
| 1.56 | 23-Nov-1997 |
enami | Remove newline after printing FRU CODE instead of removing
newline before printing SKSV information.
|
| 1.55 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.54 | 08-Oct-1997 |
mjacob | Slight formatting change.
|
| 1.53 | 08-Oct-1997 |
mjacob | The bit pointer field is a 3 bit value- not 4.
|
| 1.52 | 02-Oct-1997 |
mjacob | kern/3022: make sure that printing sense info in SCSIDEBUG case starts at
the right place and contains the right number of bytes.
|
| 1.51 | 01-Oct-1997 |
mjacob | Silly boy.... don't print out sense info if the command was marked SCSI_SILENT,
and print the opcode with a leading 0x if you're printing in hex.
|
| 1.50 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.49 | 13-Sep-1997 |
enami | Declare SCSIVERBOSE by defopt in files.scsipi. Include opt_scsiverbose.h
in scsi_base.c and st.c.
|
| 1.48 | 05-Sep-1997 |
bouyer | Restore indentation of SCSIVERBOSE output.
|
| 1.47 | 27-Aug-1997 |
bouyer | branches: 1.47.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.46 | 20-Aug-1997 |
mjacob | Add in (enabled by SCSIVERBOSE option since it adds a fair amount of string
gup to the kernel) more detailed SCSI error printing, based off of final
SCSI-2 specification. This will also print out things like progress indicator
errors, and will also dump out additional sense bytes that haven't been decoded.
Also, turned the error message 'DELAYED' to 'DEFERRED' since that's more
accurate.
|
| 1.45 | 04-Aug-1997 |
fair | %x -> 0x%x
|
| 1.44 | 09-Jun-1997 |
matthias | branches: 1.44.4;
In case of a SCSI-error print the extended sense information and not
some sort of random garbage.
|
| 1.43 | 02-Apr-1997 |
mycroft | Push the buffer cleanup code into scsi_done(), and split it so that biodone()
is called *after* the driver `done' routine. This fixes disk I/O statistics
on SCSI devices.
Also, calling the `done' routine with a `complete' argument of 0 and actually
having it do anything meaningful loses in at least 3 ways, so just nuke the
argument altogether and don't call it this way. If the driver needs to do
some error handling, that's what `err_handler' is for.
|
| 1.42 | 20-Mar-1997 |
thorpej | Back out part of the previous patch - not only does stuffing a sizeof()
into 8 bits generate (bogus) warnings on some architectures, but the
change to pad on "scsi inquiry" is no longer needed, apparently.
Thanks to Matthias Pfaller for pointing the latter out to me.
|
| 1.41 | 18-Mar-1997 |
thorpej | Two changes from Ian Dall <ian.dall@dsto.defence.gov.au>:
- Deal with devices that ignore the length specified in the
inquiry command.
- Allow asynchronous requests without using a buf (key off NOSLEEP instead).
|
| 1.40 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.39 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.38 | 03-Sep-1996 |
thorpej | Store the SCSI version in the scsi_link, and set the LUN in the CDB
if the version is <= SCSI-2. This should help some older SCSI
devices that previously needed the "NOLUNS" quirk. While this is
not strictly necessary on SCSI-2 devices, the spec allows it,
so we set it for SCSI-2 devices "just in case". See section 7.2.2 of
Draft X3T9.2 Rev 10L for details.
|
| 1.37 | 05-Jul-1996 |
christos | - Apply PR/2535: Add support for flex scsi disks.
- Add a quirk called SCSI_NOMODESENSE for drives like the iomega jaz,
that don't support mode sense.
|
| 1.36 | 03-May-1996 |
christos | - SCSIDEBUG printf() fixes.
|
| 1.35 | 22-Apr-1996 |
christos | remove include of <sys/cpu.h>
|
| 1.34 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.33 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.32 | 31-Jan-1996 |
briggs | Handle COMMAND ABORTED by forcing a retry. PR#2003, but with some more
diagnostic info.
|
| 1.31 | 12-Jan-1996 |
thorpej | Handle cases like the following:
- controller calls scsi_done() with error XS_TIMEOUT
- scsi_done() calls sddone()
- sddone() calls disk_unbusy()
- scsi_done() calls controller to retry command (missing the
call to disk_busy())
- controller calls scsi_done()
- scsi_done() calls sddone()
- sddone() calls disk_busy(), which panics because of the imbalance.
Bug noticed by Leo Weppleman, who also suggested this fix; pass an additional
boolean argument ("complete") to the device's "done" routine, with a
value of `0' passed from the previous call to "done", and add an additional
call to "done" when the xfer resources are freed.
|
| 1.30 | 26-Sep-1995 |
thorpej | Don't declare Debugger(). It's handled in <sys/systm.h>.
#include <sys/systm.h> where necessary, as suggested by Jonathan Stone.
Fixes PR #1511.
|
| 1.29 | 03-May-1995 |
mycroft | Make the byte-shifting code consistent.
|
| 1.28 | 26-Jan-1995 |
mycroft | Update copyrights.
|
| 1.27 | 26-Jan-1995 |
mycroft | Redo the `wait for spin up' code. The Mach 3 method clearly doesn't work.
Also, rearrange the first open sequences a bit; SDEV_OPEN is no longer magic.
|
| 1.26 | 16-Jan-1995 |
mycroft | Trivial simplification.
|
| 1.25 | 13-Jan-1995 |
mycroft | Add a DIAGNOSTIC check.
|
| 1.24 | 30-Dec-1994 |
mycroft | Fix some oversights in the debugging code.
|
| 1.23 | 29-Dec-1994 |
mycroft | Fix a couple of things in the sense printing code.
|
| 1.22 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.21 | 01-Dec-1994 |
mycroft | Remove some redundant code.
|
| 1.20 | 01-Dec-1994 |
mycroft | Remove unneeded override of SCSI_NOSLEEP.
|
| 1.19 | 01-Dec-1994 |
mycroft | Uncomment call to scsi_user_done(), and remove a duplicate biodone() (and
another si_free()).
|
| 1.18 | 22-Nov-1994 |
deraadt | C botch repair from bostic
|
| 1.17 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.16 | 30-Aug-1994 |
deraadt | remove an un-aligned access
|
| 1.15 | 29-Jun-1994 |
deraadt | branches: 1.15.2;
knf
|
| 1.14 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.13 | 09-May-1994 |
chopps | remove union's from sense_data struct, conditionaly define RAW_PART
|
| 1.12 | 20-Apr-1994 |
mycroft | Maintain the xs free list with LIST_*().
|
| 1.11 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags. Actually sleep a bit if
there are too many commands pending on the controller.
|
| 1.10 | 11-Apr-1994 |
mycroft | Combine scsi_start_unit() and scsi_stop_unit(), and increase the start
timeout (again).
|
| 1.9 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.8 | 16-Feb-1994 |
mycroft | Increase the timeout for start_unit.
|
| 1.7 | 14-Feb-1994 |
mycroft | PARANOID --> DIAGNOSTIC for inexpensive tests.
|
| 1.6 | 01-Feb-1994 |
mycroft | Add scsi_stop_unit(), from Mac code. Other minor nits.
|
| 1.5 | 13-Dec-1993 |
mycroft | 0xfe000000 --> KERNBASE
|
| 1.4 | 27-Nov-1993 |
mycroft | Remove remaining sleep()s.
|
| 1.3 | 25-Nov-1993 |
mycroft | Patch from Julian to fix hangage when opening a device multiple times.
|
| 1.2 | 24-Nov-1993 |
mycroft | Still under construction...
|
| 1.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.15.2.1 | 06-Oct-1994 |
mycroft | Update from trunk.
|
| 1.44.4.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.47.2.5 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.47.2.4 | 16-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.47.2.3 | 06-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.47.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.47.2.1 | 27-Aug-1997 |
thorpej | file scsi_base.c was added on branch marc-pcmcia on 1997-08-27 23:33:10 +0000
|
| 1.64.2.2 | 30-Jul-1998 |
mjacob | opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.64.2.1 | 30-Jul-1998 |
mjacob | file scsi_base.c was added on branch eeh-paddr_t on 1998-07-30 00:49:21 +0000
|
| 1.66.8.1 | 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.
|
| 1.66.6.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.66.4.2 | 23-Jan-2000 |
he | Pull up revision 1.71 (requested by bouyer):
Use SCSIPIRETRIES instead of hard-coded value '2' for number of
retries in common routines, and define it as 4 so that ATAPI
command will succeed after several downgrade.
|
| 1.66.4.1 | 24-Jun-1999 |
perry | pullup 1.66->1.67 (pk): deal with missing "raise interrupt level" code
|
| 1.69.6.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.69.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.69.2.4 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.69.2.3 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.69.2.2 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.69.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.73.6.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.73.6.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.73.6.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.73.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.74.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.74.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.77.18.1 | 11-Sep-2004 |
he | Pull up revisions 1.80-1.81 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.77.16.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.77.16.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.77.16.3 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.77.16.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.77.16.1 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.85.22.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.85.22.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.85.20.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.87.48.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.88.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.88.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.89.38.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.89.34.1 | 23-May-2012 |
yamt | sync with head.
|
| 1.90.20.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.90.16.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.90.16.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.90.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.91.8.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.92.6.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.17 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.16 | 05-Sep-2003 |
mycroft | Use SMS_DBD in a several places to avoid even thinking about block descriptors.
|
| 1.15 | 02-Sep-2001 |
thorpej | branches: 1.15.20;
Use the AUDIO_PAGE definition common to SCSI and ATAPI.
|
| 1.14 | 02-Sep-2001 |
thorpej | Centralize the definition of the page code mask and page-savable bit
in the pages returned by MODE SENSE.
|
| 1.13 | 14-May-2001 |
bouyer | branches: 1.13.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.12 | 13-Feb-1998 |
enami | branches: 1.12.26;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.11 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.10 | 17-Oct-1997 |
bouyer | kern/4267 (Takeshi Nakayama): Add a 'len' parameter to cd_scsibus_get_mode and
cd_scsibus_set_mode, to specifie the real len of "data" we need. These
functions were broken when scsi_cd_write_params_page was added to
scsi_cd_pages, thus changing the len of scsi_cd_mode_data.
|
| 1.9 | 03-Oct-1997 |
thorpej | Define the "write parameters" mode page.
|
| 1.8 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.7 | 27-Aug-1997 |
bouyer | branches: 1.7.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.6 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.5 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.4 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.3 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.2 | 20-May-1993 |
cgd | branches: 1.2.3;
add rcsids and clean up file headers
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.2.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.7.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.7.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.7.2.1 | 27-Aug-1997 |
thorpej | file scsi_cd.h was added on branch marc-pcmcia on 1997-08-27 23:33:12 +0000
|
| 1.12.26.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.12.26.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.13.2.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.15.20.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.22 | 31-Dec-2021 |
andvar | fix few typos in comments, mainly in word "parameter".
|
| 1.21 | 08-Mar-2019 |
msaitoh | s/ are are / are /
s/ a a / a /
|
| 1.20 | 11-Dec-2005 |
christos | branches: 1.20.164;
merge ktrace-lwp.
|
| 1.19 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.18 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.17 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.16 | 02-Sep-2001 |
thorpej | branches: 1.16.20; 1.16.28; 1.16.30;
Centralize the definition of the page code mask and page-savable bit
in the pages returned by MODE SENSE.
|
| 1.15 | 24-Jun-2001 |
wiz | branches: 1.15.2;
Fix some typos in comments.
|
| 1.14 | 09-Sep-1999 |
thorpej | branches: 1.14.14;
Rework the changer driver a bit:
- Take note of magazine changes, and enqueue "Element Status Changed" events
that user processes can read or select on.
- Normalize some structure names.
- Report back more status about changer elements:
* Volume tags (e.g. barcode labels on the backs of your tapes)
* External device names (for drive units in a changer)
* Last element a unit of media was moved from
* Sense information for SCSI changer elements in EXCEPT condition
* Vendor-specific data if the user requests it.
- Add support for setting volume tags.
|
| 1.13 | 22-Jul-1999 |
thorpej | Define the Volume Tag format.
|
| 1.12 | 17-Dec-1998 |
gibbs | branches: 1.12.4;
Correct the definition of the changer device capabilities page. The
previous definition confused some reserved bytes for exchange capabilities.
|
| 1.11 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.10 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.9 | 29-Sep-1997 |
mjacob | Fix a misspelling.
|
| 1.8 | 27-Aug-1997 |
bouyer | branches: 1.8.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.7 | 03-Apr-1996 |
thorpej | New SCSI media changer driver, written from scratch by me. Some
influence from the old driver (written by Stefan Grefen) and from
an autochanger driver written by the Systems Programming Group at the
University of Utah Computer Science Department (currently residing
in src/sys/arch/hp300/dev/ac.c).
|
| 1.6 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.5 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.4 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.3 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.2 | 20-May-1993 |
cgd | branches: 1.2.3;
add rcsids and clean up file headers
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.2.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.8.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.8.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.8.2.1 | 27-Aug-1997 |
thorpej | file scsi_changer.h was added on branch marc-pcmcia on 1997-08-27 23:33:13 +0000
|
| 1.12.4.1 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.14.14.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.14.14.1 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.15.2.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.16.30.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.16.28.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.16.20.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.16.20.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.20.164.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.9 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.8 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.7 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.6 | 13-Feb-1998 |
enami | branches: 1.6.48; 1.6.56;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.5 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.4 | 27-Aug-1997 |
bouyer | branches: 1.4.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.3 | 04-Apr-1997 |
matthias | * Do some KNFing.
* Make all local functions static.
* Add some untested netatalk support.
|
| 1.2 | 24-Mar-1997 |
thorpej | Fix a screwup (my fault, oops) that caused the stack to get corrupted.
From Ian Dall <Ian.Dall@dsto.defence.gov.au> on port-pc532.
|
| 1.1 | 18-Mar-1997 |
thorpej | Driver for the Cabletron EA41x SCSI Ethernet Adaptor, written by
Ian Dall <ian.dall@dsto.defence.gov.au>. Converted to "new arp"
and some other (very) minor changes by me.
|
| 1.4.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.4.2.1 | 27-Aug-1997 |
thorpej | file scsi_ctron_ether.h was added on branch marc-pcmcia on 1997-08-27 23:33:14 +0000
|
| 1.6.56.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.6.48.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.38 | 27-Feb-2025 |
jakllsch | sd(4): Translate discard to UNMAP (10) command
|
| 1.37 | 27-Feb-2025 |
jakllsch | more whitespace fixes
|
| 1.36 | 27-Feb-2025 |
jakllsch | fix comment grammar and whitespace
|
| 1.35 | 07-Dec-2023 |
andvar | branches: 1.35.2;
s/multiplcation/multiplication/ in comment.
|
| 1.34 | 10-Nov-2021 |
msaitoh | s/desciptor/descriptor/ in comment.
|
| 1.33 | 27-Dec-2019 |
msaitoh | s/transfered/transferred/
|
| 1.32 | 19-Nov-2016 |
flxd | branches: 1.32.16;
Add "getrealloc" and "setrealloc" commands to get/set automatic reallocation
parameters/enables for error recovery, similar to {get,set}cache.
Many old SCSI disks shipped with reallocation disabled, albeit supporting it.
Minor (cosmetic) fixup of scsi_disk_pages while there.
Based upon code in PR bin/29165 by Greg A. Woods.
OK christos@
|
| 1.31 | 11-Dec-2005 |
christos | branches: 1.31.116; 1.31.120; 1.31.138; 1.31.140; 1.31.144; 1.31.146;
merge ktrace-lwp.
|
| 1.30 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.29 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.28 | 07-Jan-2005 |
ginsbach | branches: 1.28.2;
Add command to read the defect data from a direct access device.
|
| 1.27 | 07-Dec-2004 |
thorpej | Define SYNCHRONIZE CACHE (16).
|
| 1.26 | 07-Dec-2004 |
thorpej | SYNCHRONIZE_CACHE -> SYNCHRONIZE_CACHE_10
|
| 1.25 | 07-Dec-2004 |
thorpej | USe more appropriate macro/struct names for READ/WRITE (6) and
READ/WRITE (10).
|
| 1.24 | 05-Sep-2003 |
mycroft | Forgot to commit this. Add the "rpm" field to the flex geometry page.
|
| 1.23 | 04-Feb-2003 |
wrstuden | branches: 1.23.2;
Add defines and structure for control mode page.
|
| 1.22 | 06-Jan-2003 |
matt | Add multiple inclusion protection.
|
| 1.21 | 24-Jul-2002 |
mjacob | Fixes kern/17573: correctly describe the bfi && physical sector defect list
formats.
Obtained from PR submitter.
|
| 1.20 | 03-Sep-2001 |
thorpej | branches: 1.20.12;
Define more caching parameters from SCSI-3.
|
| 1.19 | 02-Sep-2001 |
thorpej | Oops, name the caching parameters page structure member.
|
| 1.18 | 01-Sep-2001 |
thorpej | Add the caching parameters mode page.
|
| 1.17 | 30-Oct-1998 |
thorpej | branches: 1.17.24; 1.17.26;
Define the FORMAT UNIT command, and several related structures.
|
| 1.16 | 09-Jun-1998 |
thorpej | Define the SYNCHRONIZE CACHE command.
|
| 1.15 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.14 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.13 | 03-Oct-1997 |
thorpej | Define the REZERO UNIT command.
|
| 1.12 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.11 | 27-Aug-1997 |
bouyer | branches: 1.11.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.10 | 05-Jul-1996 |
christos | - Apply PR/2535: Add support for flex scsi disks.
- Add a quirk called SCSI_NOMODESENSE for drives like the iomega jaz,
that don't support mode sense.
|
| 1.9 | 19-Mar-1996 |
mycroft | branches: 1.9.4;
Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.8 | 15-Oct-1995 |
thorpej | Add missing spindle synch control, rotational offset, and RPM members
to the rigid geometry mode page. These missing members are at the
end of the page, in space wich was previously marked "reserved".
|
| 1.7 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.6 | 22-Nov-1994 |
deraadt | C botch repair from bostic
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.3 | 20-May-1993 |
cgd | branches: 1.3.3;
add rcsids and clean up file headers
|
| 1.2 | 12-Apr-1993 |
deraadt | new scsi subsystem.
changes also in config/mkioconf.c
i386/isa/wd.c, fd.c, and all scsi drivers.
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.3.3.2 | 01-Feb-1994 |
mycroft | Add scsi_stop_unit(), from Mac code. Other minor nits.
|
| 1.3.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.9.4.1 | 04-Mar-1997 |
mycroft | Pull up support for fetching geometry info on optical disks.
|
| 1.11.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.11.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.11.2.1 | 27-Aug-1997 |
thorpej | file scsi_disk.h was added on branch marc-pcmcia on 1997-08-27 23:33:15 +0000
|
| 1.17.26.2 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.17.26.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.17.24.3 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.17.24.2 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.17.24.1 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.20.12.1 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.23.2.5 | 17-Jan-2005 |
skrll | Sync with HEAD.
|
| 1.23.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.23.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.23.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.23.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.28.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.31.146.1 | 18-Jan-2017 |
skrll | Sync with netbsd-5
|
| 1.31.144.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.31.140.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.31.138.1 | 03-Dec-2016 |
martin | Pull up following revision(s) (requested by flxd in ticket #1276):
sys/dev/scsipi/scsi_disk.h: revision 1.32
sbin/scsictl/scsictl.8: revision 1.27-1.30
sbin/scsictl/scsictl.c: revision 1.39
Add "getrealloc" and "setrealloc" commands to get/set automatic reallocation
parameters/enables for error recovery, similar to {get,set}cache.
Many old SCSI disks shipped with reallocation disabled, albeit
supporting it.
Minor (cosmetic) fixup of scsi_disk_pages while there.
Based upon code in PR bin/29165 by Greg A. Woods.
OK christos@
Bump date for previous.
Use more and more appropriate markup while here.
Add crossreference to scsi(4) per note in PR 9627.
Whitespace, sort.
|
| 1.31.120.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.31.116.1 | 15-Jan-2017 |
snj | Pull up following revision(s) (requested by flxd in ticket #1418):
sbin/scsictl/scsictl.8: 1.27, 1.29, 1.30 via patch
sbin/scsictl/scsictl.c: revision 1.39 via patch
sys/dev/scsipi/scsi_disk.h: revision 1.32
Add "getrealloc" and "setrealloc" commands to get/set automatic reallocation
parameters/enables for error recovery, similar to {get,set}cache.
Many old SCSI disks shipped with reallocation disabled, albeit
supporting it.
Minor (cosmetic) fixup of scsi_disk_pages while there.
Based upon code in PR bin/29165 by Greg A. Woods.
OK christos@
--
Bump date for previous.
--
Add crossreference to scsi(4) per note in PR 9627.
Whitespace, sort.
|
| 1.32.16.1 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.35.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.5 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.4 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.3 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.2 | 19-Apr-2003 |
fvdl | branches: 1.2.2; 1.2.10;
Add NetBSD RCS id.
|
| 1.1 | 19-Apr-2003 |
fvdl | IU pkt/task definitions include file. From FreeBSD.
|
| 1.2.10.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.2.2.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.17 | 26-Aug-2006 |
bjh21 | QAS REQUEST is a one-byte message. Make MSG_IS1BYTE recognise this.
|
| 1.16 | 11-Dec-2005 |
christos | branches: 1.16.4; 1.16.8;
merge ktrace-lwp.
|
| 1.15 | 27-Feb-2005 |
perry | branches: 1.15.4;
nuke trailing whitespace
|
| 1.14 | 20-Apr-2003 |
fvdl | branches: 1.14.2; 1.14.10; 1.14.12;
One more FreeBSD compat define.
|
| 1.13 | 20-Apr-2003 |
fvdl | Correct typo in previous.
|
| 1.12 | 20-Apr-2003 |
fvdl | Add FreeBSD compat define for PPR_PCOM
|
| 1.11 | 19-Apr-2003 |
fvdl | Add MSG_QAS_REQUEST definition.
|
| 1.10 | 06-Dec-2002 |
fvdl | Remove duplicate define that crept in.
|
| 1.9 | 06-Dec-2002 |
fvdl | Sync this with the FreeBSD version, adding a few aliases and new definitions.
|
| 1.8 | 23-Apr-2002 |
bouyer | Add the Parallel Protocol Request message.
|
| 1.7 | 04-Nov-2001 |
tsutsui | Add common macro to check message length.
|
| 1.6 | 03-Nov-2001 |
tsutsui | Replace a magic number with macro.
|
| 1.5 | 25-Apr-2001 |
bouyer | branches: 1.5.2; 1.5.6;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.4 | 15-Mar-2000 |
fvdl | branches: 1.4.6;
Add a couple of defines/aliases for messages.
|
| 1.3 | 05-Jan-1998 |
perry | branches: 1.3.14;
RCSID Police.
|
| 1.2 | 08-Oct-1996 |
gibbs | branches: 1.2.2;
dev/microcode/aic7xxx/aic7xxx.seq,
dev/microcode/aic7xxx_seq.h,
dev/ic/aic7xxxreg.h:
Remove intrinsic knowledge about SDTR and WDTR messages and replace it
with a generic message system that allows the kernel driver to handle
SDTR, WDTR and any other type of extended message it chooses too. This
makes the sequencer code much simpler, makes extended message handling
debuggable since the bulk of the work is in the kernel driver, and saves
lots of instruction space.
Regen microcode header file.
dev/ic/aic7xxx.c, dev/ic/aic7xxxvar.h:
Add code to handle WDTR and SDTR negotiation in light of the changes in
the message interface to the sequencer. Don't reject targets that
negotiate async by sending an SDTR with a 0 offset. Use an sdtr message
with 0,0 to negotiate async when a target suggests a period that is too
long for us to handle. Some tape and cdrom drives don't like us doing
the message reject that we did in the past.
Fix a problem with handing the QUEUE FULL condition.
Fix a race condition (most likely the cause of the SCB paging problems) that
might allow the sequencer to get unpaused before the condition that caused
it to be paused (a SEQINT) was handled.
Race condition pointed out by Doug Ledford <dledford@dialnet.net> and
by "Dan Willis" <dan@plutotech.com>.
dev/pci/ahc_pci.c:
Add support for the 2940AU, an aic7860 based controller.
dev/pci/pcidevs.h, dev/pci/pcidevs_data.h:
Add product IDs for the 2940AU, aic7860 and aic7855.
Regen data file.
scsi/scsi_message.h:
Add MSG_EXT_SDTR_LEN and MSG_EXT_WDTR_LEN - the length of bytes in these
extended messages.
Thanks to Chuck Cranor <chuck@maria.wustl.edu> for testing these changes
out for me.
|
| 1.1 | 28-Dec-1994 |
mycroft | branches: 1.1.6;
Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.1.6.1 | 04-Mar-1997 |
mycroft | Add a few constants for the ahc driver.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.3.14.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.4.6.4 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.4.6.3 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.4.6.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.4.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.5.6.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 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.
|
| 1.14.12.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.14.10.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.14.2.1 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.15.4.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.16.8.1 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.16.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.11 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.10 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.9 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.8 | 07-Dec-2004 |
thorpej | branches: 1.8.2;
USe more appropriate macro/struct names for READ/WRITE (6) and
READ/WRITE (10).
|
| 1.7 | 19-Nov-2001 |
soren | branches: 1.7.16;
Remove duplicate struct scsipi_start_stop.
|
| 1.6 | 13-Feb-1998 |
enami | branches: 1.6.26; 1.6.28;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.5 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.4 | 05-Jan-1998 |
perry | RCSID Police.
|
| 1.3 | 27-Aug-1997 |
bouyer | branches: 1.3.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.2 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.1 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.3.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.3.2.1 | 27-Aug-1997 |
thorpej | file scsi_scanner.h was added on branch marc-pcmcia on 1997-08-27 23:33:17 +0000
|
| 1.6.28.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.6.26.1 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.7.16.1 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.8.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.1 | 19-Feb-2020 |
riastradh | branches: 1.1.2; 1.1.6;
Sprinkle some dtrace probes into scsi(4).
|
| 1.1.6.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.1.6.1 | 19-Feb-2020 |
martin | file scsi_sdt.h was added on branch phil-wifi on 2020-04-08 14:08:12 +0000
|
| 1.1.2.2 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.1.2.1 | 19-Feb-2020 |
ad | file scsi_sdt.h was added on branch ad-namecache on 2020-02-29 20:19:16 +0000
|
| 1.8 | 09-Feb-2024 |
andvar | fix spelling mistakes, mainly in comments and log messages.
|
| 1.7 | 27-Jan-2022 |
jakllsch | Try REPORT LUNS command to enumerate logical units.
|
| 1.6 | 28-Mar-2019 |
kardel | Add reading of supported opcodes and their timeouts
at attachment time. Though this information is optional,
it allows to override our fixed timeouts with device
provided timeouts. These timeouts will override the
hardcoded values if the device provided timeouts
exceed the hardcoded values and are less than a day.
Using the device provided timeouts avoids premature
device resets and unreliable operation due to
inadequate timeouts.
Due to the limited implementations of USB
umass devices this feature is disabled for all
umass attached devices.
|
| 1.5 | 06-Feb-2010 |
cegger | branches: 1.5.64;
scsi_sense_data: add comments taken from src/sys/dev/ieee1394/sbp.h
|
| 1.4 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.3 | 11-Dec-2005 |
christos | branches: 1.3.70; 1.3.72; 1.3.74;
merge ktrace-lwp.
|
| 1.2 | 27-Feb-2005 |
perry | branches: 1.2.2; 1.2.6; 1.2.8;
nuke trailing whitespace
|
| 1.1 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.2.8.2 | 29-Apr-2005 |
kent | sync with -current
|
| 1.2.8.1 | 27-Feb-2005 |
kent | file scsi_spc.h was added on branch kent-audio2 on 2005-04-29 11:29:16 +0000
|
| 1.2.6.2 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.2.6.1 | 27-Feb-2005 |
yamt | file scsi_spc.h was added on branch yamt-km on 2005-03-19 08:35:47 +0000
|
| 1.2.2.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.2.2.1 | 27-Feb-2005 |
skrll | file scsi_spc.h was added on branch ktrace-lwp on 2005-03-04 16:50:33 +0000
|
| 1.3.74.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.3.74.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.3.72.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.3.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.5.64.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.1 | 14-Apr-2022 |
pgoyette | Split some common stuff into scsi_subr module. This enables loading
of the iscsi module whether or not there are any scsi things built
into the kernel.
Addresses the iscsi portion of kern/56772
|
| 1.25 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.24 | 11-Dec-2005 |
christos | branches: 1.24.70; 1.24.72; 1.24.74;
merge ktrace-lwp.
|
| 1.23 | 28-Feb-2005 |
soren | Remove a bit of old cruft.
|
| 1.22 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.21 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.20 | 25-Apr-2001 |
bouyer | branches: 1.20.22; 1.20.30; 1.20.32;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.19 | 21-Feb-2000 |
mjacob | branches: 1.19.6;
Patches from msouth@scruz.net to handle QIC 3220 (HP T20).
|
| 1.18 | 30-Sep-1999 |
soren | branches: 1.18.2;
Remove outdated list of density codes. mt(1) has helpful
hints and the SCSI specification has the real, current list.
|
| 1.17 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.16 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.15 | 06-Aug-1998 |
drochner | fix incorrect length of LOCATE CDB (reported in PR kern/5306 by
Dave Huang <khym@bga.com>)
|
| 1.14 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.13 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.12 | 01-Oct-1997 |
enami | branches: 1.12.2;
Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.11 | 29-Sep-1997 |
mjacob | Add SCSI-3 DATA COMPRESSION page structure.
Add READ POSITION command structure. Add READ_POSITION and LOCATE opcodes.
|
| 1.10 | 27-Aug-1997 |
bouyer | branches: 1.10.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.9 | 24-May-1996 |
thorpej | Implement ST_Q_UNIMODAL quirk, which indicates that a drive will reject
all attempts to mode select, even if the attempted mode is supported.
Add the ST_Q_UNIMODAL quirk for the HP Colorado T4000s which exhibits
this behavior. (Someone please lob a tactical nuke in that direction.)
From David Rosenthal <dshr@vitria.com> on netbsd-bugs.
|
| 1.8 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.7 | 11-Jan-1996 |
thorpej | Honor cache request and implement erase command, and add the SCSI tape
device configuration page. Fixes PRs 807, 1201, and 1705.
From John Kohl <jtk@kolvir.blrc.ma.us>.
|
| 1.6 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.5 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.4 | 13-Apr-1994 |
mycroft | Various cleanup, but no functional differences.
|
| 1.3 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.2 | 20-May-1993 |
cgd | branches: 1.2.3;
add rcsids and clean up file headers
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.2.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.10.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.10.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.10.2.1 | 27-Aug-1997 |
thorpej | file scsi_tape.h was added on branch marc-pcmcia on 1997-08-27 23:33:18 +0000
|
| 1.12.2.1 | 30-Sep-1998 |
cgd | pull up rev 1.15 from trunk (drochner)
|
| 1.18.2.1 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.19.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.20.32.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.20.30.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.20.22.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.20.22.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.24.74.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.24.72.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.24.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.9 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.8 | 15-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.7 | 30-Jul-1998 |
mjacob | branches: 1.7.2;
opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.6 | 11-Jul-1998 |
mjacob | Clarify that Sense Key 1 is a "RECOVERED ERROR"- not just an error.
|
| 1.5 | 10-Jul-1998 |
mjacob | hmm, char sign extension bit me
|
| 1.4 | 01-Jul-1998 |
mjacob | Align output for Tape specific bits and use new Sense Key defines (for clarity)
|
| 1.3 | 24-Jun-1998 |
mjacob | When printing an error, print more than just the opcode. Print the
entire CDB, whose length is actually deterministic. This makes a *big*
difference when actually trying to bind errors to commands.
|
| 1.2 | 15-Apr-1998 |
mjacob | a tab crept in
|
| 1.1 | 28-Mar-1998 |
christos | Split out the verbose printing of scsi messages, so that user programs
can use the printing routines.
|
| 1.7.2.2 | 30-Jul-1998 |
mjacob | opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.7.2.1 | 30-Jul-1998 |
mjacob | file scsi_verbose.c was added on branch eeh-paddr_t on 1998-07-30 00:49:22 +0000
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsicd.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
bouyer | Moved atapicd and scsicd to cd_scsi and cd_atapi, to be concistent with
other bus front-end naming conventions (Suggested by Mike Long)
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.306 | 22-Nov-2024 |
mlelstv | The code tried to limit number of LUNs per target to 3, but would
only default to a single LUN when that limit is exceeded.
With the limit removed, more LUNs will be attached (up to the limit
imposed by the host adapter driver).
|
| 1.305 | 29-Sep-2024 |
nat | Add quirk for the standard PiSCSI cdrom emulation.
|
| 1.304 | 22-Jun-2024 |
palle | branches: 1.304.2;
Add quirk for sparc64/sun4v ldom virtual cd devices
|
| 1.303 | 15-Oct-2022 |
jmcneill | Add PQUIRK_ONLYBIG for Oracle OCI BlockVolumes.
Oracle cloud BlockVolumes do not appear to support SCSI READ6 or WRITE6
commands, so set PQUIRK_ONLYBIG to avoid it here.
|
| 1.302 | 14-Apr-2022 |
pgoyette | Split some common stuff into scsi_subr module. This enables loading
of the iscsi module whether or not there are any scsi things built
into the kernel.
Addresses the iscsi portion of kern/56772
|
| 1.301 | 09-Apr-2022 |
riastradh | sys: Use membar_release/acquire around reference drop.
This just goes through my recent reference count membar audit and
changes membar_exit to membar_release and membar_enter to
membar_acquire -- this should make everything cheaper on most CPUs
without hurting correctness, because membar_acquire is generally
cheaper than membar_enter.
|
| 1.300 | 12-Mar-2022 |
riastradh | scsi(9): Handle bogus number of LUNs in SCSI_REPORT_LUNS.
Reported-by: syzbot+76ef9084533d4bccec66@syzkaller.appspotmail.com
|
| 1.299 | 12-Mar-2022 |
riastradh | sys: Membar audit around reference count releases.
If two threads are using an object that is freed when the reference
count goes to zero, we need to ensure that all memory operations
related to the object happen before freeing the object.
Using an atomic_dec_uint_nv(&refcnt) == 0 ensures that only one
thread takes responsibility for freeing, but it's not enough to
ensure that the other thread's memory operations happen before the
freeing.
Consider:
Thread A Thread B
obj->foo = 42; obj->baz = 73;
mumble(&obj->bar); grumble(&obj->quux);
/* membar_exit(); */ /* membar_exit(); */
atomic_dec -- not last atomic_dec -- last
/* membar_enter(); */
KASSERT(invariant(obj->foo,
obj->bar));
free_stuff(obj);
The memory barriers ensure that
obj->foo = 42;
mumble(&obj->bar);
in thread A happens before
KASSERT(invariant(obj->foo, obj->bar));
free_stuff(obj);
in thread B. Without them, this ordering is not guaranteed.
So in general it is necessary to do
membar_exit();
if (atomic_dec_uint_nv(&obj->refcnt) != 0)
return;
membar_enter();
to release a reference, for the `last one out hit the lights' style
of reference counting. (This is in contrast to the style where one
thread blocks new references and then waits under a lock for existing
ones to drain with a condvar -- no membar needed thanks to mutex(9).)
I searched for atomic_dec to find all these. Obviously we ought to
have a better abstraction for this because there's so much copypasta.
This is a stop-gap measure to fix actual bugs until we have that. It
would be nice if an abstraction could gracefully handle the different
styles of reference counting in use -- some years ago I drafted an
API for this, but making it cover everything got a little out of hand
(particularly with struct vnode::v_usecount) and I ended up setting
it aside to work on psref/localcount instead for better scalability.
I got bored of adding #ifdef __HAVE_ATOMIC_AS_MEMBAR everywhere, so I
only put it on things that look performance-critical on 5sec review.
We should really adopt membar_enter_preatomic/membar_exit_postatomic
or something (except they are applicable only to atomic r/m/w, not to
atomic_load/store_*, making the naming annoying) and get rid of all
the ifdefs.
|
| 1.298 | 05-Feb-2022 |
hannken | Initialize "replun" -- found with KMSAN.
|
| 1.297 | 29-Jan-2022 |
martin | In some cases the gcc optimizer is not smart enough to figure out why
the luns and nluns variables are never actually used when they are not
initialized - so initialize them always.
|
| 1.296 | 28-Jan-2022 |
christos | Factor out the lun detection code to simplify control flow.
|
| 1.295 | 28-Jan-2022 |
jakllsch | shut up GCC about possibly-uninit; some KNF
|
| 1.294 | 27-Jan-2022 |
jakllsch | Try REPORT LUNS command to enumerate logical units.
|
| 1.293 | 21-Dec-2021 |
riastradh | scsi(4): Take kernel lock around entry into autoconf.
This code paths is entered by kthreads marked MP-safe, not just from
autoconf.
I'm not sure this is sufficient -- it's not clear to me whether
anything prevents concurrently scanning the same target. Someone with
a better understanding of scsi(4) locking will have to audit this.
(For example, maybe it is guaranteed only to happen only either (a)
in autoconf, or (b) in a thread that doesn't start until autoconf is
done. But I don't know -- and if it is this, it should be asserted
so we can verify it.)
|
| 1.292 | 07-Aug-2021 |
thorpej | Merge thorpej-cfargs2.
|
| 1.291 | 24-Apr-2021 |
thorpej | branches: 1.291.8;
Merge thorpej-cfargs branch:
Simplify and make extensible the config_search() / config_found() /
config_attach() interfaces: rather than having different variants for
which arguments you want pass along, just have a single call that
takes a variadic list of tag-value arguments.
Adjust all call sites:
- Simplify wherever possible; don't pass along arguments that aren't
actually needed.
- Don't be explicit about what interface attribute is attaching if
the device only has one. (More simplification.)
- Add a config_probe() function to be used in indirect configuiration
situations, making is visibly easier to see when indirect config is
in play, and allowing for future change in semantics. (As of now,
this is just a wrapper around config_match(), but that is an
implementation detail.)
Remove unnecessary or redundant interface attributes where they're not
needed.
There are currently 5 "cfargs" defined:
- CFARG_SUBMATCH (submatch function for direct config)
- CFARG_SEARCH (search function for indirect config)
- CFARG_IATTR (interface attribte)
- CFARG_LOCATORS (locators array)
- CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles)
...and a sentinel value CFARG_EOL.
Add some extra sanity checking to ensure that interface attributes
aren't ambiguous.
Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark
ports to associate those device handles with device_t instance. This
will trickle trough to more places over time (need back-end for pre-OFW
Sun OBP; any others?).
|
| 1.290 | 18-Sep-2020 |
jakllsch | branches: 1.290.4;
Revert scsiconf.c 1.288, it only worked for LUN 1.
vioscsi(4) now sets PQUIRK_FORCELUNS, which fixes the original issue for
all LUNs.
To-do: should issue REPORT LUNS and use the information it returns to
probe LUNs in an optimized way.
|
| 1.289 | 17-Sep-2020 |
jakllsch | Some misspelling-in-comments fixes for scsipi
|
| 1.288 | 11-Jul-2020 |
kim | Continue scanning a SCSI bus when a LUN is reported not present
This fixes disk attachment under Qemu when there is no disk on LUN 0 on
a SCSI bus but there is a disk on LUN 1. The inquiry for LUN 0 returns
SID_QUAL_LU_NOTPRESENT & T_NODEVICE. Quirks are only checked if neither
one of those are set, so cannot use a quirk entry.
Use case 1: Proxmox 6 configures each disk on its own bus when using
the "Virtio SCSI single" SCSI controller. However, while the "scsi0"
disk is on LUN 0, the "scsi1" disk is on LUN 1.
Use case 2: A Linode boot profile with multiple disks results in
the first disk ("sda") on LUN 1, while the second disk ("sdb") is
on LUN 0, each on their own bus.
|
| 1.287 | 02-May-2020 |
jdc | Don't attempt to read opcodes and their timeouts at attach time for
old devices. The MAINTENANCE IN command was introduced with SCSI-3
and sending it to older peripherals can cause timeouts or them not
to respond to further requests.
|
| 1.286 | 19-Feb-2020 |
riastradh | C99 initializers for scsipi_bustype. No functional change intended.
|
| 1.285 | 10-Nov-2019 |
chs | branches: 1.285.2;
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.
|
| 1.284 | 28-Mar-2019 |
kardel | branches: 1.284.4;
Add reading of supported opcodes and their timeouts
at attachment time. Though this information is optional,
it allows to override our fixed timeouts with device
provided timeouts. These timeouts will override the
hardcoded values if the device provided timeouts
exceed the hardcoded values and are less than a day.
Using the device provided timeouts avoids premature
device resets and unreliable operation due to
inadequate timeouts.
Due to the limited implementations of USB
umass devices this feature is disabled for all
umass attached devices.
|
| 1.283 | 12-Jan-2019 |
tsutsui | Add NOLUNS quirk for more SEAGATE SCA/WIDE drives.
Tested on NetBSD/luna68k and LUNA with SCA 80pin -> NARROW 50pin and
WIDE 68pin -> NARROW 50pin connectors.
|
| 1.282 | 07-Oct-2018 |
christos | Handle the SATA to USB external enclosure sold by "Sabrent" and
made by JMicro (vendor=0x152d product=0x0578). This bridge does
not understand FUA, so add a quirk for it.
|
| 1.281 | 01-Sep-2018 |
mlelstv | Wait in detach if the discovery thread is still running. Avoids crashes
when a device is attached/detached rapidly.
|
| 1.280 | 17-Jun-2017 |
mlelstv | branches: 1.280.4; 1.280.6;
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.279 | 18-Mar-2017 |
tsutsui | branches: 1.279.6;
Add NOLUNS quirk for SEAGATE ST39236LC disk drives.
Found on "SCSI drive conservation activity" for poor Tier-II machines.
|
| 1.278 | 01-Dec-2016 |
mlelstv | branches: 1.278.2;
CID 1396620: Null pointer dereferences
|
| 1.277 | 29-Nov-2016 |
mlelstv | reference count adapter mutex possibly shared by multiple channels.
fix error in atapibusdetach, when a child device cannot be detached,
keep atapibus instance alive.
|
| 1.276 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.275 | 26-Jun-2016 |
mlelstv | branches: 1.275.2;
Create a dedicated thread for the initial scsibus discovery instead
of using the completion thread. This prevents a deadlock when a
command fails during discovery which needs to be handled by the
completion thread.
|
| 1.274 | 02-May-2016 |
christos | move scsipi_strvis -> libkern:strnvisx()
change the prototype to match userland
fix sizes of strings passed to it
|
| 1.273 | 25-Jul-2014 |
dholland | branches: 1.273.4;
Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.272 | 16-Mar-2014 |
dholland | branches: 1.272.2;
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.
|
| 1.271 | 12-Oct-2013 |
christos | Pass the device name in, so we can debug what deferred drivers did not work.
|
| 1.270 | 15-Sep-2013 |
martin | Remove unused variable
|
| 1.269 | 21-Aug-2012 |
bouyer | branches: 1.269.2; 1.269.4;
If the controller supports more than 256 commands per target,
clamp it to 256 (maximum number of tags in SCSI). Newer controllers
(such as mpii(4), and mfi(4) when fixed to announce tagged queuing support)
support more than 256 outstanding commands and don't use the scsi tag,
but at this time scsipi will always allocate a tag, and panic if a periph
tries to send more than 256 commands.
|
| 1.268 | 13-May-2012 |
jakllsch | Fix locking issue triggered by drvctr -r of a scsibus(4).
Pullup to netbsd-6 should be requested.
|
| 1.267 | 20-Apr-2012 |
bouyer | Add a bustype_async_event_xfer_mode() callback to scsipi_bustype (which can
be NULL), so that transport-specific details of transfer mode setting/printing
can be handled more easily.
Move scsipi_async_event_xfer_mode() and scsipi_print_xfer_mode() to
scsi_base.c and split in parallel scsi and FC/SAS parts.
size of struct scsipi_bustype has changed, welcome to 6.99.5
|
| 1.266 | 19-Apr-2012 |
bouyer | Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.265 | 06-Apr-2012 |
christos | Add a quirk for the Apple iPod whose mode sense commands fails with not ready.
Seems to work just fine if we send a start command first...
|
| 1.264 | 12-Mar-2012 |
mrg | take the kernel lock a few more places when doing detach, to avoid
triggering KERNEL_LOCK_P() asserts in both scsi and usb code.
with this and other recent fixes i can now "drvctl -d ehci0".
|
| 1.263 | 11-Mar-2012 |
mrg | scsidevdetached ioctl path enters scsipi code without kernel lock
and this upsets the newer kasserts. take kernel lock here.
|
| 1.262 | 26-Apr-2011 |
hannken | branches: 1.262.4; 1.262.8; 1.262.10;
Fixup previous.
The bug was in scsibusdetach(), which is not doing things in the proper
order: it has to detach its children and check for error. If no error,
then it can release the resources that the children were using.
From David Young via source-changes-d.
|
| 1.261 | 25-Apr-2011 |
hannken | Don't kill outstanding requests when detaching a scsibus on shutdown.
Both the controller and tyhe targets are still running.
|
| 1.260 | 18-Apr-2011 |
rmind | Replace few simple_lock and ltsleep/wakeup uses with mutex(9) and condvar(9).
Note to all: please replace old primitives in your code! Thanks.
|
| 1.259 | 02-Apr-2011 |
macallan | Add a quirks entry for Seagate SX173404LC drives, now they will work at higher
speeds than 8bit/async
While there, also disable sync for ZIP drives - at least some of them will
pretend to support sync and then act up.
|
| 1.258 | 07-Jun-2010 |
pgoyette | branches: 1.258.2;
Update scsiverbose module to use module_autoload() rather than module_load().
Load the module right before each attempt to use its features, and let the
module subsystem handle unloading.
|
| 1.257 | 30-May-2010 |
pgoyette | Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.256 | 27-Apr-2010 |
dyoung | For clarity of scsidevdetached(), rename some variables: sc ->
self, dev -> child, ssc -> sc.
|
| 1.255 | 12-Nov-2009 |
dyoung | branches: 1.255.2; 1.255.4;
Remove superfluous activation hooks.
|
| 1.254 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.253 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.252 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.251 | 07-Apr-2009 |
dyoung | Detach atapibus(4), scsibus(4), cd(4), and sd(4) during shutdown.
Destroy sd->sc_callout in sddetach(). Delete some dead code in
cddetach().
|
| 1.250 | 16-Jul-2008 |
drochner | branches: 1.250.2; 1.250.8;
split device/softc for scsibus
|
| 1.249 | 03-Jul-2008 |
hannken | branches: 1.249.2;
The Intel SCA backplane "ESG-SHV" does not support logical units.
|
| 1.248 | 08-Jun-2008 |
tsutsui | branches: 1.248.2;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.247 | 28-Apr-2008 |
martin | branches: 1.247.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.246 | 05-Apr-2008 |
cegger | branches: 1.246.2; 1.246.4;
use aprint_*_dev and device_xname
|
| 1.245 | 04-Jan-2008 |
ad | branches: 1.245.6;
Start detangling lock.h from intr.h. This is likely to cause short term
breakage, but the mess of dependencies has been regularly breaking the
build recently anyhow.
|
| 1.244 | 09-Dec-2007 |
jmcneill | branches: 1.244.2;
Merge jmcneill-pm branch.
|
| 1.243 | 04-Mar-2007 |
christos | branches: 1.243.14; 1.243.16; 1.243.22; 1.243.24; 1.243.26;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.242 | 17-Feb-2007 |
bouyer | The SEAGATE ST318203LSUN18G announces DT only capabilitie although it also
supports ST. Because of this the HBA doesn't initiate sync/wide negotiation,
so add PQUIRK_CAP_SYNC | PQUIRK_CAP_WIDE16 for this drive; fix
problem reported by Havard Eidnes.
|
| 1.241 | 30-Nov-2006 |
christos | branches: 1.241.2; 1.241.4;
Add a quirk for Initio drives (from Rhialto)
|
| 1.240 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.239 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.238 | 03-Sep-2006 |
christos | branches: 1.238.2; 1.238.4;
add missing initializer
|
| 1.237 | 29-Aug-2006 |
bjh21 | The Fujitsu drive in my A540 supports sync transfers but apparently
doesn't advertise this in INQUIRY. It makes it nice and obvious with
SDTR, though, so add it to the list.
|
| 1.236 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.235 | 29-Mar-2006 |
thorpej | Replace device_locators() with device_locator(), and use it.
|
| 1.234 | 11-Dec-2005 |
christos | branches: 1.234.4; 1.234.6; 1.234.8; 1.234.10; 1.234.12;
merge ktrace-lwp.
|
| 1.233 | 26-Nov-2005 |
tsutsui | Fix typo (FALLTHOUGH -> FALLTHROUGH). From Jeff Ito.
|
| 1.232 | 25-Aug-2005 |
drochner | branches: 1.232.6;
kill a number of autoconf submatch functions which follow the
standard scheme:
if (<configured> != <wildcard> && <configured> != <real>)
then fail
else
ask device match function
This is handled by config_stdsubmatch() now.
|
| 1.231 | 25-Aug-2005 |
drochner | replace the "locdesc_t" structure carrying the number of locators
explicitely by a plain integer array
the length in now known to all relevant parties, so this avoids
duplication of information, and we can allocate that thing in
drivers without hacks
|
| 1.230 | 30-May-2005 |
christos | branches: 1.230.2;
- remove bogus casts
- add more const
|
| 1.229 | 27-Feb-2005 |
perry | branches: 1.229.2;
nuke trailing whitespace
|
| 1.228 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.227 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.226 | 21-Aug-2004 |
thorpej | branches: 1.226.4; 1.226.6;
Use ANSI function decls and make use of static.
|
| 1.225 | 18-Aug-2004 |
drochner | Use the new autoconf functions to rescan busses and detach devices
on user request.
This duplicates the functionality provided by a private ioctl
interface (accessible through scsictl(8)), but in a more generic way.
|
| 1.224 | 12-Aug-2004 |
mycroft | Whoops. Swap the order of the delref() and the config_pending_decr().
|
| 1.223 | 10-Aug-2004 |
mycroft | Hold a reference to the adapter until scsibus_config() has been called. This
avoids an extra enable/disable cycle on removable controllers (i.e. PCMCIA).
|
| 1.222 | 06-Aug-2004 |
bouyer | Ops, don't return before the end of the function.
Catched by enami tsugutomo.
|
| 1.221 | 05-Aug-2004 |
bouyer | scsi_kill_pending(): don't kill the commands in periph_xferq here.
The controller is handling them, calling scsipi_done() here will end up in the
xfer being scsipi_done()'ed a second time when it completes in the controller
code. In addition, the way the loop was done here would end up in an infinite
loop, because the channel kernel thread needs to run to remove a command from
this queue.
scsibusdetach(): scsipi_done() all commands from periph_xferq. The controller
is already gone, and these commands will never complete.
Shut down the channel (which will cause the kenrel thread to exit) after
detaching the childs, as they will need the kernel thread for
scsipi_wait_drain().
Fix kernel hang or deadlock when detaching devices (either by scsictl detach
or unplug) with active commands.
|
| 1.220 | 12-Mar-2004 |
bouyer | branches: 1.220.4;
Add a drive with broken tagged queuing support. From Jim Faulkner.
Fix PR kern/23815.
|
| 1.219 | 22-Feb-2004 |
mycroft | Remove PQUIRK_CDROM. It is definitely not correct -- witness that no other
OS needs such a hack, and the same drive works fine on Suns -- and is much
more likely to be a bug in the host adapter driver (which is corroborated by
the PQUIRK_NOLUNS).
|
| 1.218 | 10-Oct-2003 |
matt | Remove a quirk I added for a Hitachi drive.
|
| 1.217 | 18-Sep-2003 |
mycroft | Don't print junk if an INQUIRY fails (usually with an ILLEGAL REQUEST due to
an unused LUN). Also, if the qualifier says the LUN is non-existant, don't
try to attach a device here.
|
| 1.216 | 12-Sep-2003 |
mycroft | Some devices really want INQUIRY to be the first command they receive. Also,
the result of the extra TEST UNIT READY was being ignored anyway. So, I wrote
it, I nuke it.
|
| 1.215 | 10-Sep-2003 |
mycroft | Set up the blank fields in the INQUIRY buffer *before* executing the command --
because we can't really rely on the "additional length" being correct (any
more?). Fixes some problems with devices showing up as "<, , >".
|
| 1.214 | 09-Sep-2003 |
mycroft | Remove NOMODESENSE quirks for Iomega drives -- they're totally unneeded.
|
| 1.213 | 09-Sep-2003 |
mycroft | Delete the NOMODESENSE quirks for Maxtor USB hard drives, as these are no
doubt resolved (and were never actually needed in the first place -- someone
was let out without supervision).
|
| 1.212 | 09-Sep-2003 |
mycroft | Exorcise PQUIRK_NODOORLOCK.
|
| 1.211 | 08-Sep-2003 |
mycroft | There is absolutely no sense in having a PQUIRK_NODOORLOCK entry for a
non-removable device. Nuke it.
|
| 1.210 | 08-Sep-2003 |
mycroft | If maxlun>0, automatically set PQUIRK_FORCELUNS, rather than using quirk
table entries.
|
| 1.209 | 08-Sep-2003 |
mycroft | Do a START UNIT only if the TEST UNIT READY reports that the device is not
ready. This avoids gratuitously starting the motor on floppy and CD-ROM
drives, and eliminates the need for the audio playing test in cdopen().
Therefore, also remove PQUIRK_NOSTARTUNIT.
|
| 1.208 | 07-Aug-2003 |
jrf | Added three scsi IDs which were tested by Dave Barnes who sent in
three PRs regarding them: 17836, 17837, 17838. Did a few kernel
compiles with them just to make sure they are okay. Approved by
christos@, thanks to Dave for sending the PRs and verifying to me
that they work.
|
| 1.207 | 29-Jun-2003 |
fvdl | branches: 1.207.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.206 | 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
|
| 1.205 | 13-May-2003 |
thorpej | Use aprint*().
|
| 1.204 | 02-May-2003 |
fvdl | Local change I had lying around: add NOLUNS quirk for an older seagate
drive.
|
| 1.203 | 21-Apr-2003 |
fvdl | Add another IBM drive that incorrectly reports DT only.
|
| 1.202 | 19-Apr-2003 |
fvdl | Add a PCAP_NODT quirk. I have two older Quantum Atlas drives that work
fine at U2W, but barf at U160, with different controllers. Unfortunately
sometimes just being initialized at U160 does the trick, so setting them
to lower speed in the BIOS may also be needed.
|
| 1.201 | 14-Mar-2003 |
bouyer | Move the SCSI_DELAY before the serialization for probe. multiple SCSI busses
can sleep, only probes have to be keept ordered.
|
| 1.200 | 10-Feb-2003 |
pk | Remove NOTAG quirk for the Seagate ST11200N; the cause of the failure that
prompted this entry was actually a driver bug.
|
| 1.199 | 01-Feb-2003 |
bouyer | Add PQUIRK_NOLUNS entry for "NEC ", "CD-ROM DRIVE:502".
Fixes PR kern/18479
|
| 1.198 | 27-Jan-2003 |
bouyer | Remove revision in quirk entry for TOSHIBA XM-4101TASUNSLCD. The same drive with a newer revision also fails, so assume all revisions are bad until proven otherwise. Should fix PR kern/19912.
|
| 1.197 | 09-Jan-2003 |
pk | Add two disks capable of doing synchronous transfer mode, but don't
report SCSI2: "MICROP 1924" and "FUJITSU M2266"
|
| 1.196 | 01-Jan-2003 |
thorpej | Use aprint_normal() in cfprint routines.
|
| 1.195 | 05-Dec-2002 |
jdolecek | Add NOLUNS|NOSYNC quirk for NEC CD-ROM DRIVE:222.
Change sent in PR kern/19286 by Dave Barnes.
|
| 1.194 | 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
|
| 1.193 | 04-Oct-2002 |
soren | Slightly simplicy the SCSI device attachment message:
- Use the plain words 'disk' and 'tape' instead of 'direct' and 'sequential'.
- Media status will be printed in the frontend, so don't bother with it.
- Don't bother printing the SCSI version, which is fixed these days, or the
numberic device type.
|
| 1.192 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.191 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.190 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.189 | 27-Sep-2002 |
thorpej | Introduce a new routine, config_match(), which invokes the
cfattach->ca_match function in behalf of the caller. Use it
rather than invoking cfattach->ca_match directly.
|
| 1.188 | 19-Sep-2002 |
jmc | Force the initial probes to happen within the newly forked off kthread.
This eliminates problems where the underlying interrupt handler isn't the
specific layer calling scsipi_complete() for a given scsi transaction.
This avoids deadlocks where the kthread that called the autoconf routines
to configure a scsibus shouldn't be the one put to sleep waiting on a
scsipi_complete (only the scsibus's kthread should be doing that).
To avoid jitter this will force the scsibus's to probe in the order they
run through autoconf (so machines with multiple bus's don't move sd* devices
around on every reboot).
|
| 1.187 | 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.
|
| 1.186 | 02-Aug-2002 |
bouyer | Add yet another broken sony cdrom. This one, in addition to broken LUN
handling, has broken sync negotiation. Also add PQUIRK_NOWIDE to be safe.
Fix PR kern/17773.
|
| 1.185 | 23-Jul-2002 |
matt | Add a quirk for a HITACHI driver (CAP_SYNC|CAP_WIDE16).
|
| 1.184 | 15-May-2002 |
bouyer | branches: 1.184.2; 1.184.4;
Change PQUIRK_CAP_SYNC to set PERIPH_CAP_SYNC aslo for devices that
improperly claim to only support DT clocking.
Add a PQUIRK_CAP_WIDE16 quirk.
Add a quirk for IBM DXHS36D SCSI disk (claims to only support DT).
|
| 1.183 | 05-May-2002 |
thorpej | A statement must follow a label.
|
| 1.182 | 23-Apr-2002 |
bouyer | The IBM drive I have here supports DT, but claim version 3 only.
So check for version 3, not 4 when looking for DT support.
This should be safe as these bits are reserved for older devices, they
should be set to 0 when not supported.
|
| 1.181 | 23-Apr-2002 |
bouyer | Report SPC-2 features (DT clocking, quick arbitration and selection,
information unit transfers) though periph_cap.
|
| 1.180 | 15-Apr-2002 |
joda | Maxtor D040H2 also can't modesense
|
| 1.179 | 01-Apr-2002 |
bouyer | Add a chan_name to struct scsipi_channel, holding the channel's name.
Set this to dv_xname for scsibus and atapibus.
Set the name of the kernel thread to chan_name instead of controller's
name:channel number (so that we can use this name for controller-specific
threads).
|
| 1.178 | 16-Mar-2002 |
chs | add a quirk for some disks I have that don't like tagged queuing.
|
| 1.177 | 23-Jan-2002 |
uwe | Add notag quirk for Seagate "ST11200N SUN1.05".
|
| 1.176 | 29-Dec-2001 |
augustss | Move some USB device quirks to the umass_quirk table.
|
| 1.175 | 29-Dec-2001 |
augustss | Remove some quirk entries that are handled in umass_quirk now.
|
| 1.174 | 27-Dec-2001 |
nathanw | Use PQUIRK_ONLYBIG for all Olympus cameras, not just the C-1.
Use PQUIRK_ONLYBIG for the PQI TravelFlash.
|
| 1.173 | 14-Dec-2001 |
nathanw | The Olympus C-1 engine cameras seem to need the PQIRK_ONLYBIG quirk.
|
| 1.172 | 02-Dec-2001 |
bouyer | Clean up attach of wd/atapibus:
kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead
of ata_atapi_attach as attach arch. Create a ata_device, compatible with
scsipi_channel, to attach wd.
|
| 1.171 | 26-Nov-2001 |
fredette | Added quirk entries for Adaptec and Emulex SCSI interposer boards.
Decode a limited set of SASI/SCSI-1 sense codes, and under sun2
only, conjure up INQUIRY responses for these boards.
|
| 1.170 | 19-Nov-2001 |
tsutsui | Add a new quirk value PQUIRK_CAP_SYNC for SCSI1 devices which support
sync transfers, and set PERIPH_CAP_SYNC for them in scsi_probe_device().
As per discussion on tech-kern.
|
| 1.169 | 19-Nov-2001 |
soren | Floptical drive needs PQUIRK_NODOORLOCK as well.
|
| 1.168 | 19-Nov-2001 |
soren | Insite Floptical drive needs PQUIRK_NOLUNS.
|
| 1.167 | 18-Nov-2001 |
tsutsui | Fix formatting of scsi_quirk_patterns[] and wrap long lines.
|
| 1.166 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.165 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.164 | 06-Nov-2001 |
augustss | Add a quirk for a USB CD/RW.
|
| 1.163 | 21-Oct-2001 |
mjl | branches: 1.163.2;
Add ST15150N to quirk table, as per PR/13817.
|
| 1.162 | 16-Oct-2001 |
christos | PR/14256: Kevin P. Neal: Add quirk for micropolis drive.
|
| 1.161 | 31-Aug-2001 |
augustss | branches: 1.161.2;
Add a new quirk, PQUIRK_NOBIGMODESENSE, for devices that need big
SCSI ops, but still fails on big mode sense.
Makes M-Sys DiskOnKey work.
|
| 1.160 | 31-Aug-2001 |
augustss | Some quirks for USB M-Sys DiskOnKey. Still not enough to make it work.
|
| 1.159 | 18-Jul-2001 |
bouyer | Adn scsipi_target_detach() and scsipi_thread_call_callback() as discussed
on tech-kern. scsipi_target_detach() accept wildcard target/lun as requested.
|
| 1.158 | 11-Jun-2001 |
pk | branches: 1.158.2;
Add nolun/nosync quirk for "IBM CDRM00201 !F"
See PR#13031.
|
| 1.157 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.156 | 26-Feb-2001 |
fvdl | branches: 1.156.2;
Add quirks for VMware emulated disks, for NetBSD as a guest OS.
|
| 1.155 | 16-Feb-2001 |
pk | Turn off command tagging by default for pre-SCSI2 devices.
|
| 1.154 | 18-Jan-2001 |
jdolecek | constify
|
| 1.153 | 03-Dec-2000 |
ad | I2O HBAs provide an abstracted view of the bus; use SCBUSIOLLSCAN to give
them an oppertunity to re-scan it before we do.
|
| 1.152 | 14-Nov-2000 |
pk | Mark the "TOSHIBA, XM-4101TASUNSLCD" CD-R for a-sync operation only.
|
| 1.151 | 22-Sep-2000 |
ad | Add a new, optional method to scsipi_adapter (scsipi_accesschk()), and use
it when considering whether to attach devices. This is to facilitate
`non-SCSI' RAID controller drivers that want to provide SCSI pass-through
services to the kernel.
|
| 1.150 | 13-Aug-2000 |
mjacob | Add "SYMBIOS" processor type to the SDEV_NOLUN category. These are usually
the GEM chips on a Sun D1000- and they cause Qlogic SBus firmware to blow
chunks if you access past lun 0.
|
| 1.149 | 08-Aug-2000 |
mjacob | During probe, use SCBUSACCEL ioctl to adapters that support it to
enable fast/wide/tagged.
|
| 1.148 | 03-Aug-2000 |
bouyer | Add quirk entry for NEDICOM CRD-BP2, from kern/10738.
|
| 1.147 | 09-Jun-2000 |
enami | branches: 1.147.2;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.146 | 30-May-2000 |
augustss | Add NOMODESENSE quirk for TEAC USB floppy.
|
| 1.145 | 30-May-2000 |
augustss | Add NOMODESENSE quirk for Y-E data floppy (thanks Jason for reminding me).
|
| 1.144 | 29-May-2000 |
bouyer | ADEV_CDROM -> SDEV_CDROM, for consistency.
|
| 1.143 | 28-May-2000 |
gmcgarry | Handle ADEV_CDROM in quirks table.
Quirk entry for another Toshiba cdrom.
|
| 1.142 | 15-May-2000 |
dante | branches: 1.142.2;
Add few peripheral device type
Remove "???" from T_IT8_1/2. They actually are pre-press devices for graphic arts as described by ASC IT8
Zeros and blanks scsipi_inquiry_data from byte 58 to byte 74 if additional_length is less than 58
|
| 1.141 | 14-May-2000 |
dante | Change scsipi_inquiry_data strucure to be ANSI SPC-2 rev16 compliant
|
| 1.140 | 19-Apr-2000 |
enami | Allocate the variable `inqbuf' in scsi_probedev on stack rather than
statically. Since this function may called for another luns immediately,
allocating it statically doesn't make sense and may cause race condition
as pointed out by PR#9749.
|
| 1.139 | 02-Apr-2000 |
augustss | Change the initial field in struct ata_atapi_attach and struct scsipi_link
slightly to allow scsibus and atapibus to attach to the same device.
Furthermore, only attach a scsibus when the bus type is BUS_SCSI.
|
| 1.138 | 27-Mar-2000 |
augustss | In attach message, spell it "target" not "targ", as in the locator.
|
| 1.137 | 19-Mar-2000 |
sjg | Added NOLUNS quirk for YAMAHA CRW8424S
|
| 1.136 | 17-Mar-2000 |
soren | scsiprint() is needed even without scsibus'es, so move it to scsi_base.c .
|
| 1.135 | 13-Mar-2000 |
martin | Added Artec/Ultima A6000C scanner to the quirks table (no LUNs).
|
| 1.134 | 20-Jan-2000 |
mjacob | Nobody said no to adding a pointer to original scsi inquiry data
to the scsibus attach args. Make sure it's nulled for ATAPI. Also,
for scsiconf.c, modify SENA's quirk entry.
|
| 1.133 | 13-Jan-2000 |
nisimura | Add tweaks for TEAC compact cassette tape drive.
|
| 1.132 | 14-Nov-1999 |
soren | Allow SCSI_DELAY to be shorter than 2 seconds.
|
| 1.131 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.130 | 11-Oct-1999 |
hwr | branches: 1.130.2; 1.130.4;
Add ad quirk for CDU-561 CD-ROM. From PR kern/8608
by SUNAGAWA Keiki <kei_sun@ba2.so-net.ne.jp>.
|
| 1.129 | 10-Oct-1999 |
hwr | Add a quirk to recognize a Toshiba XM-3401TA SCSI CD-ROM as
removable SCSI device. From Matthew Fredette <fredette@mit.edu> in
kern/7438.
|
| 1.128 | 30-Sep-1999 |
thorpej | branches: 1.128.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.127 | 19-Sep-1999 |
nathanw | Add NOLUNS quirk for another Texel CD-ROM revision.
|
| 1.126 | 11-Sep-1999 |
thorpej | Implement detaching of SCSI busses.
|
| 1.125 | 09-Sep-1999 |
hwr | Add a quirk for Wangtec SCSI-36 (QIC-120) tape drive.
From Izumi Tsutsiu in PR 8357.
|
| 1.124 | 26-Jul-1999 |
explorer | make the JVC 2626 match more than one version, since all seem to be
returning errors on the LUN probe.
|
| 1.123 | 14-Jul-1999 |
tron | Skip LUN check for all versions of the UMAX Astra 1220S as suggested
by Dan McMahill in PR kern/7991.
|
| 1.122 | 10-Dec-1998 |
mjacob | branches: 1.122.2; 1.122.4;
add Adaptec RAID units as devices that do not return geometry pages
|
| 1.121 | 08-Dec-1998 |
thorpej | When allocating a device's scsipi_link, initialize the pending_xfers
queue.
|
| 1.120 | 05-Dec-1998 |
mjacob | NOMODESENSE for all Seagate ST19171- not just FC
|
| 1.119 | 05-Dec-1998 |
mjacob | Eliminate the moreluns entry as it makes no sense for fat SCSI busses (e.g,
FC loops). Change the semantics of scsi_probedev so that it returns 1 if
you should continue probing at this target, else 0 for not.
Replace the blanket use of '7' with the use of the new sc_maxlun property
that is now gathered from HBAs. Allocate scsipi_link arrays based upon this.
Fix a really nasty and silly bug that has been there for a while where the
number of first level scsipi_link structures was one less than it needed
to be.
|
| 1.118 | 26-Nov-1998 |
leo | The IOMEGA ZIP 100, J.03 does not grok LUN's.
|
| 1.117 | 19-Nov-1998 |
thorpej | Add a reference to the adapter when the scsibus is opened, and delete it
when it is closed.
|
| 1.116 | 19-Nov-1998 |
thorpej | Add a reference to the adapter before probing the bus, and delete it
once we are done probing.
|
| 1.115 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.114 | 20-Oct-1998 |
thorpej | Fix a sight open flags buglet pointed out by Matthias Scheler.
|
| 1.113 | 10-Oct-1998 |
thorpej | Enforce open-for-writing on ioctls that change the bus's state.
Implement ioctl pass-through to the host bus adapter, allowing both
SCBUS* ioctls handled at that level and host adapter-specific ioctls
to be implemented. Implement SCBUSIORESET as a pass-through.
Inspired by PR #6090, from Matt Jacob.
|
| 1.112 | 10-Oct-1998 |
thorpej | Implement the SCBUSIOSCAN ioctl. Rescans the bus for new devices.
|
| 1.111 | 10-Oct-1998 |
thorpej | Add the open/close/ioctl entry points for the SCSI bus, i.e. /dev/scsibusN.
|
| 1.110 | 08-Oct-1998 |
thorpej | Add a NOSYNCCACHE quirk for the Micropolis 2217-15MQ1091501, as reported
by Matt Jacob, PR #6027.
|
| 1.109 | 08-Oct-1998 |
thorpej | Add a NOLUNS quirk for the Sony CDL1100 changer, from Chris Jones, PR #6238.
|
| 1.108 | 08-Sep-1998 |
mjacob | Reflect changes in quirk flags, and also now add case post retrieval
of Inquiry data during probe where SDEV_NOSYNC, SDEV_NOTAG and SDEV_NOWIDE
can be set (with quirk data overriding) per device probed.
|
| 1.107 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.106 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.105 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.104 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.103 | 31-Jul-1998 |
thorpej | Use the pool allocator for scsipi_xfer structures.
|
| 1.102 | 30-Jul-1998 |
fvdl | branches: 1.102.2;
Add NOLUNS quirk for the UMAX Astra 1220 scanner.
|
| 1.101 | 22-Jun-1998 |
bouyer | Ricoh IS60 to the quirk table (SDEV_NOLUN). Fixes PR 5473 by
Feico Dillema < dillema@acm.org>
|
| 1.100 | 04-May-1998 |
thorpej | The DEC RRD42 doesn't deal with LUNs very well either.
kern/5376, Simon Burge <simonb@telstra.com.au>
|
| 1.99 | 04-May-1998 |
thorpej | Add the UMAX Astra 1200S scanner to the list of the LUN-clueless.
kern/5390, Rene Hexel <rh@vip.at>
|
| 1.98 | 23-Apr-1998 |
explorer | Ignore LUN on yet another cdrom
|
| 1.97 | 10-Apr-1998 |
mjacob | trim ID of ST19171FC so that SUN drives are caught
|
| 1.96 | 07-Mar-1998 |
scottr | Quirk table entries for IBM H3171 and Quantum ELS85S disks.
|
| 1.95 | 21-Jan-1998 |
mikel | add NOLUNS quirk for UMAX S-12; from Andreas Wrede in PR kern/4849
|
| 1.94 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.93 | 03-Oct-1997 |
thorpej | branches: 1.93.2;
Clean up and comment the device types, and add the "enclosure services",
"storage array", and "IT8" types.
|
| 1.92 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.91 | 29-Sep-1997 |
bouyer | Add quirks entries for Hitachi DK515C disk drives and Cipher ST150S tape drive.
Closes PR kern/4171 from jbernard@tater.mines.edu.
|
| 1.90 | 19-Sep-1997 |
mjacob | Another couple NOMODESENSE drives.
|
| 1.89 | 09-Sep-1997 |
pk | Add SDEV_NOLUNS to another CD-ROM player.
|
| 1.88 | 27-Aug-1997 |
bouyer | branches: 1.88.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.87 | 17-Aug-1997 |
mjacob | A few more SDEV_NOLUNS entries- one tape drive (SONY Beta/VHS) and a couple
of PCMCIA card readers (one as T_DIRECT, the other as T_PROCESSOR).
|
| 1.86 | 16-Aug-1997 |
mjacob | This fibre channel disk doesn't support the geometry page.
|
| 1.85 | 14-Aug-1997 |
explorer | add SDEV_NOLUNS for my scanner (UMAX S-6E)
|
| 1.84 | 17-Jul-1997 |
perry | branches: 1.84.2;
More CD quirks, from Dave Huang (pr-3843)
|
| 1.83 | 24-Apr-1997 |
mycroft | Another silly CD-ROM drive...
|
| 1.82 | 20-Apr-1997 |
thorpej | And Yet Another Broken IBM disk (the 664 this time), once again
reported by Hubert Feyrer <feyrer@grizu.fh-regensburg.de>.
|
| 1.81 | 20-Apr-1997 |
thorpej | Add Yet Another Broken IBM disk to the SDEV_AUTOSAVE list. This one
doesn't report a vendor string (yuck)!. Problem reported by
Hubert Feyrer <feyrer@grizu.fh-regensburg.de>.
|
| 1.80 | 19-Apr-1997 |
pk | Two more quirks:
Python 28454-XXX tape drive: NOLUNS
1588-15MBSUN0669 disk: AUTOSAVE
|
| 1.79 | 08-Apr-1997 |
scottr | Add another IBM disk that needs SDEV_AUTOSAVE.
|
| 1.78 | 02-Apr-1997 |
mycroft | Add a SDEV_AUTOSAVE quirk for the Emulex SCSI<->ESDI bridge.
|
| 1.77 | 01-Apr-1997 |
mikel | loosen the version number check in the quirk list for the CDR-H98MV;
from Koji Imada in PR kern/3419.
|
| 1.76 | 27-Mar-1997 |
scottr | Add SDEV_AUTOSAVE quirks for IBM 0663H{08,12} disks
|
| 1.75 | 25-Mar-1997 |
scottr | Add SDEV_NOLUNS quirk for Seagate ST125N.
|
| 1.74 | 11-Mar-1997 |
mikel | SDEV_NOMODESENSE quirks are no longer needed for optical devices; from
Enami Tsugutomo in PR kern/3308.
|
| 1.73 | 14-Dec-1996 |
mycroft | branches: 1.73.6;
Make sure to initialize the memory we just allocated.
|
| 1.72 | 10-Dec-1996 |
thorpej | Add a "max_target" member to struct scsi_link, which is filled in by
host adapter drivers, indicating the highest SCSI target they can
address. Use this value to dynamically allocate data structures, rather
than hard-coding 8 targets.
These changes allow targets > 7 to be addressed on wide SCSI busses.
Fixes PRs #1674 and #2892.
|
| 1.71 | 05-Dec-1996 |
cgd | update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.70 | 29-Nov-1996 |
thorpej | Add Yet Another Broken CD-ROM Drive to The List.
From Lennart Augustsson <augustss@cs.chalmers.se>, PR #2971.
|
| 1.69 | 08-Nov-1996 |
explorer | Another quirky cdrom drive ; pr kern/2917
|
| 1.68 | 06-Nov-1996 |
mikel | Add another broken CDROM. Fixes PR kern/2913.
Some misc. cleanup.
|
| 1.67 | 23-Oct-1996 |
matthias | * At least the Teac FC-1 Shugart-SCSI bridgeboard does motor on/off
management by itself. But when it gets a start unit request, it keeps
the floppy motor running all the time. This adds code for dealing with
yet another quirk (SDEV_NOSTARTUNIT) that prevents sd.c from sending
start unit requests. A entry for the Teac FC-1 is added to the quirk
table.
|
| 1.66 | 18-Oct-1996 |
explorer | Add quirk entries for two more optical drives; closes pr kern/2861
|
| 1.65 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.64 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.63 | 03-Sep-1996 |
thorpej | Store the SCSI version in the scsi_link, and set the LUN in the CDB
if the version is <= SCSI-2. This should help some older SCSI
devices that previously needed the "NOLUNS" quirk. While this is
not strictly necessary on SCSI-2 devices, the spec allows it,
so we set it for SCSI-2 devices "just in case". See section 7.2.2 of
Draft X3T9.2 Rev 10L for details.
|
| 1.62 | 28-Aug-1996 |
cgd | (1) add a 'channel' field to scsi_link.
(2) in scsibusmatch, match channel as appropriate.
(3) add a scsiprint() function, to do the "scsibus at..."
and channel (if not SCSI_CHANNEL_ONLY_ONE) printing,
i.e. the common functionality that all SCSI drivers currently
should be doing.
|
| 1.61 | 27-Aug-1996 |
cgd | change cfprint_t type definition to take a const char *, rather than
a char *, because that's what was really intended, and because
if the print function modifies the string, various things could become
unhappy (so the string should _not_ be modified).
|
| 1.60 | 24-Jul-1996 |
thorpej | Match any Maxtor XT-8760S for the SDEV_NOLUNS quirk. Some of these
drives carry Sun OEM identifiers, as well.
Noticed by Taras Ivanenko <ivanenko@ctpa03.mit.edu>.
|
| 1.59 | 05-Jul-1996 |
explorer | add ZIP quirk for SDEV_NOMODESENSE
|
| 1.58 | 05-Jul-1996 |
christos | - Apply PR/2535: Add support for flex scsi disks.
- Add a quirk called SCSI_NOMODESENSE for drives like the iomega jaz,
that don't support mode sense.
|
| 1.57 | 02-May-1996 |
neil | branches: 1.57.4;
Added another quirked SCSI device, re PR 2320 from Manuel Bouyer
|
| 1.56 | 22-Apr-1996 |
christos | remove include of <sys/cpu.h>
|
| 1.55 | 21-Mar-1996 |
scottr | Add a SDEV_NOLUNS quirk for the Epson OMD-5010 removable MO cartridge drive.
|
| 1.54 | 18-Mar-1996 |
hpeyerl | Fix for NEC D3847 1.6gB drive which reported itself at lun[1-7].
|
| 1.53 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.52 | 05-Mar-1996 |
thorpej | Add another broken drive that doesn't deal with LUNs properly to The List.
Fixes PR 1961, from Kevin P. Neal.
|
| 1.51 | 05-Mar-1996 |
thorpej | Normalize SCSI autoconfiguration output.
From Chris Demetriou <cgd@NetBSD.ORG>. Fixes PR #1958.
|
| 1.50 | 22-Feb-1996 |
mycroft | Add a SDEV_AUTOSAVE quirk for the DEC RZ55.
|
| 1.49 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.48 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.47 | 12-Jan-1996 |
pk | LUN deficient device: Tandberg 3600 (Ken Raeburn; PR#1933).
|
| 1.46 | 31-Dec-1995 |
thorpej | Add another broken CD-ROM drive to The List (Chinon CDS-525), per PR #1686.
From Kortelainen Mika <k125374@cs.tut.fi>.
|
| 1.45 | 11-Dec-1995 |
mycroft | Trim NULs, in addition to spaces, in scsi_strvis().
|
| 1.44 | 01-Nov-1995 |
pk | More rogues.
|
| 1.43 | 13-Oct-1995 |
gwr | Add quirk entries for Wangtek SCSI tapes. Also, add entry for
Tandberg 3800 (fixes PR#1592 - from Jochen Pohl).
|
| 1.42 | 26-Sep-1995 |
thorpej | Add another broken NEC CD-ROM drive to The List. Fixes PR #1451.
|
| 1.41 | 21-Aug-1995 |
pk | Set the scsi link before calling config_attach(), so target/lun info is
available to, say, dk_establish().
|
| 1.40 | 14-Aug-1995 |
briggs | CHINON CDS-535 version Q117 also needs NOLUNS. Assume all 535s do.
|
| 1.39 | 13-Aug-1995 |
briggs | Another SDEV_NOLUNS device from paul@pgoyette.bdt.com (Paul Goyette).
|
| 1.38 | 06-Aug-1995 |
mycroft | Add another Seagate disk to the list, and make all the disks match any
firmware revision.
|
| 1.37 | 12-Jul-1995 |
cgd | add/change a few quirks:
(1) all Chinon CDS-431 CD-ROMs (regardless of revision)
are forced to only having LUN 0, at the suggestion
of Michael Hitch.
(2) _force_ searching of extra LUNs for the Emulex MD21/S2
ESDI bridge. It's pre-SCSI 1, but knows about LUNs.
"amazing." From Jason Thorpe.
(3) recognize an Emulex tape adapter in front of a QIC-36
tape, and have it forced to only LUN 0. This is
an odd one; vendor, name, and rev strings are all
spaces. Anything that mathes this is very likely
broken, anyway, so might as well give it a shot.
Again from Jason Thorpe.
|
| 1.36 | 12-Jul-1995 |
cgd | implement SDEV_FORCELUNS flag, as suggested by Jason Thorpe.
|
| 1.35 | 09-Jul-1995 |
cgd | Morningstar SnapLink -> SDEV_NOLUNS, per Peter Galbavy in pr 1192.
|
| 1.34 | 09-Jul-1995 |
cgd | play the alphabetization game
|
| 1.33 | 09-Jul-1995 |
cgd | QUANTUM P105S 910-10-94x A.3 -> SDEV_NOLUNS
|
| 1.32 | 27-Jun-1995 |
cgd | another CD-ROM quirk, a la PR 1166
|
| 1.31 | 24-Jun-1995 |
cgd | add entry for another tweaked Texel CD-ROM. from pr 1151
|
| 1.30 | 18-Jun-1995 |
mycroft | Add another CD-ROM drive to The List.
|
| 1.29 | 02-Jun-1995 |
pk | Add `Maxtor LXT-213S', this time without a Sun OEM addition.
|
| 1.28 | 04-Apr-1995 |
mycroft | Add another broken CD-ROM drive to The List.
|
| 1.27 | 22-Feb-1995 |
mycroft | Be less picky about revision numbers for some broken devices.
|
| 1.26 | 09-Feb-1995 |
pk | Another Maxtor botch.
|
| 1.25 | 01-Feb-1995 |
mycroft | Add Exabyte 8200 to the list.
|
| 1.24 | 30-Jan-1995 |
mycroft | All revisions of the TDC 3600 are rogues.
|
| 1.23 | 16-Jan-1995 |
mycroft | Don't clear the debug flags when copying the prototype scsi_link.
|
| 1.22 | 12-Jan-1995 |
mycroft | Add two more broken devices.
|
| 1.21 | 01-Jan-1995 |
mycroft | Put Dan's broken tape drive in The List.
|
| 1.20 | 30-Dec-1994 |
mycroft | And a little better...
|
| 1.19 | 30-Dec-1994 |
mycroft | Deal with short inquiry response a little better.
|
| 1.18 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.17 | 03-Nov-1994 |
mycroft | Oops; update scsibusmatch() arg list.
|
| 1.16 | 03-Nov-1994 |
mycroft | Always use direct configuration for SCSI devices.
|
| 1.15 | 03-Nov-1994 |
mycroft | Remove a couple of i386-specific hacks, and make a couple of others #ifdef
CONFIG_INDIRECT instead.
|
| 1.14 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.13 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.12 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.11 | 25-Mar-1994 |
mycroft | Put controller target in scsi_switch.
|
| 1.10 | 17-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.9 | 27-May-1993 |
deraadt | branches: 1.9.3;
scsi probing spent too much time on the bus -- sped it up now,
and simplified the message printing code (works the same)
|
| 1.8 | 25-May-1993 |
deraadt | patch00149 by Julian Elischer <julian@jules.dialix.oz.au> & Rodney Grimes.
When an error was encountered, the sd/cd drivers printed blockno&ff0000
rather that blockno.
|
| 1.7 | 20-May-1993 |
cgd | add rcsids and clean up file headers
|
| 1.6 | 05-May-1993 |
deraadt | fixed bug from last commit
|
| 1.5 | 04-May-1993 |
deraadt | support for making dev->id_alive be set, this is for iostat to
find disk devices. wee bit of a kludge. sub-device attach()
routines must now return 1 for successful attach(), 0 otherwise.
Other bsd's do this too..
|
| 1.4 | 20-Apr-1993 |
deraadt | a test for mycroft..
|
| 1.3 | 19-Apr-1993 |
mycroft | Don't overwrite global kernel version string.
|
| 1.2 | 12-Apr-1993 |
deraadt | new scsi subsystem.
changes also in config/mkioconf.c
i386/isa/wd.c, fd.c, and all scsi drivers.
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.9.3.7 | 22-Jan-1994 |
briggs | lu -> lun in SCSI_DEBUG section...
|
| 1.9.3.6 | 25-Nov-1993 |
mycroft | Compare the device name rather than the attach function.
|
| 1.9.3.5 | 24-Nov-1993 |
mycroft | Final tweaks to get it going.
|
| 1.9.3.4 | 24-Nov-1993 |
mycroft | Some cleanup, and fix a bug I just introduced.
|
| 1.9.3.3 | 24-Nov-1993 |
mycroft | More construction...
|
| 1.9.3.2 | 24-Nov-1993 |
mycroft | Still under construction...
|
| 1.9.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.57.4.1 | 31-Jul-1996 |
jtc | Pulled up diff between revs 1.59 and 1.60 by request from Jason Thorpe
|
| 1.73.6.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.84.2.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.88.2.5 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.88.2.4 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.88.2.3 | 16-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.88.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.88.2.1 | 27-Aug-1997 |
thorpej | file scsiconf.c was added on branch marc-pcmcia on 1997-08-27 23:33:19 +0000
|
| 1.93.2.1 | 05-May-1998 |
mycroft | Pull up 1.96, per request of scottr. Also do 1.95 and 1.97-1.100.
|
| 1.102.2.2 | 08-Aug-1998 |
eeh | Revert cdevsw mmap routines to return int.
|
| 1.102.2.1 | 30-Jul-1998 |
eeh | file scsiconf.c was added on branch eeh-paddr_t on 1998-08-08 03:06:51 +0000
|
| 1.122.4.1 | 02-Aug-1999 |
thorpej | Update from trunk.
|
| 1.122.2.3 | 16-Jan-2000 |
he | Apply patch (requested by ad):
When probing a SCSI target, do not report an error if the target
indicates that LUNs are not supported.
|
| 1.122.2.2 | 27-Nov-1999 |
he | Pull up revision 1.123 (requested by dmcmahill):
Add SCSI quirk for UMAX Astra 1220S scanner: skip LUN check,
fixes PR#7991.
|
| 1.122.2.1 | 10-Nov-1999 |
he | Pull up revisions 1.129-1.130 (requested by hwr):
Add quirks for Toshiba XM-3401TA SCSI CD-ROM and CDU-561 CD-ROM,
fixing PR#7438 and PR#8608.
|
| 1.128.2.2 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.128.2.1 | 21-Dec-1999 |
wrstuden | Initial commit of recent changes to make DEV_BSIZE go away.
Runs on i386, needs work on other arch's. Main kernel routines should be
fine, but a number of the stand programs need help.
cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512
byte block devices. vnd, raidframe, and lfs need work.
Non 2**n block support is automatic for LKM's and conditional for kernels
on "options NON_PO2_BLOCKS".
|
| 1.130.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.130.2.13 | 23-Apr-2001 |
ad | If SCSIPI_CHAN_NOSETTLE is set in the channel's flags, then don't bother
with SCSI_DELAY.
|
| 1.130.2.12 | 11-Apr-2001 |
mjacob | oops: reversed conditional
|
| 1.130.2.11 | 11-Apr-2001 |
mjacob | Make scsipi_channel_init a function returning an int- non-zero means
it failed to initialize the channel (this should be acceptable)- in
which case we complain and don't schedule bus probing for later.
We make the internal memory allocations for the periph and the chan_periphs
array M_NOWAIT- this way we have a hope of booting instead of silently hanging
during boot if we've run out of memory.
|
| 1.130.2.10 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.130.2.9 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.130.2.8 | 08-Dec-2000 |
bouyer | Sync with HEAD.
|
| 1.130.2.7 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.130.2.6 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.130.2.5 | 04-Feb-2000 |
thorpej | Make sure the channel's completion thread exits when the channel
is detached, and completely encapsulate the way periph's are stored
by the channel.
|
| 1.130.2.4 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.130.2.3 | 26-Oct-1999 |
thorpej | Deal a little more gracefully with the fact that xfer mode parameters
are for the I_T Nexus, and make all xfer mode updates `async events'.
|
| 1.130.2.2 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.130.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.142.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.147.2.6 | 11-Feb-2003 |
msaitoh | Pull up revision 1.199 (requested by bouyer):
Add SDEV_NOLUNS quirk entry for "NEC ", "CD-ROM DRIVE:502".
Fixes kern/18479 .
|
| 1.147.2.5 | 04-Sep-2002 |
itojun | sys/dev/scsipi/scsipiconf.c 1.186 via patch (bouyer)
add quirk entry for yet another broken sony cdrom. Fix kern/17773
|
| 1.147.2.4 | 13-Nov-2001 |
he | Pull up revision 1.162 (via patch, requested by christos):
Add a ``no tagged queueing'' for an old Micropolis drive. Fixes
PR#14256.
|
| 1.147.2.3 | 25-Oct-2001 |
he | Pull up revision 1.153 (requested by ad):
Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
|
| 1.147.2.2 | 28-Aug-2000 |
mjacob | Add support for accelleration SCSI ioctls (that allow for the
midlayer to enable sync/wide/tagged).
Approved for pullup to netbsd-1-5 by thorpej@netbsd.org.
|
| 1.147.2.1 | 03-Aug-2000 |
bouyer | pullup 1.147 -> 1.148 (approved by thorpej):
Add quirk entry for NEDICOM CRD-BP2, from kern/10738.
|
| 1.156.2.21 | 15-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.156.2.20 | 03-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.156.2.19 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.156.2.18 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.156.2.17 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.16 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.15 | 13-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.14 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.13 | 12-Jul-2002 |
nathanw | No longer need to pull in lwp.h; proc.h pulls it in for us.
|
| 1.156.2.12 | 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).
|
| 1.156.2.11 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.10 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.9 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.156.2.8 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.7 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.156.2.6 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.156.2.5 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.156.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.156.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.156.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.156.2.1 | 05-Mar-2001 |
nathanw | Initial commit of scheduler activations and lightweight process support.
|
| 1.158.2.7 | 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
|
| 1.158.2.6 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.158.2.5 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.158.2.4 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.158.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.158.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.158.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.161.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.161.2.1 | 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).
|
| 1.163.2.1 | 12-Nov-2001 |
thorpej | Sync the thorpej-mips-cache branch with -current.
|
| 1.184.4.4 | 15-Mar-2004 |
jmc | Pullup rev 1.220 (requested by bouyer in ticket #1625)
Add a drive with broken tagged queuing support. PR#23815
|
| 1.184.4.3 | 02-Jun-2003 |
tron | Pull up revision 1.198 (requested by bouyer in ticket #1155):
Remove revision in quirk entry for TOSHIBA XM-4101TASUNSLCD.
The same drive with a newer revision also fails, so assume all revisions
are bad until proven otherwise.
|
| 1.184.4.2 | 02-Jun-2003 |
tron | Pull up revision 1.199 (requested by bouyer):
Add PQUIRK_NOLUNS entry for "NEC ", "CD-ROM DRIVE:502".
Fixes PR kern/18479
|
| 1.184.4.1 | 02-Aug-2002 |
lukem | Pull up revision 1.186 (requested by bouyer in ticket #608):
Add yet another broken sony cdrom. This one, in addition to broken LUN
handling, has broken sync negotiation. Also add PQUIRK_NOWIDE to be safe.
Fix PR kern/17773.
|
| 1.184.2.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.184.2.1 | 16-May-2002 |
gehenna | Add the character device switch.
|
| 1.207.2.9 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.207.2.8 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.207.2.7 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.207.2.6 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.207.2.5 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.207.2.4 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.207.2.3 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.207.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.207.2.1 | 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.
|
| 1.220.4.1 | 20-Feb-2007 |
tron | Pull up following revision(s) (requested by bouyer in ticket #11051):
sys/dev/scsipi/scsiconf.c: revision 1.242 via patch
The SEAGATE ST318203LSUN18G announces DT only capabilitie although it also
supports ST. Because of this the HBA doesn't initiate sync/wide negotiation,
so add PQUIRK_CAP_SYNC | PQUIRK_CAP_WIDE16 for this drive; fix
problem reported by Havard Eidnes.
|
| 1.226.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.226.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.229.2.1 | 19-Feb-2007 |
tron | Pull up following revision(s) (requested by bouyer in ticket #1668):
sys/dev/scsipi/scsiconf.c: revision 1.242 via patch
The SEAGATE ST318203LSUN18G announces DT only capabilitie although it also
supports ST. Because of this the HBA doesn't initiate sync/wide negotiation,
so add PQUIRK_CAP_SYNC | PQUIRK_CAP_WIDE16 for this drive; fix
problem reported by Havard Eidnes.
|
| 1.230.2.5 | 21-Jan-2008 |
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.
|
| 1.232.6.1 | 29-Nov-2005 |
yamt | sync with head.
|
| 1.234.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.234.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.234.8.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.234.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.234.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.234.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.238.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.238.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.238.2.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.238.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.241.4.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.241.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.241.2.1 | 19-Feb-2007 |
riz | Pull up following revision(s) (requested by bouyer in ticket #452):
sys/dev/scsipi/scsiconf.c: revision 1.242
The SEAGATE ST318203LSUN18G announces DT only capabilitie although it also
supports ST. Because of this the HBA doesn't initiate sync/wide negotiation,
so add PQUIRK_CAP_SYNC | PQUIRK_CAP_WIDE16 for this drive; fix
problem reported by Havard Eidnes.
|
| 1.243.26.1 | 11-Dec-2007 |
yamt | sync with head.
|
| 1.243.24.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.243.22.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.243.16.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.243.14.2 | 08-Dec-2007 |
jmcneill | Rename pnp(9) -> pmf(9), as requested by many.
|
| 1.243.14.1 | 06-Nov-2007 |
joerg | Refactor PNP API:
- Make suspend/resume directly a device functionality. It consists of
three layers (class logic, device logic, bus logic), all of them being
optional. This replaces D0/D3 transitions.
- device_is_active returns true if the device was not disabled and was
not suspended (even partially), device_is_enabled returns true if the
device was enabled.
- Change pnp_global_transition into pnp_system_suspend and
pnp_system_resume. Before running any suspend/resume handlers, check
that all currently attached devices support power management and bail
out otherwise. The latter is not done for the shutdown/panic case.
- Make the former bus-specific generic network handlers a class handler.
- Make PNP message like volume up/down/toogle PNP events. Each device
can register what events they are interested in and whether the handler
should be global or not.
- Introduce device_active API for devices to mark themselve in use from
either the system or the device. Use this to implement the idle handling
for audio and input devices. This is intended to replace most ad-hoc
watchdogs as well.
- Fix somes situations in which audio resume would lose mixer settings.
- Make USB host controllers better deal with suspend in the light of
shared interrupts.
- Flush filesystem cache on suspend.
- Flush disk caches on suspend. Put ATA disks into standby on suspend as
well.
- Adopt drivers to use the new PNP API.
- Fix a critical bug in the generic cardbus layer that made D0->D3
break.
- Fix ral(4) to set if_stop.
- Convert cbb(4) to the new PNP API.
- Apply the PCI Express SCI fix on resume again.
|
| 1.244.2.1 | 08-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.245.6.5 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.245.6.4 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.245.6.3 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.245.6.2 | 06-Apr-2008 |
mjf | - after some discussion with agc@ i agreed it would be a good idea to move
device_unregister_* to device_deregister_* to be more like the pmf(9)
functions, especially since a lot of the time the function calls are next
to each other.
- add device_register_name() support for dk(4).
|
| 1.245.6.1 | 05-Apr-2008 |
mjf | - add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing
as these are always needed.
- convert many, many drivers over to the New Devfs World Order. For a
list of device drivers yet to be converted see,
http://www.netbsd.org/~mjf/devfs-todo.html.
- add a new device_unregister_all(device_t) function to remove all device
names associated with a device_t, which saves us having to construct
device names when the driver is detached.
- add a DEV_AUDIO type for devices.
|
| 1.246.4.5 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.246.4.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.246.4.3 | 16-May-2009 |
yamt | sync with head
|
| 1.246.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.246.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.246.2.2 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.246.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.247.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.247.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.248.2.2 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.248.2.1 | 03-Jul-2008 |
simonb | Sync with head.
|
| 1.249.2.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.250.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.250.2.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.255.4.3 | 31-May-2011 |
rmind | sync with head
|
| 1.255.4.2 | 21-Apr-2011 |
rmind | sync with head
|
| 1.255.4.1 | 30-May-2010 |
rmind | sync with head
|
| 1.255.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.258.2.1 | 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.262.10.4 | 03-Sep-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #523):
sys/dev/scsipi/scsiconf.c: revision 1.269
If the controller supports more than 256 commands per target,
clamp it to 256 (maximum number of tags in SCSI). Newer controllers
(such as mpii(4), and mfi(4) when fixed to announce tagged queuing support)
support more than 256 outstanding commands and don't use the scsi tag,
but at this time scsipi will always allocate a tag, and panic if a periph
tries to send more than 256 commands.
|
| 1.262.10.3 | 22-May-2012 |
riz | branches: 1.262.10.3.2;
Pull up following revision(s) (requested by jakllsch in ticket #277):
sys/dev/scsipi/scsiconf.c: revision 1.268
Fix locking issue triggered by drvctr -r of a scsibus(4).
Pullup to netbsd-6 should be requested.
|
| 1.262.10.2 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.262.10.1 | 19-Mar-2012 |
riz | Pull up following revision(s) (requested by mrg in ticket #125):
sys/dev/scsipi/scsiconf.c: revision 1.263
sys/dev/scsipi/scsiconf.c: revision 1.264
sys/dev/ic/ncr53c9x.c: revision 1.144
sys/dev/ic/ninjascsi32.c: revision 1.22
sys/dev/usb/uhub.c: revision 1.117
take the kernel lock in functions called from attach*().
scsidevdetached ioctl path enters scsipi code without kernel lock
and this upsets the newer kasserts. take kernel lock here.
take the kernel lock a few more places when doing detach, to avoid
triggering KERNEL_LOCK_P() asserts in both scsi and usb code.
with this and other recent fixes i can now "drvctl -d ehci0".
|
| 1.262.10.3.2.1 | 01-Nov-2012 |
matt | sync with netbsd-6-0-RELEASE.
|
| 1.262.8.3 | 02-Jun-2012 |
mrg | sync to latest -current.
|
| 1.262.8.2 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.262.8.1 | 12-Mar-2012 |
mrg | merge scsiconf.c 1.264.
|
| 1.262.4.4 | 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")
|
| 1.262.4.3 | 30-Oct-2012 |
yamt | sync with head
|
| 1.262.4.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.262.4.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.269.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.269.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.269.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.272.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.273.4.4 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.273.4.3 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.273.4.2 | 09-Jul-2016 |
skrll | Sync with HEAD
|
| 1.273.4.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.275.2.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.275.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.278.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.279.6.3 | 19-Dec-2020 |
martin | Pull up following revision(s) (requested by tsutsui in ticket #1640):
sys/dev/scsipi/scsiconf.c: revision 1.283
Add NOLUNS quirk for more SEAGATE SCA/WIDE drives.
Tested on NetBSD/luna68k and LUNA with SCA 80pin -> NARROW 50pin and
WIDE 68pin -> NARROW 50pin connectors.
|
| 1.279.6.2 | 13-Jul-2020 |
martin | Pull up following revision(s) (requested by kim in ticket #1571):
sys/dev/scsipi/scsiconf.c: revision 1.288
Continue scanning a SCSI bus when a LUN is reported not present
This fixes disk attachment under Qemu when there is no disk on LUN 0 on
a SCSI bus but there is a disk on LUN 1. The inquiry for LUN 0 returns
SID_QUAL_LU_NOTPRESENT & T_NODEVICE. Quirks are only checked if neither
one of those are set, so cannot use a quirk entry.
Use case 1: Proxmox 6 configures each disk on its own bus when using
the "Virtio SCSI single" SCSI controller. However, while the "scsi0"
disk is on LUN 0, the "scsi1" disk is on LUN 1.
Use case 2: A Linode boot profile with multiple disks results in
the first disk ("sda") on LUN 1, while the second disk ("sdb") is
on LUN 0, each on their own bus.
|
| 1.279.6.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.280.6.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.280.6.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.280.6.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.280.4.3 | 18-Jan-2019 |
pgoyette | Synch with HEAD
|
| 1.280.4.2 | 20-Oct-2018 |
pgoyette | Sync with head
|
| 1.280.4.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.284.4.2 | 13-Jul-2020 |
martin | Pull up following revision(s) (requested by kim in ticket #1000):
sys/dev/scsipi/scsiconf.c: revision 1.288
Continue scanning a SCSI bus when a LUN is reported not present
This fixes disk attachment under Qemu when there is no disk on LUN 0 on
a SCSI bus but there is a disk on LUN 1. The inquiry for LUN 0 returns
SID_QUAL_LU_NOTPRESENT & T_NODEVICE. Quirks are only checked if neither
one of those are set, so cannot use a quirk entry.
Use case 1: Proxmox 6 configures each disk on its own bus when using
the "Virtio SCSI single" SCSI controller. However, while the "scsi0"
disk is on LUN 0, the "scsi1" disk is on LUN 1.
Use case 2: A Linode boot profile with multiple disks results in
the first disk ("sda") on LUN 1, while the second disk ("sdb") is
on LUN 0, each on their own bus.
|
| 1.284.4.1 | 04-May-2020 |
martin | Pull up following revision(s) (requested by jdc in ticket #885):
sys/dev/scsipi/scsiconf.c: revision 1.287
Don't attempt to read opcodes and their timeouts at attach time for
old devices. The MAINTENANCE IN command was introduced with SCSI-3
and sending it to older peripherals can cause timeouts or them not
to respond to further requests.
|
| 1.285.2.1 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.290.4.4 | 03-Apr-2021 |
thorpej | config_attach_loc() -> config_attach() with CFARG_LOCATORS argument.
|
| 1.290.4.3 | 28-Mar-2021 |
thorpej | Unwrap a line.
|
| 1.290.4.2 | 21-Mar-2021 |
thorpej | CFARG_IATTR usage audit:
If a device carries only one interface attribute, there is no need
to specify it when calling config_search(); that specification is
meant only to disambiguate which interface attribute (which is a
proxy for "what kind of attach args are being used") is having
children attached. cfparent_match() will take care of ensuring that
any potential children can attach to one of the parent's iterface
attributes, and if the parent only carries one, no disambiguation is
necessary.
|
| 1.290.4.1 | 20-Mar-2021 |
thorpej | The proliferation if config_search_*() and config_found_*() combinations
is a little absurd, so begin to tidy this up:
- Introduce a new cfarg_t enumerated type, that defines the types of
tag-value variadic arguments that can be passed to the various
config_*() functions (CFARG_SUBMATCH, CFARG_IATTR, and CFARG_LOCATORS,
for now, plus a CFARG_EOL sentinel).
- Collapse config_search_*() into config_search() that takes these
variadic arguments.
- Convert all call sites of config_search_*() to the new signature.
Noticed several incorrect usages along the way, which will be
audited in a future commit.
|
| 1.291.8.1 | 04-Aug-2021 |
thorpej | Adapt to CFARGS().
|
| 1.304.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.58 | 20-Apr-2012 |
bouyer | Add a bustype_async_event_xfer_mode() callback to scsipi_bustype (which can
be NULL), so that transport-specific details of transfer mode setting/printing
can be handled more easily.
Move scsipi_async_event_xfer_mode() and scsipi_print_xfer_mode() to
scsi_base.c and split in parallel scsi and FC/SAS parts.
size of struct scsipi_bustype has changed, welcome to 6.99.5
|
| 1.57 | 19-Apr-2012 |
bouyer | Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.56 | 16-Jul-2008 |
drochner | branches: 1.56.28; 1.56.32; 1.56.34;
split device/softc for scsibus
|
| 1.55 | 28-Apr-2008 |
martin | branches: 1.55.2; 1.55.4; 1.55.6;
Remove clause 3 and 4 from TNF licenses
|
| 1.54 | 11-Dec-2005 |
christos | branches: 1.54.70; 1.54.72; 1.54.74;
merge ktrace-lwp.
|
| 1.53 | 17-Sep-2004 |
mycroft | Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.52 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.51 | 21-Aug-2004 |
thorpej | - De-__P.
- Use ANSI function decls.
|
| 1.50 | 06-Jan-2003 |
matt | branches: 1.50.2; 1.50.4;
Add multiple inclusion protection.
|
| 1.49 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.48 | 20-Oct-1999 |
enami | branches: 1.48.8;
Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.47 | 05-Dec-1998 |
mjacob | branches: 1.47.10; 1.47.12; 1.47.14;
Eliminate the moreluns entry as it makes no sense for fat SCSI busses (e.g,
FC loops). Change the semantics of scsi_probedev so that it returns 1 if
you should continue probing at this target, else 0 for not.
Replace the blanket use of '7' with the use of the new sc_maxlun property
that is now gathered from HBAs. Allocate scsipi_link arrays based upon this.
Fix a really nasty and silly bug that has been there for a while where the
number of first level scsipi_link structures was one less than it needed
to be.
|
| 1.46 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.45 | 20-Oct-1998 |
thorpej | Fix a sight open flags buglet pointed out by Matthias Scheler.
|
| 1.44 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.43 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.42 | 28-Mar-1998 |
christos | Split out the verbose printing of scsi messages, so that user programs
can use the printing routines.
|
| 1.41 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.40 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.39 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.38 | 27-Aug-1997 |
bouyer | branches: 1.38.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.37 | 20-Aug-1997 |
mjacob | Add in prototype for routine to print SCSI errors in detail.
|
| 1.36 | 07-Jul-1997 |
cgd | branches: 1.36.2;
mark prototypes for static inline functions as possibly unused (with
__attribute__ ((unused))), to avoid generating warnings when compiling
without optimization but with most ports' default warning flags.
|
| 1.35 | 02-Apr-1997 |
mycroft | Push the buffer cleanup code into scsi_done(), and split it so that biodone()
is called *after* the driver `done' routine. This fixes disk I/O statistics
on SCSI devices.
Also, calling the `done' routine with a `complete' argument of 0 and actually
having it do anything meaningful loses in at least 3 ways, so just nuke the
argument altogether and don't call it this way. If the driver needs to do
some error handling, that's what `err_handler' is for.
|
| 1.34 | 10-Dec-1996 |
thorpej | Add a "max_target" member to struct scsi_link, which is filled in by
host adapter drivers, indicating the highest SCSI target they can
address. Use this value to dynamically allocate data structures, rather
than hard-coding 8 targets.
These changes allow targets > 7 to be addressed on wide SCSI busses.
Fixes PRs #1674 and #2892.
|
| 1.33 | 23-Oct-1996 |
matthias | * At least the Teac FC-1 Shugart-SCSI bridgeboard does motor on/off
management by itself. But when it gets a start unit request, it keeps
the floppy motor running all the time. This adds code for dealing with
yet another quirk (SDEV_NOSTARTUNIT) that prevents sd.c from sending
start unit requests. A entry for the Teac FC-1 is added to the quirk
table.
|
| 1.32 | 03-Sep-1996 |
thorpej | Store the SCSI version in the scsi_link, and set the LUN in the CDB
if the version is <= SCSI-2. This should help some older SCSI
devices that previously needed the "NOLUNS" quirk. While this is
not strictly necessary on SCSI-2 devices, the spec allows it,
so we set it for SCSI-2 devices "just in case". See section 7.2.2 of
Draft X3T9.2 Rev 10L for details.
|
| 1.31 | 28-Aug-1996 |
cgd | (1) add a 'channel' field to scsi_link.
(2) in scsibusmatch, match channel as appropriate.
(3) add a scsiprint() function, to do the "scsibus at..."
and channel (if not SCSI_CHANNEL_ONLY_ONE) printing,
i.e. the common functionality that all SCSI drivers currently
should be doing.
|
| 1.30 | 05-Jul-1996 |
christos | - Apply PR/2535: Add support for flex scsi disks.
- Add a quirk called SCSI_NOMODESENSE for drives like the iomega jaz,
that don't support mode sense.
|
| 1.29 | 19-Mar-1996 |
mycroft | branches: 1.29.4;
Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.28 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.27 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.26 | 12-Jan-1996 |
thorpej | Handle cases like the following:
- controller calls scsi_done() with error XS_TIMEOUT
- scsi_done() calls sddone()
- sddone() calls disk_unbusy()
- scsi_done() calls controller to retry command (missing the
call to disk_busy())
- controller calls scsi_done()
- scsi_done() calls sddone()
- sddone() calls disk_busy(), which panics because of the imbalance.
Bug noticed by Leo Weppleman, who also suggested this fix; pass an additional
boolean argument ("complete") to the device's "done" routine, with a
value of `0' passed from the previous call to "done", and add an additional
call to "done" when the xfer resources are freed.
|
| 1.25 | 12-Aug-1995 |
mycroft | minphys() functions really should return void.
|
| 1.24 | 24-Jul-1995 |
cgd | restore a deleted semicolon
|
| 1.23 | 24-Jul-1995 |
cgd | kill structure offest comments in non-fixed-size structures.
In some of our worlds, pointers just Aren't 4 bytes.
|
| 1.22 | 24-Jul-1995 |
cgd | update SCSI minphys routines' definitions to match standard minphys()
definition and usage.
|
| 1.21 | 12-Jul-1995 |
cgd | add a SDEV_FORCELUNS flag, as suggested by Jason Thorpe, to force
the SCSI code to look for LUNs on prehistoric (pre-SCSI 1) drives &
controllers, where normally it would not.
|
| 1.20 | 26-Jan-1995 |
mycroft | Update copyrights.
|
| 1.19 | 26-Jan-1995 |
mycroft | Redo the `wait for spin up' code. The Mach 3 method clearly doesn't work.
Also, rearrange the first open sequences a bit; SDEV_OPEN is no longer magic.
|
| 1.18 | 30-Dec-1994 |
mycroft | Fix some oversights in the debugging code.
|
| 1.17 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.16 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.15 | 03-Nov-1994 |
mycroft | Always use direct configuration for SCSI devices.
|
| 1.14 | 03-Nov-1994 |
mycroft | Remove a couple of i386-specific hacks, and make a couple of others #ifdef
CONFIG_INDIRECT instead.
|
| 1.13 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.12 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.11 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.10 | 12-May-1994 |
mycroft | Nuke some junk in scsi_link, as pointed out by Gordon.
|
| 1.9 | 20-Apr-1994 |
mycroft | Maintain the xs free list with LIST_*().
|
| 1.8 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.7 | 11-Apr-1994 |
mycroft | Combine scsi_start_unit() and scsi_stop_unit(), and increase the start
timeout (again).
|
| 1.6 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.5 | 25-Mar-1994 |
mycroft | Put controller target in scsi_switch.
|
| 1.4 | 20-May-1993 |
cgd | branches: 1.4.3;
add rcsids and clean up file headers
|
| 1.3 | 04-May-1993 |
deraadt | support for making dev->id_alive be set, this is for iostat to
find disk devices. wee bit of a kludge. sub-device attach()
routines must now return 1 for successful attach(), 0 otherwise.
Other bsd's do this too..
|
| 1.2 | 12-Apr-1993 |
deraadt | new scsi subsystem.
changes also in config/mkioconf.c
i386/isa/wd.c, fd.c, and all scsi drivers.
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.4.3.5 | 01-Feb-1994 |
mycroft | Add scsi_stop_unit(), from Mac code. Other minor nits.
|
| 1.4.3.4 | 28-Nov-1993 |
mycroft | Correct some function return types.
|
| 1.4.3.3 | 24-Nov-1993 |
mycroft | More construction...
|
| 1.4.3.2 | 24-Nov-1993 |
mycroft | Still under construction...
|
| 1.4.3.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.29.4.1 | 04-Mar-1997 |
mycroft | Pull up support for fetching geometry info on optical disks.
|
| 1.36.2.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.38.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.38.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.38.2.1 | 27-Aug-1997 |
thorpej | file scsiconf.h was added on branch marc-pcmcia on 1997-08-27 23:33:21 +0000
|
| 1.47.14.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.47.12.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.47.10.2 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.47.10.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.48.8.2 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.48.8.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.50.4.1 | 11-Sep-2004 |
he | Pull up revision 1.52 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.50.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.50.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.50.2.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.54.74.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.54.74.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.54.72.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.54.70.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.54.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.55.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.55.4.1 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.55.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.56.34.1 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.56.32.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.56.28.1 | 23-May-2012 |
yamt | sync with head.
|
| 1.35 | 10-Feb-2025 |
jakllsch | Fix misspelling of unused #define
|
| 1.34 | 09-Nov-2024 |
mlelstv | Add VPD descriptor definitions.
|
| 1.33 | 25-Dec-2007 |
perry | branches: 1.33.136;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.32 | 01-Dec-2006 |
martin | branches: 1.32.24; 1.32.30; 1.32.34; 1.32.38;
On Christos' request, reinstantiate the packed->__packed__ part of the
backed out change.
|
| 1.31 | 01-Dec-2006 |
martin | Back out align(4) for all struct decls - it breaks various things all
over the tree (for example esiop on sparc64 won't boot any more).
|
| 1.30 | 30-Nov-2006 |
christos | PR/34202: Dave Huang: Accessing cd device on alpha causes kernel trap
From the PR:
Since struct scsipi_read_discinfo_data is an odd number of bytes long
(35 bytes), ti ends up being unaligned. wdc_datain_pio() uses
bus_space_read_multi_4() to read the response from the drive, and
since ti is unaligned, it causes an unaligned access fault.
Add __aligned__(4) to all the structures.
|
| 1.29 | 11-Dec-2005 |
christos | branches: 1.29.20; 1.29.22;
merge ktrace-lwp.
|
| 1.28 | 21-Feb-2005 |
thorpej | branches: 1.28.4;
Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.27 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.26 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.25 | 16-Oct-2003 |
mycroft | branches: 1.25.8; 1.25.10;
Do a 36-byte SCSI 2 inquiry first, and iff that returns an additional length
>32 do a 74-byte inquiry. Fixes problems with devices that barf on longer
inquiries. (Linux uses 36 bytes almost everywhere, as a data point.)
|
| 1.24 | 08-Sep-2003 |
mycroft | Perform the exorcism on scsipi_mode_select() too.
|
| 1.23 | 08-Sep-2003 |
mycroft | Exorcise something evil and wrong in scsipi_mode_sense().
|
| 1.22 | 19-Nov-2001 |
soren | branches: 1.22.16;
Remove duplicate struct scsipi_start_stop.
|
| 1.21 | 02-Sep-2001 |
thorpej | Comment what the PF bit means in the MODE SELECT command, and
define a scsipi_mode_page_header, which has the common fields
(page code and page length) of all mode pages.
|
| 1.20 | 02-Sep-2001 |
thorpej | Centralize the definition of the page code mask and page-savable bit
in the pages returned by MODE SENSE.
|
| 1.19 | 22-May-2001 |
bouyer | branches: 1.19.2;
Remove one extra byte from scsipi_mode_select; pointed out by Juergen
Hannken-Illjes.
|
| 1.18 | 18-May-2001 |
is | When abusing structs to describe on-wire (or in-hardware) data structures,
they need to be packed. (Without this, breaks on at least m68k.)
|
| 1.17 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.16 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.15 | 15-May-2000 |
dante | branches: 1.15.6;
Add few peripheral device type
Remove "???" from T_IT8_1/2. They actually are pre-press devices for graphic arts as described by ASC IT8
Zeros and blanks scsipi_inquiry_data from byte 58 to byte 74 if additional_length is less than 58
|
| 1.14 | 14-May-2000 |
dante | Fix a typo
|
| 1.13 | 14-May-2000 |
dante | Change scsipi_inquiry_data strucure to be ANSI SPC-2 rev16 compliant
|
| 1.12 | 30-Dec-1998 |
dbj | branches: 1.12.8;
Added parens around ADD_BYTES_LIM() macro definition to protects its
use of the low precedence ?: conditional operator. This fixes a bug
in scsipi_interpret_sense when printing out debugging info.
|
| 1.11 | 19-Nov-1998 |
thorpej | Protect against multiple inclusion.
|
| 1.10 | 11-Jul-1998 |
mjacob | Clarify that Sense Key 1 is a "RECOVERED ERROR"- not just an error.
|
| 1.9 | 01-Jul-1998 |
mjacob | oh, for goodness sake, define the bloody Sense Keys...
|
| 1.8 | 24-Jun-1998 |
mjacob | add the SCSI-2/SCSI-3 Group ID definitions
|
| 1.7 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.6 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.5 | 03-Oct-1997 |
thorpej | Clean up and comment the device types, and add the "enclosure services",
"storage array", and "IT8" types.
|
| 1.4 | 02-Oct-1997 |
mjacob | kern/3022: make sure that printing sense info in SCSIDEBUG case starts at
the right place and contains the right number of bytes.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipi_all.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: some spaces -> tabs, indenting
consistency in prototypes.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipi_all.h was added on branch marc-pcmcia on 1997-08-27 23:33:22 +0000
|
| 1.12.8.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.12.8.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.15.6.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.15.6.2 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.15.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.19.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.19.2.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.22.16.4 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.22.16.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.22.16.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.22.16.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.25.10.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.25.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.28.4.2 | 21-Jan-2008 |
yamt | sync with head
|
| 1.28.4.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.29.22.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.29.20.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.32.38.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.32.34.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.32.30.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.32.24.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.33.136.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.193 | 29-Oct-2024 |
nat | Revert to previous - scsipi_done_once is not necessary.
A follow up commit will contain a simpler change (from riastradh@) to
ncr5380sbc.
|
| 1.192 | 28-Oct-2024 |
nat | Allow medium errors to be retried.
This was necessary for some variants of the BlueSCSI-v2. For these devices
retrying upon receiving a medium error they would work.
Addreses kern/58452.
As posted to tech-kern:
https://mail-index.netbsd.org/tech-kern/2024/08/02/msg029652.html
Ok thorpej@.
|
| 1.191 | 28-Oct-2024 |
nat | Introduce scsipi_done_once.
This allows for transfers to be sucessfully aborted on the ncr5380sbc(4).
This may be usefull in future for other scsi controllers.
Callers of scsipi_done are not affected by this change.
Part of kern/58452.
As posted to tech-kern:
https://mail-index.netbsd.org/tech-kern/2024/08/02/msg029652.html
Ok thorpej@.
|
| 1.190 | 14-Jun-2024 |
kardel | branches: 1.190.2;
Ignore unit attention caused EIO errors when attempting to fetch
supported op-codes and their timeout values during device attachment.
|
| 1.189 | 09-Apr-2022 |
riastradh | sys: Use membar_release/acquire around reference drop.
This just goes through my recent reference count membar audit and
changes membar_exit to membar_release and membar_enter to
membar_acquire -- this should make everything cheaper on most CPUs
without hurting correctness, because membar_acquire is generally
cheaper than membar_enter.
|
| 1.188 | 12-Mar-2022 |
riastradh | sys: Membar audit around reference count releases.
If two threads are using an object that is freed when the reference
count goes to zero, we need to ensure that all memory operations
related to the object happen before freeing the object.
Using an atomic_dec_uint_nv(&refcnt) == 0 ensures that only one
thread takes responsibility for freeing, but it's not enough to
ensure that the other thread's memory operations happen before the
freeing.
Consider:
Thread A Thread B
obj->foo = 42; obj->baz = 73;
mumble(&obj->bar); grumble(&obj->quux);
/* membar_exit(); */ /* membar_exit(); */
atomic_dec -- not last atomic_dec -- last
/* membar_enter(); */
KASSERT(invariant(obj->foo,
obj->bar));
free_stuff(obj);
The memory barriers ensure that
obj->foo = 42;
mumble(&obj->bar);
in thread A happens before
KASSERT(invariant(obj->foo, obj->bar));
free_stuff(obj);
in thread B. Without them, this ordering is not guaranteed.
So in general it is necessary to do
membar_exit();
if (atomic_dec_uint_nv(&obj->refcnt) != 0)
return;
membar_enter();
to release a reference, for the `last one out hit the lights' style
of reference counting. (This is in contrast to the style where one
thread blocks new references and then waits under a lock for existing
ones to drain with a condvar -- no membar needed thanks to mutex(9).)
I searched for atomic_dec to find all these. Obviously we ought to
have a better abstraction for this because there's so much copypasta.
This is a stop-gap measure to fix actual bugs until we have that. It
would be nice if an abstraction could gracefully handle the different
styles of reference counting in use -- some years ago I drafted an
API for this, but making it cover everything got a little out of hand
(particularly with struct vnode::v_usecount) and I ended up setting
it aside to work on psref/localcount instead for better scalability.
I got bored of adding #ifdef __HAVE_ATOMIC_AS_MEMBAR everywhere, so I
only put it on things that look performance-critical on 5sec review.
We should really adopt membar_enter_preatomic/membar_exit_postatomic
or something (except they are applicable only to atomic r/m/w, not to
atomic_load/store_*, making the naming annoying) and get rid of all
the ifdefs.
|
| 1.187 | 17-Sep-2020 |
jakllsch | Some misspelling-in-comments fixes for scsipi
|
| 1.186 | 13-Apr-2020 |
chs | slightly change and fix the semantics of pool_set*wat(), pool_sethardlimit()
and pool_prime() (and their pool_cache_* counterparts):
- the pool_set*wat() APIs are supposed to specify thresholds for the count of
free items in the pool before pool pages are automatically allocated or freed
during pool_get() / pool_put(), whereas pool_sethardlimit() and pool_prime()
are supposed to specify minimum and maximum numbers of total items
in the pool (both free and allocated). these were somewhat conflated
in the existing code, so separate them as they were intended.
- change pool_prime() to take an absolute number of items to preallocate
rather than an increment over whatever was done before, and wait for
any memory allocations to succeed. since pool_prime() can no longer fail
after this, change its return value to void and adjust all callers.
- pool_setlowat() is documented as not immediately attempting to allocate
any memory, but it was changed some time ago to immediately try to allocate
up to the lowat level, so just fix the manpage to describe the current
behaviour.
- add a pool_cache_prime() to complete the API set.
|
| 1.185 | 19-Feb-2020 |
riastradh | branches: 1.185.4;
Sprinkle some dtrace probes into scsi(4).
|
| 1.184 | 10-Nov-2019 |
chs | branches: 1.184.2;
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.
|
| 1.183 | 19-Sep-2019 |
msaitoh | Use unsigned to avoid undefined behavior in scsipi_{get,put}_tag().
Found by kUBSan.
|
| 1.182 | 28-Mar-2019 |
kardel | branches: 1.182.4;
Add reading of supported opcodes and their timeouts
at attachment time. Though this information is optional,
it allows to override our fixed timeouts with device
provided timeouts. These timeouts will override the
hardcoded values if the device provided timeouts
exceed the hardcoded values and are less than a day.
Using the device provided timeouts avoids premature
device resets and unreliable operation due to
inadequate timeouts.
Due to the limited implementations of USB
umass devices this feature is disabled for all
umass attached devices.
|
| 1.181 | 05-Feb-2019 |
mrg | avoid a fallthru checker complaint and make one case like the rest. NFCI.
|
| 1.180 | 24-Nov-2018 |
bouyer | Add more KASSERTS about locking.
|
| 1.179 | 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!)
|
| 1.178 | 14-Jul-2017 |
christos | branches: 1.178.2; 1.178.4; 1.178.6;
Async event can be called before the adapter is running (pmax tc asc)
|
| 1.177 | 19-Jun-2017 |
mlelstv | pass config_detach error to caller.
|
| 1.176 | 17-Jun-2017 |
mlelstv | The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.175 | 22-Dec-2016 |
mlelstv | branches: 1.175.8;
fix comment. request_sense is called unlocked.
|
| 1.174 | 18-Dec-2016 |
skrll | KNF
No functional change.
|
| 1.173 | 18-Dec-2016 |
skrll | Whitespace
|
| 1.172 | 18-Dec-2016 |
skrll | mlelstv accidentaly dropped a mutex_enter
|
| 1.171 | 18-Dec-2016 |
mlelstv | The mutex passed to cv_wait must also be held when calling cv_broadcast.
Also optimizing mutex handling in completion thread.
From nick@.
|
| 1.170 | 16-Dec-2016 |
mlelstv | Add locking for periph_active and flags. The operations aren't atomic.
|
| 1.169 | 29-Nov-2016 |
mlelstv | reference count adapter mutex possibly shared by multiple channels.
fix error in atapibusdetach, when a child device cannot be detached,
keep atapibus instance alive.
|
| 1.168 | 21-Nov-2016 |
mlelstv | Lock is already taken when handling async events, don't lock again
in scsipi_lookup_periph.
Fixes PR kern/51641.
|
| 1.167 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.166 | 02-Oct-2016 |
jdolecek | change scsipi_execute_xs() to default to simple tags for !XS_CTL_URGENT
if not specified by caller; this is mostly for documentation purposes
only, as sd(4) and cd(4) explicitly use simple tags already
|
| 1.165 | 24-Aug-2015 |
pooka | branches: 1.165.2;
would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.164 | 18-Nov-2014 |
joerg | branches: 1.164.2;
Use size for the size argument of memcmp, not the result of a compare.
|
| 1.163 | 17-Nov-2014 |
christos | PR/49054: Add a quirk for the ES-6600 RAID controller which does not do
INQUIRY3 properly. Unfortunately looking at the length does not solve
the problem since other devices send greater lengths too.
|
| 1.162 | 18-Oct-2014 |
snj | src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.161 | 06-Oct-2014 |
christos | PR/49054: Uwe Toenjes: Some RAID controllers return more bytes in the
scsi 3 inquiry command than expected by the size of the scsi 3 inquiry
packet. This can be detected by looking at the additional_length field
returned by the scsi 2 inquiry. If that's the case, avoid doing the
scsi 3 inquiry because we can't handle the extra bytes later.
XXX: Pullup -7
|
| 1.160 | 13-Jul-2014 |
dholland | branches: 1.160.2;
"peripheral", not "peripherial".
|
| 1.159 | 20-Apr-2012 |
bouyer | branches: 1.159.2; 1.159.12;
Add a bustype_async_event_xfer_mode() callback to scsipi_bustype (which can
be NULL), so that transport-specific details of transfer mode setting/printing
can be handled more easily.
Move scsipi_async_event_xfer_mode() and scsipi_print_xfer_mode() to
scsi_base.c and split in parallel scsi and FC/SAS parts.
size of struct scsipi_bustype has changed, welcome to 6.99.5
|
| 1.158 | 19-Apr-2012 |
bouyer | Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.157 | 18-Apr-2012 |
bouyer | Fix KASSERT(): autoconf doesn't run under the KERNEL_LOCK
|
| 1.156 | 20-Feb-2012 |
mrg | assert kernel lock is held in a few places in inside scsipi.
lock the kernel when calling into scsipi from umass and usscanner.
with these two in place on usbmp branch, umass appears stable.
|
| 1.155 | 13-Nov-2010 |
uebayasi | branches: 1.155.8; 1.155.12; 1.155.14;
Don't pull in the whole uvm(9) API to access only PAGE_SIZE and
some other constants. These are provided by sys/param.h now.
|
| 1.154 | 23-Aug-2010 |
pooka | Convert one second(!!) delay to kpause. It was originally done
from a callout, so delay was the only option (in those days). Then
the caller was converted to a thread, but left as a delay. It
still may block the scsipi completion thread processing (so I
seriously doubt this code path is executed very often on a live
system).
|
| 1.153 | 07-Jun-2010 |
pgoyette | Update scsiverbose module to use module_autoload() rather than module_load().
Load the module right before each attempt to use its features, and let the
module subsystem handle unloading.
|
| 1.152 | 30-May-2010 |
pgoyette | Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.151 | 12-Feb-2010 |
pooka | branches: 1.151.2;
delay() is used here, so need to include machine/param.h.
|
| 1.150 | 21-Oct-2009 |
rmind | branches: 1.150.2;
Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.149 | 07-Apr-2009 |
dyoung | Destroy a scsipi_xfer's callout before putting it back into the pool.
|
| 1.148 | 11-May-2008 |
mlelstv | branches: 1.148.6; 1.148.12;
Reintroduce the NODOORLOCK quirk. Helps with PR kern/23875.
|
| 1.147 | 28-Apr-2008 |
martin | branches: 1.147.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.146 | 05-Apr-2008 |
cegger | branches: 1.146.2; 1.146.4;
use aprint_*_dev and device_xname
|
| 1.145 | 09-Jul-2007 |
ad | branches: 1.145.28;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.144 | 12-Mar-2007 |
ad | branches: 1.144.2;
Pass an ipl argument to pool_init/POOL_INIT to be used when initializing
the pool's lock.
|
| 1.143 | 09-Feb-2007 |
ad | branches: 1.143.2; 1.143.6;
Merge newlock2 to head.
|
| 1.142 | 26-Nov-2006 |
itohy | If the block size reported by Read Capacity looks valid, just use it.
Use Request Sense only if Read Capacity succeeded and did not return
valid block size.
Discussed on tech-kern.
Fix the easier part of NetBSD PR kern/26537.
(The harder part is the device hangs on large (>= 8KB) transfer.
Possibly umass BBB problem?)
Remove scsipi_size() and scsipi_validate_secsize() from scsipi_base.c
and add their functions to sd.c since they are used only by sd.c.
Use SCSI term `block' instead of `sector' where applicable.
|
| 1.141 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.140 | 20-Oct-2006 |
scw | Validate the sector size returned by READ CAPACITY. If it looks bogus
print a warning and fallback to a suitable default.
Fixes a problem on hp700 reported by skrll@
|
| 1.139 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.138 | 09-Oct-2006 |
scw | Some removable umass(4) devices don't respond to mode sense page 6, or
simply return zero for logical block size. In either case, use the sector
length reported by READ_CAPACITY instead of defaulting to 512 bytes.
This partially addresses the problems reported in PR port-i386/34707 and
PR kern/34737. Namely the incorrectly reported drive geometry and the
'hanging' issue.
However, since the device in question reports 2048-byte physical sectors
it will remain unusable until DEV_BSIZE is banished.
|
| 1.137 | 11-Sep-2006 |
reinoud | branches: 1.137.2;
Aparently i've misread, backing out change.
|
| 1.136 | 11-Sep-2006 |
reinoud | Don't add one to the capacity returned of READ CAPACITY! it results in
reporting a sd* disc that is one sector too big (!) Normally not much a
problem in FFS since its clustered but other filingsystems *do* care.
|
| 1.135 | 17-Apr-2006 |
nathanw | branches: 1.135.8;
In scsipi_done(), bail out early if the scsipi_xfer has already been
marked as done. Works around one problem with detaching in-use sd
devices, as described on tech-kern:
http://mail-index.netbsd.org/tech-kern/2005/09/22/0002.html
|
| 1.134 | 20-Feb-2006 |
thorpej | branches: 1.134.2; 1.134.4; 1.134.6;
Use device_is_active() rather than testing dv_flags for DVF_ACTIVE
directly.
|
| 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 | 31-May-2005 |
xtraeme | branches: 1.131.2;
Add missing const.
|
| 1.130 | 29-May-2005 |
christos | - Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.129 | 04-Apr-2005 |
yamt | scsipi_completion_thread:
if we seemt to need more resources after doing ADAPTER_REQ_GROW_RESOURCES,
yield cpu rather than busy-looping. PR/25164.
|
| 1.128 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.127 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.126 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.125 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.124 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.123 | 07-Dec-2004 |
thorpej | branches: 1.123.2; 1.123.4;
READ_CAPACITY -> READ_CAPACITY_10
|
| 1.122 | 03-Dec-2004 |
thorpej | And thus spake SBC-3:
If the number of logical blocks exceeds the maximum value that is able to be
specified in the RETURNED LOGICAL BLOCK ADDRESS field, the device server
shall set the RETURNED LOGICAL BLOCK ADDRESS field to FFFFFFFFh. The
application client should then issue a READ CAPACITY (16) command (see 5.11)
to retrieve the READ CAPACITY (16) parameter data.
Implement this in scsipi_size().
First issue in kern/28514.
|
| 1.121 | 01-Oct-2004 |
enami | Don't touch free'ed object. Fixes l_holdcnt leak reported by Andreas Wrede
on current-users.
|
| 1.120 | 18-Sep-2004 |
mycroft | If our enqueue failed -- because we're polling and there is already something
in the queue -- do not attempt to requeue it. We only poll in two cases:
1) We have a non-interrupting controller. In this case, execution of the
previous command should have left the queue empty. (Perhaps there should be
a KASSERT() to this effect.)
2) We're in the shutdown path, either doing a cache sync or a dump. In these
cases, the retry behavior is useless, because we will no longer get
interrupts to notify us that the earlier commands completed. Instead we
just spin for a few seconds and fail anyway. (XXX We should probably clear
the queue explicitly so that the shutdown/dump will succeed.)
|
| 1.119 | 18-Sep-2004 |
mycroft | Since we always defer probing SCSI and ATAPI devices now, we can never be
"cold" -- so change the check for this to a KASSERT().
|
| 1.118 | 18-Sep-2004 |
mycroft | Oops; eliminate the EJUSTRETURN return value from scsipi_execute_xs(). Don't
know how this worked when I tested it.
|
| 1.117 | 18-Sep-2004 |
mycroft | Minor rearrangement. Whitespace and #include cleanup.
|
| 1.116 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.115 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.114 | 17-Sep-2004 |
mycroft | Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.113 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.112 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.111 | 02-Sep-2004 |
chs | make this compile on platforms where PAGE_SIZE is not a compile-time constant.
|
| 1.110 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.109 | 23-Aug-2004 |
bouyer | Fix typo pointed out by Patrick Welche on current-users@
|
| 1.108 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.107 | 18-Aug-2004 |
drochner | Use the new autoconf functions to rescan busses and detach devices
on user request.
This duplicates the functionality provided by a private ioctl
interface (accessible through scsictl(8)), but in a more generic way.
|
| 1.106 | 04-Aug-2004 |
bouyer | Fix some problems in the scsipi detach code:
- only call periph_switch->psw_start() if the device is active; if it is not
psw_start() may try to access invalid data.
- remove the TAILQ_FIRST(&periph->periph_xferq) != NULL diagnostic in
scsipi_kill_pending(). This can't be true at this point (unless the device
was idle at the time of the detach), because the scsipi channel kernel
thread has to run for the queue to be flushed.
There are still other problems to be fixed here ...
|
| 1.105 | 27-Apr-2004 |
bouyer | Revert part of 1.102:
Don't decrease/check xs_retries when the device report "Power On, Reset, or
Bus Device Reset" sense condition, just retry the command. The initial bus
reset would cause the first TEST_UNIT_READY to report this condition,
and as xs_retries is set to 0 when XS_CTL_DISCOVERY is set, it would report an
error instead of being retried, causing the disk probe to report "drive
offline" instead of the geometry and capacity. Checking/decreasing
xs_retries on the bus reset reported by the adapter is enouth to avoid the
problem reported by rev 1.102.
Problem analysed by Paul Kranenburg, fix confirmed by Anders Hjalmars,
and explaination as to why the INQUIRY wasn't affected by this provided by
Bill Studenmund.
|
| 1.104 | 16-Mar-2004 |
bouyer | branches: 1.104.2;
cbd -> cdb
Command Block Descriptor -> Command Descriptor Block
Pointed out by Allen Briggs.
|
| 1.103 | 15-Mar-2004 |
bouyer | Extract the code printing the CBD from scsipi_print_sense(), so that it's
usable in other context.
Use the new scsipi_print_cbd() to dump the command in case of timeout
in siop/esiop.
|
| 1.102 | 10-Mar-2004 |
bouyer | Decrease xs_retries before retrying aborted commands, and report EIO if it
reaches 0. Avoids looping on aborded command in some special cases.
|
| 1.101 | 03-Jan-2004 |
thorpej | callout_init() after memset().
|
| 1.100 | 18-Nov-2003 |
briggs | Use aprint_normal instead of printf in scsipi_print_xfer_mode().
|
| 1.99 | 28-Oct-2003 |
simonb | Fix bogus uninitialised warning.
|
| 1.98 | 17-Oct-2003 |
mycroft | Add a comment explaining the INQUIRE behavior.
|
| 1.97 | 16-Oct-2003 |
mycroft | Whoops, set cmd_length correctly for the 36-byte INQUIRE.
|
| 1.96 | 16-Oct-2003 |
mycroft | Do a 36-byte SCSI 2 inquiry first, and iff that returns an additional length
>32 do a 74-byte inquiry. Fixes problems with devices that barf on longer
inquiries. (Linux uses 36 bytes almost everywhere, as a data point.)
|
| 1.95 | 12-Oct-2003 |
thorpej | Revert previous change.
|
| 1.94 | 10-Oct-2003 |
thorpej | Warn if a sync period requiring DT was negotiated on a peripheral not
capable of DT.
|
| 1.93 | 09-Sep-2003 |
mycroft | Exorcise PQUIRK_NODOORLOCK.
|
| 1.92 | 08-Sep-2003 |
mycroft | Do a START UNIT only if the TEST UNIT READY reports that the device is not
ready. This avoids gratuitously starting the motor on floppy and CD-ROM
drives, and eliminates the need for the audio playing test in cdopen().
Therefore, also remove PQUIRK_NOSTARTUNIT.
|
| 1.91 | 08-Sep-2003 |
mycroft | On further investigation... don't panic if we try to do a 6-byte mode
sense/select on ATAPI. Some tape drives require it. And who knows? Some
fool might wire a RBC device to ATAPI.
|
| 1.90 | 08-Sep-2003 |
mycroft | Perform the exorcism on scsipi_mode_select() too.
|
| 1.89 | 08-Sep-2003 |
mycroft | Exorcise something evil and wrong in scsipi_mode_sense().
|
| 1.88 | 19-Apr-2003 |
fvdl | branches: 1.88.2;
A device should always respond to inquiry or test unit ready; disable
retries for these during discovery. From Pascal Renauld at Network
Storage Solutions, Inc
|
| 1.87 | 16-Apr-2003 |
nathanw | Remove extra right paren.
|
| 1.86 | 16-Apr-2003 |
thorpej | * Change the APIs that handle the sync period to work with 100ths of ns,
rather than 10ths of ns. This is necessary in order to represent
Ultra320 SCSI.
* Add Ultra320 SCSI to the scsipi_syncparams[] table.
We're not going to bother bumping any version numbers with this change; only
the "hba" driver uses scsipi_sync_period_to_factor(), and the uses of
scsipi_sync_factor_to_period() are all internal to the scsipi code. Most
things just pass the factor around, which is unchanged by this.
Reviewed by Frank van der Linden.
|
| 1.85 | 03-Apr-2003 |
fvdl | Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.84 | 03-Feb-2003 |
thorpej | Test callout_pending(), not callout_active(), and eliminate now-unnecessary
callout_deactivate() calls.
|
| 1.83 | 20-Jan-2003 |
simonb | Only declare and set the "info" variable #ifndef SCSIVERBOSE.
|
| 1.82 | 24-Nov-2002 |
scw | Quell an uninitialised variable warning.
|
| 1.81 | 09-Nov-2002 |
thorpej | Fix signed/unsigned comparison warnings.
|
| 1.80 | 04-Oct-2002 |
soren | Leave error printing to the callers of scsipi_size().
|
| 1.79 | 19-Sep-2002 |
jmc | Force the initial probes to happen within the newly forked off kthread.
This eliminates problems where the underlying interrupt handler isn't the
specific layer calling scsipi_complete() for a given scsi transaction.
This avoids deadlocks where the kthread that called the autoconf routines
to configure a scsibus shouldn't be the one put to sleep waiting on a
scsipi_complete (only the scsibus's kthread should be doing that).
To avoid jitter this will force the scsibus's to probe in the order they
run through autoconf (so machines with multiple bus's don't move sd* devices
around on every reboot).
|
| 1.78 | 26-Jul-2002 |
wiz | Spell '[Rr]ight' correctly. From Jim Bernard.
|
| 1.77 | 05-Jun-2002 |
mjacob | As per a discussion on tech-kern a while back- honor retries for
XS_SELTIMEOUT and XS_TIMEOUT errors- but only do so if the device exists
already.
Devices that are being probed won't be found via periph_lookup, so we can
use that to find if a device exists for the purposes of honoring retries.
|
| 1.76 | 03-Jun-2002 |
bouyer | Doh, the return of scsipi_lookup_periph() was meant to be assigned to
periph.
periph was used uninitialised, and caused a panic when the scsibus is reset
with siop or esiop, and possibly others HBA drivers.
|
| 1.75 | 17-May-2002 |
mjacob | branches: 1.75.2;
Give XS_DRIVER_STUFFUP a case all by itself.
|
| 1.74 | 16-May-2002 |
thorpej | Don't use a 2-dimensional array for the channel's periph table. Instead,
hash the target and lun together and use a hash table. This will be
necessary in order to support very large (64-bit) LUN number spaces.
|
| 1.73 | 15-May-2002 |
bouyer | branches: 1.73.2;
scsipi_print_xfer_mode(): PERIPH_CAP_DT is an equivalent of
(PERIPH_CAP_SYNC | PERIPH_CAP_WIDE16) here.
|
| 1.72 | 05-May-2002 |
bouyer | If periph->periph_callout is already active, don't freeze the periph again:
scispi_periph_timed_thaw() will be called only one time anyway.
|
| 1.71 | 01-Apr-2002 |
bouyer | Add a chan_name to struct scsipi_channel, holding the channel's name.
Set this to dv_xname for scsibus and atapibus.
Set the name of the kernel thread to chan_name instead of controller's
name:channel number (so that we can use this name for controller-specific
threads).
|
| 1.70 | 28-Mar-2002 |
christos | PR/16110: Chris Jepeway: scsipi_complete() calls (*psw_done)() b/4
setting buffer fields
|
| 1.69 | 16-Mar-2002 |
bouyer | Present the same interface to periph driver for ASYNC scsipi_command() in the
normal case, and in the polling or "no thread yet" cases:
don't return an error from scsipi_complete(), as the error should already have
been handled in scsipi_complete() and eventually periph driver callbacks.
Should fix kern/15190.
|
| 1.68 | 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.
|
| 1.67 | 21-Feb-2002 |
enami | Fix broken indentation.
|
| 1.66 | 12-Jan-2002 |
tsutsui | Call malloc(9) with M_ZERO flag instead of memset() after malloc().
|
| 1.65 | 28-Nov-2001 |
fredette | Added the new defopt SCSI_OLD_NOINQUIRY; this is used instead
of sun2 to wrap code that conjures up INQUIRY responses for
certain specific old devices.
|
| 1.64 | 26-Nov-2001 |
fredette | Added quirk entries for Adaptec and Emulex SCSI interposer boards.
Decode a limited set of SASI/SCSI-1 sense codes, and under sun2
only, conjure up INQUIRY responses for these boards.
|
| 1.63 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.62 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.61 | 14-Oct-2001 |
bouyer | Refuse to register a callback if the completion thread isn't started yet.
|
| 1.60 | 14-Oct-2001 |
bouyer | Call ADAPTER_REQ_GROW_RESOURCES from the completion thread, if possible.
This allows HBA drivers to call bus_dmamem_map() safely.
|
| 1.59 | 14-Oct-2001 |
bouyer | Split channel flags in chan_flags used for communications between
scsipi and HBA, and chan_tflags used for communications between scsipi
and its kernel thread. No functionnal change.
|
| 1.58 | 27-Sep-2001 |
mjacob | Restore previous functionality- in scsipi_periph_timed_thaw check to
make sure the completion thread is running before you try to schedule
it.
Fixes port-i386/14013
|
| 1.57 | 21-Sep-2001 |
fvdl | Unfortunately, the previous change seems to make most (all?) configurations
using the ahc driver hang when probing scsi devices. The problem may be in
the ahc driver and not in this change, but back it out until this has been
fixed anyway.
|
| 1.56 | 18-Sep-2001 |
mjacob | Rather than run a periph's queue from scsipi_periph_timed_thaw which is
called via a callout, kick the completion thread to run it for us (uses
a new flag, SCSIPI_CHAN_KICK).
If we've received BUSY status and we haven't started the completion
thread yet, don't freeze do a callout to scsipi_periph_timed_thaw which
then will try and kick the completion thread- instead treat the command
as if it were a polled command and just call delay for 1 second.
If DIAGNOSTIC is defined, and the periph qfreeze count is less than
zero, panic because some HBA has corrupted the periph structure's
accounting.
|
| 1.55 | 01-Sep-2001 |
mjacob | branches: 1.55.2;
Add table value for Ultra3, so configuring an Ultra3 disk shows:
sd1: sync (12.5ns offset 14), 16-bit (160.000MB/s) transfers, tagged queueing enabled
instead of:
sd1: sync (36.0ns offset 14), 16-bit (55.554MB/s) transfers, tagged queueing enabled
|
| 1.54 | 20-Aug-2001 |
ad | Cosmetic change.
|
| 1.53 | 20-Aug-2001 |
ad | Medium Not Present has number of defined ASCQs. Pointed out by Sergey
Svishchev <svs@ropnet.ru> in PR 8326.
|
| 1.52 | 19-Aug-2001 |
yamt | fix scsipi_target_detach with wildcard target.
|
| 1.51 | 18-Jul-2001 |
bouyer | Adn scsipi_target_detach() and scsipi_thread_call_callback() as discussed
on tech-kern. scsipi_target_detach() accept wildcard target/lun as requested.
|
| 1.50 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.49 | 13-Jul-2001 |
bouyer | scsipi_set_xfer_mode(): issue a ADAPTER_REQ_SET_XFER_MODE request to adapter
only if we succesfully attached at last one device for this I_T.
|
| 1.48 | 27-Jun-2001 |
ross | branches: 1.48.2;
compile tweak for non-SCSIVERBOSE
|
| 1.47 | 27-Jun-2001 |
bouyer | Better diagnostic when a REQUEST SENSE is terminated with CHECK CONDITION.
|
| 1.46 | 26-Jun-2001 |
bouyer | Add a XS_CTL_SILENT_NODEV flag: if the sense info is "not ready, medium
not present" don't print any message but still return ENODEV.
Use this in cd driver to allow open of character raw partition even if
the drive is empty (older drives fails at LOAD_UNLOAD time, newer
ones fail at TEST_UNIT_READY time).
|
| 1.45 | 13-Jun-2001 |
bjh21 | Add explicit support for IDE and SCSI adaptors which don't support interrupts.
On such adaptors, all transfers are done in polling mode.
OK'ed by Manuel on tech-kern.
|
| 1.44 | 23-May-2001 |
bouyer | In scsipi_channel_thaw(), if the count drops to 0, call scsipi_run_queue().
|
| 1.43 | 18-May-2001 |
enami | Don't capitalize the word sync or async. It's inconsistient with other
messages.
|
| 1.42 | 18-May-2001 |
bouyer | Workaround for broken drives (explained to me by Chris G. Demetriou):
some devices can't handle tag number larger than some values and always
reject commands with QUEUE FULL if the tag number is larger than this.
Under some circonstances the scsipi system may not decrease periph_openings
(as a workaround of other odd behavior) and we may end up requeuing the command
with a hight tag value again, and the driver could loop on this.
Workaround: in case of queue full, decrease periph_openings to
min(periph_active - 1, periph_openings - 1) so that, after some iteration,
periph_openings is less than the max tag value acceptable by the device.
Solves the problem with tagged queuing on ncr53c9x for me.
|
| 1.41 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.40 | 27-Apr-2001 |
bouyer | Don't forget to call psw_done() !
|
| 1.39 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.38 | 09-Jun-2000 |
enami | branches: 1.38.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.37 | 31-May-2000 |
fvdl | Initialize xs_status to 0 after allocating a scsipi_xfer struct. Makes life
easier for driver debugging.
|
| 1.36 | 27-May-2000 |
fvdl | At least try to do something useful in the XS_BUSY case; don't cause
a panic by sleeping in an interrupt context.
|
| 1.35 | 23-May-2000 |
bouyer | branches: 1.35.2;
scsipi_get_xs(): if we have XS_CTL_URGENT, return an xfer even if
active >= openings. An XS_CTL_URGENT command could otherwise fail, especially
if openings == 1.
|
| 1.34 | 03-Apr-2000 |
enami | When killing pending xfers on device detach, we can't expect scsipi_done
to remove all xfers from the pending queue. It removes only xfers for
asynchronous transactions. So, simply loop over all pending xfers
with calling scsipi_done and wait xfers to drain. Addresses PR#9703.
|
| 1.33 | 03-Apr-2000 |
enami | Fold long line.
|
| 1.32 | 02-Apr-2000 |
augustss | With SCSIVERBOSE, only print sense data if there actually is a problem.
|
| 1.31 | 02-Apr-2000 |
enami | Add missing protect from disk interrupt while calling scsipi_free_xs.
|
| 1.30 | 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.
|
| 1.29 | 17-Jan-2000 |
bouyer | Use SCSIPIRETRIES instead of hard-coded value '2' for number of retries in
common routines. Define SCSIPIRETRIES as 4, so that the command will finally
succeed after several consecutive downgrades from UDMA2 to PIO4.
|
| 1.28 | 14-Jan-2000 |
mjacob | Print out the contents of an otherwise unreported undecodable
sense data buffer. This helps catch adapter breakage mostly.
|
| 1.27 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.26 | 17-Oct-1999 |
enami | branches: 1.26.2; 1.26.4;
ENODEV is not a value supposed to be assigned to xs->error. Use
XS_DRIVER_STUFFUP instead.
|
| 1.25 | 06-Oct-1999 |
mjacob | If we're discovering, don't tsleep on lbolt if we had a busy status.
|
| 1.24 | 30-Sep-1999 |
thorpej | branches: 1.24.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.23 | 11-Sep-1999 |
thorpej | Implement a function to kill off all commands pending for a given
scsipi_link.
|
| 1.22 | 16-Jun-1999 |
pk | In scsipi_done(), call scsipi_free_xs() at splbio().
|
| 1.21 | 07-Apr-1999 |
bouyer | Now that we do real use of CDIOCCLOSE, we can have SSS_START|SSS_LOEJ
in scsipi_start(). Adjust test so that timeout will let enouth time to
the drive to spin up.
|
| 1.20 | 02-Feb-1999 |
bouyer | branches: 1.20.2;
If sense_key == UNIT_ATTENTION and ASC/ASQ == "Power On, Reset, or Bus Device
Reset", retry the command instead of returning an error. XS_RESET is useless
without this, as the retryed command will die with this unit attention.
|
| 1.19 | 29-Jan-1999 |
bouyer | Return ENODEV instead of EIO when we are trying to open a device without media
in the drive. restrict "opening of empty drive" to character devices only
(reading a block device returns a short read instead of ENODEV, which can lead
to confusion).
|
| 1.18 | 19-Jan-1999 |
bouyer | Move test for SDEV_NOSTARTUNIT quirk from sd.c to scsipi_start().
Add a SDEV_NOSTARTUNIT quirk entry for BCD-16X 1997-04-25", "", "VER 2.2"
CD-rom (from Michael Santos).
|
| 1.17 | 10-Jan-1999 |
mjacob | Make the error condition of neither 0x70 or 0x71 error codes a tad
more informative than the lower case 'error code %d'.
|
| 1.16 | 08-Dec-1998 |
thorpej | Actually, restructure scsipi_wait_drain() slightly.
|
| 1.15 | 08-Dec-1998 |
thorpej | - When allocating a scsipi_xfer, add it to the specified device's
pending_xfers queue.
- When freeing a scsipi_xfer, remove it from the device's pending_xfers
queue. If the queue is empty, and SDEV_WAITDRAIN is set, wakeup
those waiting for the queue to drain.
- Implement scsipi_wait_drain(), which waits for a device's pending_xfers
queue to drain.
|
| 1.14 | 19-Nov-1998 |
thorpej | Add support for reference counting and enabling/disabling SCSI and
ATAPI controllers. This will eventually be used for power management
of e.g. PCMCIA SCSI and IDE controller cards.
|
| 1.13 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.12 | 10-Oct-1998 |
thorpej | If the adapter returned XS_RESET and the xfer specifies a retry count,
attempt to reissue the command (which was destroyed by the bus reset).
Slightly modified from PR #6090, from Matt Jacob.
|
| 1.11 | 18-Sep-1998 |
scottr | SCSI_ASYNCREQ turns out to be redundant; we can simply free the scsipi_xfer
in scsipi_done() if the transfer is asynchronous. This reduces the size
of the critical section in scsipi_execute_xs() somewhat (in fact,
back to its original size).
|
| 1.10 | 16-Sep-1998 |
scottr | Found a race in scsipi_execute_xs(): if an asynchronous transfer completes
(probably due to an interrupt) between the time it is scheduled and the
time we get around to setting the SCSI_ASYNCREQ flag, we can lose the xs.
Fix this by checking to see if the transfer has already completed after
the scsi_cmd function returns SUCCESSFULLY_QUEUED, and just return to the
caller if so.
|
| 1.9 | 14-Sep-1998 |
scottr | Fix a problem uncovered when we started to use the pool allocator to manage
scsipi_xfer structures.
When scsipi_execute_xs() calls the driver's scsi_cmd function, it assumes
that it can still dereference a pointer to the scsipi_xfer struct. Since
scsipi_done() has already been called, which in turn has called
scsipi_free_xs(), the struct has already been returned to the pool! In
other words, xs->flags has been compromised, but we are still testing it.
These changes resolve the problem by doing the following:
- In scsipi_execute_xs(), if the lower-level driver's scsi_cmd function
returns SUCCESSFULLY_QUEUED and SCSI_NOSLEEP is set in xs->flags, set a
new flag (SCSI_ASYNCREQ). This indicates that scsipi_done() should free
the scsipi_xfer struct.
If the lower-level driver's scsi_cmd function returns SUCCESSFULLY_QUEUED
but SCSI_NOSLEEP is not set, we wait (via tsleep()) for the request to
complete, then fall through to the COMPLETE case.
If the lower-level driver's scsi_cmd function returns COMPLETE, we now
simply return any actual errors, or 0 if none occurred. (Previously,
we may have returned EJUSTRETURN, of which the sole effect was to
avoid freeing the scsipi_xfer struct in our caller. No code seems
to depend on this behavior, however.)
- In scsipi_done(), only free the scsipi_xfer struct for async requests.
The contents of the struct will otherwise remain valid until the
function that initiated the transfer frees it.
With this change, responsibility for freeing the struct now lies in two
places, depending on the type of the request:
- For synchronous requests, the routine calling scsipi_execute_xs()
must clean up.
- For asynchronous requests, scsipi_done() cleans up (as it always has).
[Note: this change also corrects a problem with sddump(): scsipi_done()
was attempting to return a static scsipi_xfer struct to the pool! Since
dumps are performed synchronously, we now handle this correctly.]
This solution was provided by Jason Thorpe, after I got him to look at
some related (but insufficient) attempts of my own.
|
| 1.8 | 15-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.7 | 11-Aug-1998 |
scottr | Eliminate a potential (but not common) NULL dereference.
|
| 1.6 | 31-Jul-1998 |
thorpej | Use the pool allocator for scsipi_xfer structures.
|
| 1.5 | 10-Feb-1998 |
thorpej | branches: 1.5.2;
Clear the scsipi_xfer's status byte before executing the command. This
fixes a condition where stale data can be left in the status byte, causing
user programs that interpret it to fail.
Fixes kern/4964 from Chris Jones <cjones@honors.montana.edu>.
|
| 1.4 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipi_base.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 14-Aug-1997 |
bouyer | Sync with trunk.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipi_base.c was added on branch marc-pcmcia on 1997-08-27 23:33:23 +0000
|
| 1.5.2.1 | 08-Aug-1998 |
eeh | Revert cdevsw mmap routines to return int.
|
| 1.20.2.4 | 23-Jan-2000 |
he | Pull up revision 1.29 (via patch, requested by bouyer):
Use SCSIPIRETRIES instead of hard-coded value '2' for number of
retries in common routines, and define it as 4 so that ATAPI
command will succeed after several downgrade.
|
| 1.20.2.3 | 16-Jan-2000 |
he | Apply patch (requested by ad):
When probing a SCSI target, do not report an error if the target
indicates that LUNs are not supported.
|
| 1.20.2.2 | 24-Jun-1999 |
perry | pullup 1.21->1.22 (pk): deal with missing "raise interrupt level" code
|
| 1.20.2.1 | 07-Apr-1999 |
bouyer | branches: 1.20.2.1.2; 1.20.2.1.4;
Pull up rev 1.21: fix timeout for SSS_START|SSS_LOEJ in scsipi_start().
|
| 1.20.2.1.4.1 | 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.
|
| 1.20.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.24.2.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.26.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.26.2.17 | 23-Apr-2001 |
bouyer | In channel completion thread, tsleep() at splbio, to avoid race condition
where scsipi_done would put something in the queue and wakeup between
splx(s) and tsleep().
|
| 1.26.2.16 | 23-Apr-2001 |
mjacob | Don't let the chan_qfreeze count go negative.
It's possible that the adapters shouldn't do this. But if you think about
it, this is a function that could be called from anywhere. In any case,
the specific reason here was that there is *sometimes* one more Loop UP
message from QLogic FC cards than then Loop Down, so making the chan_qfreeze
go negative blocked the queue needlessly.
Add a splbio/splx around the test against XS_STS_DONE in scsipi_execute_xs
for the non-async case so we don't do the classic sleep after test crock.
|
| 1.26.2.15 | 22-Apr-2001 |
bouyer | Keep track of the number of commands per periph really sent to the adapter.
In scsipi_run_queue(), compare this to periph_openings instead of periph_active
When a drive reject all queued commands with QUEUE FULL (my IBM DDRS34560D
does this on large writes) we end up with all commands in the queue, and
periph_active > periph_openings.
While doing this I found a bug in scsipi_async_event_channel_reset(): calling
scsipi_done() would corrupt accounting.
|
| 1.26.2.14 | 11-Apr-2001 |
mjacob | Make scsipi_channel_init a function returning an int- non-zero means
it failed to initialize the channel (this should be acceptable)- in
which case we complain and don't schedule bus probing for later.
We make the internal memory allocations for the periph and the chan_periphs
array M_NOWAIT- this way we have a hope of booting instead of silently hanging
during boot if we've run out of memory.
|
| 1.26.2.13 | 03-Apr-2001 |
bouyer | When we send a untagged command, block the queue until this command is
complete.
|
| 1.26.2.12 | 28-Mar-2001 |
bouyer | Better handling of errors for request sense commands.
|
| 1.26.2.11 | 27-Mar-2001 |
bouyer | Kill scsipi_link from comments.
|
| 1.26.2.10 | 22-Jan-2001 |
bouyer | Don't forget to remove xfer from the channel's queue in
scsipi_async_event_channel_reset()
thaw the periph when request_sense failed(), and when calling user done,
as we're not going though the usual path.
Implement XS_REQUEUE, for drivers needing to requeue a xfer.
|
| 1.26.2.9 | 15-Jan-2001 |
bouyer | Handle CHECK CONDITION status in mid-layer:
the REQUEST_SENSE command is generated from scsipi_complete() so it can
tsleep() (we use a regular scsipi_command() call for this). Add a new
periph_flag, PERIPH_SENSE, and a new xs_control, XS_CTL_REQSENSE.
When PERIPH_SENSE is set only xfer with XS_CTL_REQSENSE may be sent
(same logic as PERIPH_RECOVERING but with higther priority). xfer with
XS_CTL_REQSENSE need to have XS_CTL_URGENT urgent too.
XS_CTL_USERCMD xfers are now handled in scsipi_complete().
We need to pay special attention to SCSI resets, as we may have:
- an aborted REQUEST_SENSE. In this case we need to requeue the original
command, not the request sense.
- sense pending for a command no longer in the queue but for which a
request sense has not yet been queued. In this case we should not issue
the request sense but requeue the original command instead.
For this:
- the xfer with the CHECK CONDITION status is stored in the scsipi_periph
(periph_xscheck); the CHECK CONDITION is tested in scsipi_done, PERIPH_SENSE
and periph_xscheck is set here.
- XS_CTL_REQSENSE xfers are not allowed to be requeued, and are terminated
with EINTR in case of reset.
- we have a new async event, "ASYNC_EVENT_RESET", which cleanup the
xfer in periph_xscheck.
- appropriate splbio/splx to avoid race condition (especially,
scsipi_request_sense() runs at splbio. Should not be a real problem as it
doesn't happen often.
While I'm there kill req_sense_length from struct scsipi_xfer, it's always set
to 0 and only checked by a few drivers.
|
| 1.26.2.8 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.26.2.7 | 04-Feb-2000 |
thorpej | Make sure the channel's completion thread exits when the channel
is detached, and completely encapsulate the way periph's are stored
by the channel.
|
| 1.26.2.6 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.26.2.5 | 26-Oct-1999 |
thorpej | Deal a little more gracefully with the fact that xfer mode parameters
are for the I_T Nexus, and make all xfer mode updates `async events'.
|
| 1.26.2.4 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.26.2.3 | 20-Oct-1999 |
thorpej | Add channel freeze/thaw, and when a timed thaw occurs on either a
channel or periph, make sure to kick the channel queue.
|
| 1.26.2.2 | 19-Oct-1999 |
thorpej | Add an asynchronous event for the adapter to notify the midlayer that
xfer mode parameters have changed for an I_T Nexus.
|
| 1.26.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.35.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.38.4.15 | 11-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.38.4.14 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.38.4.13 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.38.4.12 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.38.4.11 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.38.4.10 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.38.4.9 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.38.4.8 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.38.4.7 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.38.4.6 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.38.4.5 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.38.4.4 | 08-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.38.4.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.38.4.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.38.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.48.2.9 | 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
|
| 1.48.2.8 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.48.2.7 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.48.2.6 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.48.2.5 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.48.2.4 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.48.2.3 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.48.2.2 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.48.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.55.2.1 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.73.2.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.73.2.2 | 20-Jun-2002 |
gehenna | catch up with -current.
|
| 1.73.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.75.2.1 | 04-Jun-2002 |
lukem | Pull up revision 1.76 (requested by bouyer in ticket #149):
Doh, the return of scsipi_lookup_periph() was meant to be assigned to
periph.
periph was used uninitialised, and caused a panic when the scsibus is reset
with siop or esiop, and possibly others HBA drivers.
|
| 1.88.2.11 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.88.2.10 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.88.2.9 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.88.2.8 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.88.2.7 | 24-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.88.2.6 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.88.2.5 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.88.2.4 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.88.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.88.2.2 | 12-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.88.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.104.2.2 | 11-Sep-2004 |
he | Pull up revisions 1.110-1.112 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.104.2.1 | 29-Apr-2004 |
jmc | Pullup rev 1.105 (requested by bouyer in ticket #214)
Revert part of 1.102:
Don't decrease/check xs_retries when the device report "Power On, Reset, or
Bus Device Reset" sense condition, just retry the command. The initial bus
reset would cause the first TEST_UNIT_READY to report this condition,
and as xs_retries is set to 0 when XS_CTL_DISCOVERY is set, it would report an
error instead of being retried, causing the disk probe to report "drive
offline" instead of the geometry and capacity. Checking/decreasing
xs_retries on the bus reset reported by the adapter is enouth to avoid the
problem reported by rev 1.102.
|
| 1.123.4.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.123.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.131.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.131.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.131.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.131.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.133.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.133.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.133.2.1 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.134.6.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.134.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.134.2.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.135.8.6 | 06-Feb-2007 |
ad | Now that kthreads always run with kernel priority, don't rely on preempt()
to actually yield the CPU.
|
| 1.135.8.5 | 30-Jan-2007 |
ad | Remove support for SA. Ok core@.
|
| 1.135.8.4 | 19-Jan-2007 |
ad | Put back arg to preempt() for nwo.
|
| 1.135.8.3 | 19-Jan-2007 |
ad | Acquire proclist_mutex before sending signals.
|
| 1.135.8.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.135.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.137.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.137.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.143.6.6 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.143.6.5 | 13-May-2007 |
ad | - Pass the error number and residual count to biodone(), and let it handle
setting error indicators. Prepare to eliminate B_ERROR.
- Add a flag argument to brelse() to be set into the buf's flags, instead
of doing it directly. Typically used to set B_INVAL.
- Add a "struct cpu_info *" argument to kthread_create(), to be used to
create bound threads. Change "bool mpsafe" to "int flags".
- Allow exit of LWPs in the IDL state when (l != curlwp).
- More locking fixes & conversion to the new API.
|
| 1.143.6.4 | 10-Apr-2007 |
ad | Nuke the deferred kthread creation stuff, as it's no longer needed.
Pointed out by thorpej@.
|
| 1.143.6.3 | 09-Apr-2007 |
ad | - Add two new arguments to kthread_create1: pri_t pri, bool mpsafe.
- Fork kthreads off proc0 as new LWPs, not new processes.
|
| 1.143.6.2 | 05-Apr-2007 |
ad | Compile fixes.
|
| 1.143.6.1 | 13-Mar-2007 |
ad | Sync with head.
|
| 1.143.2.1 | 24-Mar-2007 |
yamt | sync with head.
|
| 1.144.2.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.145.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.146.4.5 | 09-Oct-2010 |
yamt | sync with head
|
| 1.146.4.4 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.146.4.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.146.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.146.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.146.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.147.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.148.12.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.148.6.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.150.2.3 | 22-Oct-2010 |
uebayasi | Sync with HEAD (-D20101022).
|
| 1.150.2.2 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.150.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.151.2.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.151.2.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.155.14.1 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.155.12.2 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.155.12.1 | 20-Feb-2012 |
mrg | pullup from -current:
>assert kernel lock is held in a few places in inside scsipi.
>lock the kernel when calling into scsipi from umass and usscanner.
>
>with these two in place on usbmp branch, umass appears stable.
|
| 1.155.8.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.155.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.159.12.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.159.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.159.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.160.2.1 | 26-Mar-2015 |
martin | Pull up the following revisions, requested by christos in #644:
sys/dev/scsipi/scsipi_base.c 1.161 - 1.164
Use size for the size argument of memcmp, not the result of a compare.
PR/49054: Add a quirk for the ES-6600 RAID controller which does not do
INQUIRY3 properly. Unfortunately looking at the length does not solve
the problem since other devices send greater lengths too.
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
PR/49054: Uwe Toenjes: Some RAID controllers return more bytes in the
scsi 3 inquiry command than expected by the size of the scsi 3 inquiry
packet. This can be detected by looking at the additional_length field
returned by the scsi 2 inquiry. If that's the case, avoid doing the
scsi 3 inquiry because we can't handle the extra bytes later.
|
| 1.164.2.5 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.164.2.4 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.164.2.3 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.164.2.2 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.164.2.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.165.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.165.2.1 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.175.8.2 | 07-Sep-2018 |
martin | Pull up following revision(s) (requested by mrg in ticket #1010):
sys/dev/scsipi/scsipi_base.c: revision 1.178
Async event can be called before the adapter is running (pmax tc asc)
|
| 1.175.8.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.178.6.4 | 21-Apr-2020 |
martin | Sync with HEAD
|
| 1.178.6.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.178.6.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.178.6.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.178.4.2 | 26-Nov-2018 |
pgoyette | Sync with HEAD, resolve a couple of conflicts
|
| 1.178.4.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.178.2.2 | 14-Jul-2017 |
christos | 2975092
|
| 1.178.2.1 | 14-Jul-2017 |
christos | file scsipi_base.c was added on branch perseant-stdc-iso10646 on 2017-07-14 17:50:12 +0000
|
| 1.182.4.1 | 31-Jan-2020 |
martin | Pull up following revision(s) (requested by msaitoh in ticket #672):
sys/dev/ic/tulip.c: revision 1.198
sys/dev/pci/if_jme.c: revision 1.45
sys/dev/pci/agp.c: revision 1.86
sys/dev/pci/if_lii.c: revision 1.27
sys/dev/acpi/thinkpad_acpi.c: revision 1.47
sys/dev/scsipi/scsipi_base.c: revision 1.183
sys/dev/ic/aic6915reg.h: revision 1.6
Fix undefined behavior in thinkpad_mask_init(). Found by kUBSan.
Use unsigned when initialize the transmit completion ring to avoid undefined
behavior. Found by kUBSan.
Avoid undefined behavior when setting multicast address. found by kUBSan.
Use unsigned to avoid undefined behavior in agpattach(). Found by kUBSan.
Avoid undefined behavior in jme_mii_write(). Found by kUBSan.
Use unsigned to avoid undefined behavior in lii_setmulti().
Use unsigned to avoid undefined behavior in scsipi_{get,put}_tag().
Found by kUBSan.
|
| 1.184.2.1 | 29-Feb-2020 |
ad | Sync with head.
|
| 1.185.4.1 | 20-Apr-2020 |
bouyer | Sync with HEAD
|
| 1.190.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.24 | 26-Feb-2017 |
maya | Add a DEBUG assert for the problem noted in PR kern/51731
of some drivers not sufficiently locking.
The relevant drivers might not work, but if someone gets around to them
then the problem could be visible.
sd doesn't trip the assert.
|
| 1.23 | 20-Nov-2016 |
mlelstv | branches: 1.23.2;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.22 | 28-Apr-2008 |
martin | branches: 1.22.44; 1.22.64; 1.22.68;
Remove clause 3 and 4 from TNF licenses
|
| 1.21 | 16-Nov-2006 |
christos | branches: 1.21.48; 1.21.50; 1.21.52;
__unused removal on arguments; approved by core.
|
| 1.20 | 16-Feb-2006 |
perry | branches: 1.20.14; 1.20.16;
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@.
|
| 1.19 | 24-Dec-2005 |
perry | branches: 1.19.2; 1.19.4; 1.19.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.18 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.17 | 01-Feb-2005 |
reinoud | branches: 1.17.6;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.16 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.15 | 18-Sep-2004 |
mycroft | branches: 1.15.4;
Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.14 | 21-Aug-2004 |
thorpej | - De-__P.
- Use ANSI function decls.
|
| 1.13 | 06-Jan-2003 |
matt | branches: 1.13.2;
Add multiple inclusion protection.
|
| 1.12 | 18-Jul-2001 |
thorpej | bcopy -> memcpy
|
| 1.11 | 25-Apr-2001 |
bouyer | branches: 1.11.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.10 | 28-Feb-2000 |
thorpej | branches: 1.10.6;
Make sure to initialize xs->req_sense_length (to 0). This appears to be
a historical leftover, and very few drivers actually look at it. But the
PCI `ncr' driver does, and if this field is not initialized to its default
value, that driver will fall over flat on its face when it attempts to get
sense data from a target.
|
| 1.9 | 08-Dec-1998 |
thorpej | branches: 1.9.4; 1.9.10;
Remove xs_free_list; it no longer exists.
|
| 1.8 | 15-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.7 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.6 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.5 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.4 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.3 | 19-Sep-1997 |
cgd | mark scsipi_make_xs() as potentially unused, so that this can be
compiled with -Wunused without optimization/inlining enabled.
Also, make it a little bit closer to KNF.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipi_base.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.3 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipi_base.h was added on branch marc-pcmcia on 1997-08-27 23:33:24 +0000
|
| 1.9.10.4 | 15-Jan-2001 |
bouyer | Handle CHECK CONDITION status in mid-layer:
the REQUEST_SENSE command is generated from scsipi_complete() so it can
tsleep() (we use a regular scsipi_command() call for this). Add a new
periph_flag, PERIPH_SENSE, and a new xs_control, XS_CTL_REQSENSE.
When PERIPH_SENSE is set only xfer with XS_CTL_REQSENSE may be sent
(same logic as PERIPH_RECOVERING but with higther priority). xfer with
XS_CTL_REQSENSE need to have XS_CTL_URGENT urgent too.
XS_CTL_USERCMD xfers are now handled in scsipi_complete().
We need to pay special attention to SCSI resets, as we may have:
- an aborted REQUEST_SENSE. In this case we need to requeue the original
command, not the request sense.
- sense pending for a command no longer in the queue but for which a
request sense has not yet been queued. In this case we should not issue
the request sense but requeue the original command instead.
For this:
- the xfer with the CHECK CONDITION status is stored in the scsipi_periph
(periph_xscheck); the CHECK CONDITION is tested in scsipi_done, PERIPH_SENSE
and periph_xscheck is set here.
- XS_CTL_REQSENSE xfers are not allowed to be requeued, and are terminated
with EINTR in case of reset.
- we have a new async event, "ASYNC_EVENT_RESET", which cleanup the
xfer in periph_xscheck.
- appropriate splbio/splx to avoid race condition (especially,
scsipi_request_sense() runs at splbio. Should not be a real problem as it
doesn't happen often.
While I'm there kill req_sense_length from struct scsipi_xfer, it's always set
to 0 and only checked by a few drivers.
|
| 1.9.10.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.9.10.2 | 04-Feb-2000 |
thorpej | Make sure the channel's completion thread exits when the channel
is detached, and completely encapsulate the way periph's are stored
by the channel.
|
| 1.9.10.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.9.4.1 | 28-Feb-2000 |
he | Pull up revision 1.10 (requested by thorpej):
Make sure to initialize the ``req_sense_length'' field. Fixes
a serious problem with the PCI ``ncr'' driver.
|
| 1.10.6.3 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.10.6.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.10.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.11.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.13.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.13.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.13.2.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.15.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.17.6.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.19.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.19.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.19.2.1 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.20.16.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.20.14.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.21.52.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.21.50.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.21.48.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.22.68.2 | 20-Mar-2017 |
pgoyette | Sync with HEAD
|
| 1.22.68.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.22.64.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.22.64.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.22.44.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.23.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.21 | 01-Apr-2009 |
reinoud | Implement MMC_TRACKINFO_DATA and MMC_TRACKINFO_AUDIO detection for cdrom and
dvdrom devices. These two compatibility codes were forgotten when those were
introduced.
|
| 1.20 | 08-Sep-2008 |
gmcgarry | branches: 1.20.2; 1.20.4; 1.20.8;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.
|
| 1.19 | 02-May-2008 |
reinoud | branches: 1.19.2; 1.19.6;
Add definitions for the close track/session, reserve track and repair
track SCSI commands.
|
| 1.18 | 25-Dec-2007 |
perry | branches: 1.18.6; 1.18.8; 1.18.10;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.17 | 01-Dec-2006 |
martin | branches: 1.17.24; 1.17.30; 1.17.34; 1.17.38;
On Christos' request, reinstantiate the packed->__packed__ part of the
backed out change.
|
| 1.16 | 01-Dec-2006 |
martin | Back out align(4) for all struct decls - it breaks various things all
over the tree (for example esiop on sparc64 won't boot any more).
|
| 1.15 | 30-Nov-2006 |
christos | PR/34202: Dave Huang: Accessing cd device on alpha causes kernel trap
From the PR:
Since struct scsipi_read_discinfo_data is an odd number of bytes long
(35 bytes), ti ends up being unaligned. wdc_datain_pio() uses
bus_space_read_multi_4() to read the response from the drive, and
since ti is unaligned, it causes an unaligned access fault.
Add __aligned__(4) to all the structures.
|
| 1.14 | 20-Dec-2005 |
reinoud | branches: 1.14.20; 1.14.22;
The as `unused' specified field in the readdiscinfo SCSI command is really
application_code to be able to write to -RW media when its in restricted
mode.
|
| 1.13 | 11-Dec-2005 |
simonb | Remove blank line added to end of file by previous change.
|
| 1.12 | 11-Dec-2005 |
reinoud | Add SCSI calls for read discinfo and read trackinfo. Also add the various
read toc formats returned.
|
| 1.11 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.10 | 01-Feb-2005 |
reinoud | branches: 1.10.6;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.9 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.8 | 30-Jan-2005 |
reinoud | The code asumes all programs want to know is TOC response format 0. Support
for TOC response format 1 and 2 are mandatory on CD/DVD too and provide
more information.
Next an IOCTL needs to be implemented that can read all TOC formats in a
generic way. This is pending.
|
| 1.7 | 07-Sep-2003 |
mycroft | branches: 1.7.8; 1.7.10;
Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.6 | 02-Sep-2001 |
thorpej | branches: 1.6.20;
Centralize the definition of the page code mask and page-savable bit
in the pages returned by MODE SENSE.
|
| 1.5 | 13-Feb-1998 |
enami | branches: 1.5.26; 1.5.28;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.4 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipi_cd.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: some spaces -> tabs, indenting
consistency in prototypes.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipi_cd.h was added on branch marc-pcmcia on 1997-08-27 23:33:25 +0000
|
| 1.5.28.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.5.26.1 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.6.20.4 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.6.20.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.6.20.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.6.20.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.7.10.1 | 12-Feb-2005 |
yamt | sync with head.
|
| 1.7.8.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.10.6.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.10.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.10.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.14.22.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.14.20.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.17.38.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.17.34.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.17.30.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.17.24.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.18.10.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.18.10.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.18.8.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.18.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.18.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.19.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.19.2.1 | 24-Sep-2008 |
wrstuden | Merge in changes between wrstuden-revivesa-base-2 and
wrstuden-revivesa-base-3.
|
| 1.20.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.20.4.1 | 08-Apr-2009 |
snj | Pull up following revision(s) (requested by reinoud in ticket #633):
sys/dev/scsipi/cd.c: revision 1.291
sys/dev/scsipi/scsipi_cd.h: revision 1.21
Implement MMC_TRACKINFO_DATA and MMC_TRACKINFO_AUDIO detection for cdrom and
dvdrom devices. These two compatibility codes were forgotten when those were
introduced.
|
| 1.20.2.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.17 | 28-Apr-2008 |
martin | Remove clause 3 and 4 from TNF licenses
|
| 1.16 | 11-Dec-2005 |
christos | branches: 1.16.70; 1.16.72; 1.16.74;
merge ktrace-lwp.
|
| 1.15 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.14 | 30-May-2001 |
mrg | branches: 1.14.22; 1.14.30; 1.14.32;
use _KERNEL_OPT
|
| 1.13 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.12 | 13-Feb-1998 |
enami | branches: 1.12.14; 1.12.26;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.11 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.10 | 24-Oct-1997 |
christos | PR/4328: Matthias Scheler: Syntax errors in debug macros.
|
| 1.9 | 03-Oct-1997 |
enami | branches: 1.9.2;
- Rename PIDEBUG back to SCSIDEBUG, since other files are still
using SCSIDEBUG.
- Don't #undef PIDEBUG unconditionally.
|
| 1.8 | 27-Aug-1997 |
bouyer | branches: 1.8.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.7 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.6 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.5 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.4 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.3 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.2 | 22-Jan-1994 |
briggs | If SCSI_DEBUG is on, a structure member wasn't updated from adapter_unit
to scsibus.
|
| 1.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.8.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.8.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.8.2.1 | 27-Aug-1997 |
thorpej | file scsipi_debug.h was added on branch marc-pcmcia on 1997-08-27 23:33:26 +0000
|
| 1.9.2.1 | 24-Oct-1997 |
mellon | Pull rev 1.10 up from trunk
|
| 1.12.26.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.12.14.1 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.14.32.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.14.30.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.14.22.1 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.16.74.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.16.72.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.16.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.24 | 27-Feb-2025 |
jakllsch | add more READ CAPACITY 16 data definitions
|
| 1.23 | 04-Nov-2017 |
jdolecek | branches: 1.23.42;
fix comment - scsipi_rw_big was renamed to scsipi_rw_10 a while ago
|
| 1.22 | 05-Apr-2017 |
jdolecek | expose disk device FUA/DPO support via DIOCGCACHE, and allow the flags
to be set for I/O; implement support in sd(4) and nvme(4)
discussed on tech-kern
|
| 1.21 | 25-Dec-2007 |
perry | branches: 1.21.54; 1.21.74; 1.21.78; 1.21.82;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.20 | 01-Jul-2007 |
tsutsui | branches: 1.20.8; 1.20.14; 1.20.16; 1.20.20;
Add __attribute__((__packed__)) to all command structures.
|
| 1.19 | 01-Dec-2006 |
martin | branches: 1.19.8; 1.19.10;
On Christos' request, reinstantiate the packed->__packed__ part of the
backed out change.
|
| 1.18 | 01-Dec-2006 |
martin | Back out align(4) for all struct decls - it breaks various things all
over the tree (for example esiop on sparc64 won't boot any more).
|
| 1.17 | 30-Nov-2006 |
christos | PR/34202: Dave Huang: Accessing cd device on alpha causes kernel trap
From the PR:
Since struct scsipi_read_discinfo_data is an odd number of bytes long
(35 bytes), ti ends up being unaligned. wdc_datain_pio() uses
bus_space_read_multi_4() to read the response from the drive, and
since ti is unaligned, it causes an unaligned access fault.
Add __aligned__(4) to all the structures.
|
| 1.16 | 11-Dec-2005 |
christos | branches: 1.16.20; 1.16.22;
merge ktrace-lwp.
|
| 1.15 | 01-Feb-2005 |
reinoud | branches: 1.15.6;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.14 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.13 | 07-Dec-2004 |
thorpej | branches: 1.13.2;
READ_CAPACITY -> READ_CAPACITY_10
|
| 1.12 | 07-Dec-2004 |
thorpej | USe more appropriate macro/struct names for READ/WRITE (6) and
READ/WRITE (10).
|
| 1.11 | 04-Dec-2004 |
thorpej | Use READ/WRITE (16) if the LBA is larger than 32 bits. Another bit
of kern/28514.
|
| 1.10 | 03-Dec-2004 |
thorpej | And thus spake SBC-3:
If the number of logical blocks exceeds the maximum value that is able to be
specified in the RETURNED LOGICAL BLOCK ADDRESS field, the device server
shall set the RETURNED LOGICAL BLOCK ADDRESS field to FFFFFFFFh. The
application client should then issue a READ CAPACITY (16) command (see 5.11)
to retrieve the READ CAPACITY (16) parameter data.
Implement this in scsipi_size().
First issue in kern/28514.
|
| 1.9 | 17-Sep-2003 |
mycroft | If READ CAPACITY fails, try a READ FORMAT CAPACITIES. Separate this into
another function, always doing a page 0 MODE SENSE to get the block
descriptor if we use READ CAPACITY, and use SMS_DBD on the page 4/5 MODE
SENSE. This does one extra command in some cases, but it separates and
simplifies the code a little.
Why do we prefer READ CAPACITY over READ FORMAT CAPACITIES? Two reasons:
1) It's much older and is much less likely to have had its command code
abused, and is thus "safer" to try first. 2) ALL of my USB flash readers
and pen drives screw up their capacity descriptors -- mostly off-by-one
errors in the size (they return the maximum LBA number instead, a la READ
CAPACITY, which has *never* been how READ FORMAT CAPACITIES was documented
in the MMC spec), and one returns the "no media" code on slots that have
media inserted (despite returning almost-correct data otherwise)!
F*** me with a chainsaw.
|
| 1.8 | 06-Jan-2003 |
matt | branches: 1.8.2;
Add multiple inclusion protection.
|
| 1.7 | 19-Nov-2001 |
soren | Remove duplicate struct scsipi_start_stop.
|
| 1.6 | 02-Sep-2001 |
thorpej | Add the Force Unit Access and Disable Page Out bits to the
READ(10) and WRITE(10) commands.
|
| 1.5 | 13-Feb-1998 |
enami | branches: 1.5.26; 1.5.28;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.4 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipi_disk.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipi_disk.h was added on branch marc-pcmcia on 1997-08-27 23:33:27 +0000
|
| 1.5.28.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.5.28.1 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.5.26.3 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.5.26.2 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.5.26.1 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.8.2.4 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.8.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.8.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.8.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.13.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.15.6.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.15.6.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.15.6.1 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.16.22.1 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.16.20.1 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.19.10.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.19.8.1 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.20.20.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.20.16.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.20.14.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.20.8.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.21.82.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.21.78.1 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.21.74.1 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.21.54.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.23.42.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.73 | 27-Dec-2019 |
msaitoh | s/transfered/transferred/
|
| 1.72 | 30-May-2019 |
mlelstv | use correct size when copying outgoing sense data.
|
| 1.71 | 26-May-2019 |
mlelstv | Add sanity checks to SCIOCCOMMAND, adapter drivers might be confused or trigger
assertions (e.g. umass).
|
| 1.70 | 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!)
|
| 1.69 | 20-Nov-2016 |
mlelstv | branches: 1.69.8; 1.69.14; 1.69.16;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.68 | 24-Aug-2015 |
pooka | branches: 1.68.2;
to garnish, dust with _KERNEL_OPT
|
| 1.67 | 19-Apr-2012 |
bouyer | branches: 1.67.2; 1.67.14; 1.67.16; 1.67.18; 1.67.22;
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.66 | 14-Jul-2008 |
drochner | branches: 1.66.28; 1.66.32; 1.66.34;
set the buffer's "resid" if a raw SCSI command fails, otherwise a
diagnostic check in kern_physio is set up
|
| 1.65 | 28-Apr-2008 |
martin | branches: 1.65.2; 1.65.4; 1.65.6;
Remove clause 3 and 4 from TNF licenses
|
| 1.64 | 02-Jan-2008 |
ad | branches: 1.64.6; 1.64.8; 1.64.10;
Merge vmlocking2 to head.
|
| 1.63 | 29-Jul-2007 |
ad | branches: 1.63.6; 1.63.12; 1.63.14; 1.63.18; 1.63.22;
It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.62 | 04-Mar-2007 |
christos | branches: 1.62.2; 1.62.10;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.61 | 14-Apr-2006 |
christos | branches: 1.61.14;
Coverity CID 1101: Protect against address argument being NULL.
|
| 1.60 | 28-Mar-2006 |
thorpej | Use device_unit().
|
| 1.59 | 25-Mar-2006 |
thorpej | Use device_parent().
|
| 1.58 | 01-Mar-2006 |
yamt | branches: 1.58.2; 1.58.4; 1.58.6;
merge yamt-uio_vmspace branch.
- use vmspace rather than proc or lwp where appropriate.
the latter is more natural to specify an address space.
(and less likely to be abused for random purposes.)
- fix a swdmover race.
|
| 1.57 | 14-Dec-2005 |
reinoud | branches: 1.57.2; 1.57.4; 1.57.6;
Allow SCIOCCOMMAND ioctl that executes a SCSI command to also be issued
within the kernel when the FKIOCTL flag is passed to the ioctl.
No code actually uses this yet but is committed for completion.
|
| 1.56 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.55 | 01-Nov-2005 |
martin | Rework the funny "user_strat: No ioctl" error messages - the functions
have not been called user_strat for quite some time and the message is
confusing if you do not know the code.
|
| 1.54 | 01-Nov-2005 |
bouyer | Don't claim there is a data in or out phase if the datalen is 0 (userland
shouldn't claim it either, but a buggy software shouldn't be able to crash
the kernel anyway). Should fix port-sparc64/31925 by Johan A.van Zanten
(which should really be kern/31925).
Analysed and patch tested by Martin Husemann.
|
| 1.53 | 31-Oct-2005 |
yamt | fix ioctl problems after the recent physio changes
in some drivers including wd and scsi.
- physio: if a caller provided a buf, stick to use it
because some drivers use it as an identifier.
- sprinkle simple_locks.
- scsistrategy: rather than issueing an async request and
waiting for its completion, simply issue a sync request.
the way to wait for the completion had an assumption that
B_CALL is never used. it isn't the case after the recent
physio() changes.
pointed/analyzed/tested by Martin Husemann.
|
| 1.52 | 01-Feb-2005 |
reinoud | branches: 1.52.4; 1.52.6; 1.52.8;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.51 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.50 | 18-Sep-2004 |
mycroft | branches: 1.50.4;
Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.49 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.48 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.47 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.46 | 29-Jun-2003 |
fvdl | branches: 1.46.2; 1.46.4;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.45 | 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
|
| 1.44 | 23-Feb-2003 |
yamt | initialize b_interlock properly.
(for ioctl)
|
| 1.43 | 12-Jan-2002 |
tsutsui | branches: 1.43.10;
Call malloc(9) with M_ZERO flag instead of memset() after malloc().
|
| 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 | 18-Jul-2001 |
thorpej | branches: 1.40.2;
bcopy -> memcpy
|
| 1.39 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.38 | 25-Apr-2001 |
bouyer | branches: 1.38.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.37 | 30-Sep-1999 |
thorpej | branches: 1.37.2; 1.37.8; 1.37.14;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.36 | 11-Jan-1999 |
thorpej | Add missing XS_SELTIMEOUT case in user command completion path, returning
SCCMD_TIMEOUT. Based on PR #6720, Andreas Wrede.
|
| 1.35 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.34 | 10-Oct-1998 |
thorpej | Garbage-collect the SCIOCREPROBE and OSCIOCREPROBE ioctls. This is a
bus-oriented command. (How much sense does it make to open a device
to rescan the bus?!)
|
| 1.33 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.32 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.31 | 05-Jul-1998 |
jonathan | * defopt COMPAT_{09,10,11,12,13} and COMPAT_NOMID.
TODO: revisit interaction between native compat and emul compat usage.
|
| 1.30 | 25-Jun-1998 |
thorpej | defopt COMPAT_FREEBSD
|
| 1.29 | 15-Jun-1998 |
bouyer | Treat OSCIOCIDENTIFY just as SCIOCIDENTIFY regarding permissions, don't
require the file descriptor to be open read/write. Fix PR kern/5592
from John F. Woods.
|
| 1.28 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.27 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.26 | 19-Sep-1997 |
enami | Add new ioctl OSCIOCREPROBE and OSCIOCIDENTIFY to support binary
compatibility with NetBSD 1.2.
|
| 1.25 | 27-Aug-1997 |
bouyer | branches: 1.25.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.24 | 26-Apr-1997 |
augustss | Don't require that a device is open for writing to perform a raw
SCSI command that just reads.
|
| 1.23 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.22 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.21 | 13-Sep-1996 |
thorpej | For ioctl commands which may change the device's state, ensure that
the caller has the device open for writing.
|
| 1.20 | 14-Feb-1996 |
christos | branches: 1.20.4;
scsi prototypes
|
| 1.19 | 26-Sep-1995 |
thorpej | Don't declare Debugger(). It's handled in <sys/systm.h>.
#include <sys/systm.h> where necessary, as suggested by Jonathan Stone.
Fixes PR #1511.
|
| 1.18 | 30-Jan-1995 |
mycroft | Fix invocation of scsi_scsi_cmd(), per John Kohl.
|
| 1.17 | 23-Jan-1995 |
mycroft | Remove old, misleading comment.
|
| 1.16 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.15 | 01-Dec-1994 |
mycroft | Set UIO residual count.
|
| 1.14 | 01-Dec-1994 |
mycroft | Always set the device number correctly, even though we don't need it.
|
| 1.13 | 01-Dec-1994 |
mycroft | Redux.
|
| 1.12 | 01-Dec-1994 |
mycroft | Uncomment call to scsi_user_done(), and remove a duplicate biodone() (and
another si_free()).
|
| 1.11 | 01-Dec-1994 |
mycroft | Remove the silly statically allocated array, and simply malloc the data
structures as needed. Also, put the buf in the malloc()ed area. Remove a
duplicate si_free() (though this was harmless before).
|
| 1.10 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.9 | 23-Oct-1994 |
mycroft | Make this more likely to work.
|
| 1.8 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.7 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.6 | 16-Jun-1994 |
mycroft | b_un.b_addr -> b_data
|
| 1.5 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.4 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.3 | 29-Nov-1993 |
mycroft | Use new struct buf fields.
|
| 1.2 | 24-Nov-1993 |
mycroft | Still under construction...
|
| 1.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.20.4.1 | 14-Nov-1996 |
thorpej | Pull up from trunk:
> For ioctl commands which may change the device's state, ensure that
> the caller has the device open for writing.
|
| 1.25.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.25.2.3 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.25.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.25.2.1 | 27-Aug-1997 |
thorpej | file scsipi_ioctl.c was added on branch marc-pcmcia on 1997-08-27 23:33:28 +0000
|
| 1.37.14.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.37.14.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.37.14.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.37.14.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.37.14.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.37.8.1 | 21-Dec-1999 |
wrstuden | Initial commit of recent changes to make DEV_BSIZE go away.
Runs on i386, needs work on other arch's. Main kernel routines should be
fine, but a number of the stand programs need help.
cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512
byte block devices. vnd, raidframe, and lfs need work.
Non 2**n block support is automatic for LKM's and conditional for kernels
on "options NON_PO2_BLOCKS".
|
| 1.37.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.37.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.38.2.3 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.38.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.38.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.40.2.1 | 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).
|
| 1.43.10.1 | 03-Nov-2005 |
riz | Pull up following revision(s) (requested by bouyer in ticket #5929):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.54
Don't claim there is a data in or out phase if the datalen is 0 (userland
shouldn't claim it either, but a buggy software shouldn't be able to crash
the kernel anyway). Should fix port-sparc64/31925 by Johan A.van Zanten
(which should really be kern/31925).
Analysed and patch tested by Martin Husemann.
|
| 1.46.4.2 | 03-Nov-2005 |
riz | Pull up following revision(s) (requested by bouyer in ticket #5961):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.54
Don't claim there is a data in or out phase if the datalen is 0 (userland
shouldn't claim it either, but a buggy software shouldn't be able to crash
the kernel anyway). Should fix port-sparc64/31925 by Johan A.van Zanten
(which should really be kern/31925).
Analysed and patch tested by Martin Husemann.
|
| 1.46.4.1 | 11-Sep-2004 |
he | branches: 1.46.4.1.2; 1.46.4.1.4;
Pull up revision 1.48 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.46.4.1.4.1 | 03-Nov-2005 |
riz | Pull up following revision(s) (requested by bouyer in ticket #5961):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.54
Don't claim there is a data in or out phase if the datalen is 0 (userland
shouldn't claim it either, but a buggy software shouldn't be able to crash
the kernel anyway). Should fix port-sparc64/31925 by Johan A.van Zanten
(which should really be kern/31925).
Analysed and patch tested by Martin Husemann.
|
| 1.46.4.1.2.1 | 03-Nov-2005 |
riz | Pull up following revision(s) (requested by bouyer in ticket #5961):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.54
Don't claim there is a data in or out phase if the datalen is 0 (userland
shouldn't claim it either, but a buggy software shouldn't be able to crash
the kernel anyway). Should fix port-sparc64/31925 by Johan A.van Zanten
(which should really be kern/31925).
Analysed and patch tested by Martin Husemann.
|
| 1.46.2.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.46.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.46.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.46.2.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.46.2.1 | 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.
|
| 1.50.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.52.8.1 | 02-Nov-2005 |
yamt | sync with head.
|
| 1.52.6.3 | 21-Jan-2008 |
yamt | sync with head
|
| 1.52.6.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.52.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.52.4.1 | 03-Nov-2005 |
tron | Pull up following revision(s) (requested by bouyer in ticket #928):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.54
Don't claim there is a data in or out phase if the datalen is 0 (userland
shouldn't claim it either, but a buggy software shouldn't be able to crash
the kernel anyway). Should fix port-sparc64/31925 by Johan A.van Zanten
(which should really be kern/31925).
Analysed and patch tested by Martin Husemann.
|
| 1.57.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.57.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.57.2.1 | 05-Feb-2006 |
yamt | adapt scsipi.
|
| 1.58.6.3 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.58.6.2 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.58.6.1 | 28-Mar-2006 |
tron | Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
|
| 1.58.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.58.2.2 | 24-May-2006 |
yamt | sync with head.
|
| 1.58.2.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.61.14.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.62.10.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.62.2.6 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.62.2.5 | 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.62.2.4 | 19-Aug-2007 |
ad | - Back out the biodone() changes.
- Eliminate B_ERROR (from HEAD).
|
| 1.62.2.3 | 17-Jun-2007 |
ad | - Increase the number of thread priorities from 128 to 256. How the space
is set up is to be revisited.
- Implement soft interrupts as kernel threads. A generic implementation
is provided, with hooks for fast-path MD code that can run the interrupt
threads over the top of other threads executing in the kernel.
- Split vnode::v_flag into three fields, depending on how the flag is
locked (by the interlock, by the vnode lock, by the file system).
- Miscellaneous locking fixes and improvements.
|
| 1.62.2.2 | 13-May-2007 |
ad | - Pass the error number and residual count to biodone(), and let it handle
setting error indicators. Prepare to eliminate B_ERROR.
- Add a flag argument to brelse() to be set into the buf's flags, instead
of doing it directly. Typically used to set B_INVAL.
- Add a "struct cpu_info *" argument to kthread_create(), to be used to
create bound threads. Change "bool mpsafe" to "int flags".
- Allow exit of LWPs in the IDL state when (l != curlwp).
- More locking fixes & conversion to the new API.
|
| 1.62.2.1 | 13-Mar-2007 |
ad | Pull in the initial set of changes for the vmlocking branch.
|
| 1.63.22.2 | 29-Jul-2007 |
ad | It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.63.22.1 | 29-Jul-2007 |
ad | file scsipi_ioctl.c was added on branch matt-mips64 on 2007-07-29 12:50:24 +0000
|
| 1.63.18.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.63.14.1 | 04-Dec-2007 |
ad | Pull the vmlocking changes into a new branch.
|
| 1.63.12.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.63.6.1 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.64.10.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.64.10.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.64.8.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.64.6.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.64.6.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.65.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.65.4.1 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.65.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.66.34.1 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.66.32.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.66.28.1 | 23-May-2012 |
yamt | sync with head.
|
| 1.67.22.1 | 31-May-2019 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1698):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.72 (via patch)
use correct size when copying outgoing sense data.
|
| 1.67.18.1 | 31-May-2019 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1698):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.72 (via patch)
use correct size when copying outgoing sense data.
|
| 1.67.16.2 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.67.16.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.67.14.1 | 31-May-2019 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1698):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.72 (via patch)
use correct size when copying outgoing sense data.
|
| 1.67.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.68.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.69.16.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.69.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.69.14.1 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.69.8.1 | 31-May-2019 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1278):
sys/dev/scsipi/scsipi_ioctl.c: revision 1.72 (via patch)
use correct size when copying outgoing sense data.
|
| 1.37 | 27-Jun-2025 |
andvar | Grammar and spelling fixes, mainly in comments. A few in documentation,
logging, test description, and SCSI ASC/ASCQ assignment descriptions.
|
| 1.36 | 28-Dec-2023 |
andvar | branches: 1.36.2;
s/informtion/information/ in comments.
|
| 1.35 | 05-Jun-2021 |
pgoyette | As with usbverbose and pciverbose, these modules are not safe to be
auto-unloaded. Disable for now.
All of these need to be updated with an appropriate refcount mechanism
to ensure that the code and/or tables aren't unloaded while they are
being used.
|
| 1.34 | 16-Sep-2018 |
mrg | branches: 1.34.16; 1.34.20;
don't print a blank line that ends up in logs and console output.
|
| 1.33 | 17-Sep-2016 |
kardel | branches: 1.33.8; 1.33.14; 1.33.16;
ASC/ASCQ table updated from http://www.t10.org/lists/asc-num.txt 2016-07-30
|
| 1.32 | 25-Jul-2010 |
pgoyette | branches: 1.32.18; 1.32.36; 1.32.40;
Move setting of scsi_verbose_loaded flag into the module's init routine.
This ensures that the flag is set even if the module was manually loaded
by the user rather than just auto-loaded.
|
| 1.31 | 07-Jun-2010 |
pgoyette | Update scsiverbose module to use module_autoload() rather than module_load().
Load the module right before each attempt to use its features, and let the
module subsystem handle unloading.
|
| 1.30 | 30-May-2010 |
pgoyette | Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.29 | 28-Apr-2008 |
martin | branches: 1.29.20; 1.29.22;
Remove clause 3 and 4 from TNF licenses
|
| 1.28 | 11-Dec-2005 |
christos | branches: 1.28.70; 1.28.72; 1.28.74;
merge ktrace-lwp.
|
| 1.27 | 29-May-2005 |
christos | - Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.26 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.25 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.24 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.23 | 21-Aug-2004 |
thorpej | branches: 1.23.4; 1.23.6;
Use ANSI function decls and make use of static.
|
| 1.22 | 23-Apr-2004 |
itojun | use bounded string ops
|
| 1.21 | 22-Apr-2004 |
itojun | sprintf -> snprintf
|
| 1.20 | 16-Mar-2004 |
bouyer | cbd -> cdb
Command Block Descriptor -> Command Descriptor Block
Pointed out by Allen Briggs.
|
| 1.19 | 15-Mar-2004 |
bouyer | Extract the code printing the CBD from scsipi_print_sense(), so that it's
usable in other context.
Use the new scsipi_print_cbd() to dump the command in case of timeout
in siop/esiop.
|
| 1.18 | 03-Dec-2003 |
bjh21 | Synchronise with <ftp://ftp.t10.org/t10/drafts/spc3/asc-num.txt>:
SCSI ASC/ASCQ Assignments
Numeric Sorted Listing
as of 9/12/03
|
| 1.17 | 22-Oct-2003 |
mycroft | Add a code that appeared in a PR recently.
|
| 1.16 | 28-May-2002 |
drochner | branches: 1.16.6;
add translations for two ASC/ASCQs which occur before
the DVD player's region code is dealt with
|
| 1.15 | 15-Nov-2001 |
lukem | branches: 1.15.8;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.14 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.13 | 20-Aug-2001 |
ad | Add descriptions for Medium Not Present ASCQs.
|
| 1.12 | 25-Apr-2001 |
bouyer | branches: 1.12.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.11 | 30-Mar-2000 |
augustss | branches: 1.11.6;
Get rid of register declarations.
|
| 1.10 | 24-Feb-1999 |
jwise | branches: 1.10.8;
Fix typo.
Fixes last part of PR bin/6883 from Soren S. Jorvang <soren@t.dk>.
|
| 1.9 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.8 | 15-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.7 | 30-Jul-1998 |
mjacob | opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.6 | 11-Jul-1998 |
mjacob | Clarify that Sense Key 1 is a "RECOVERED ERROR"- not just an error.
|
| 1.5 | 10-Jul-1998 |
mjacob | hmm, char sign extension bit me
|
| 1.4 | 01-Jul-1998 |
mjacob | Align output for Tape specific bits and use new Sense Key defines (for clarity)
|
| 1.3 | 24-Jun-1998 |
mjacob | When printing an error, print more than just the opcode. Print the
entire CDB, whose length is actually deterministic. This makes a *big*
difference when actually trying to bind errors to commands.
|
| 1.2 | 15-Apr-1998 |
mjacob | a tab crept in
|
| 1.1 | 28-Mar-1998 |
christos | Split out the verbose printing of scsi messages, so that user programs
can use the printing routines.
|
| 1.10.8.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.10.8.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.11.6.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.11.6.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.11.6.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.11.6.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.11.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.12.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.12.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.12.2.1 | 25-Aug-2001 |
thorpej | Merge Aug 24 -current into the kqueue branch.
|
| 1.15.8.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.16.6.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.16.6.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.16.6.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.16.6.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.16.6.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.16.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.23.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.23.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.28.74.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.28.74.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.28.72.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.28.70.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.29.22.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.29.22.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.29.20.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.32.40.1 | 04-Nov-2016 |
pgoyette | Sync with HEAD
|
| 1.32.36.1 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.32.18.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.33.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.33.14.1 | 30-Sep-2018 |
pgoyette | Ssync with HEAD
|
| 1.33.8.1 | 23-Sep-2018 |
martin | Pull up following revision(s) (requested by mrg in ticket #1023):
sys/dev/scsipi/scsipi_verbose.c: revision 1.34
don't print a blank line that ends up in logs and console output.
|
| 1.34.20.1 | 06-Jun-2021 |
cjep | sync with head
|
| 1.34.16.1 | 17-Jun-2021 |
thorpej | Sync w/ HEAD.
|
| 1.36.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.45 | 28-Mar-2019 |
kardel | Add reading of supported opcodes and their timeouts
at attachment time. Though this information is optional,
it allows to override our fixed timeouts with device
provided timeouts. These timeouts will override the
hardcoded values if the device provided timeouts
exceed the hardcoded values and are less than a day.
Using the device provided timeouts avoids premature
device resets and unreliable operation due to
inadequate timeouts.
Due to the limited implementations of USB
umass devices this feature is disabled for all
umass attached devices.
|
| 1.44 | 11-Apr-2017 |
christos | branches: 1.44.12;
Put back module_autoload code; the module_autoload code now checks that
root is mounted.
|
| 1.43 | 10-Apr-2017 |
jdolecek | just do not autoload scsiverbose module, it causes deadlock if it happens
while root fs is being mounted
adresses second part of PR kern/52147 by Michael van Elst, thank you
|
| 1.42 | 20-Nov-2016 |
mlelstv | branches: 1.42.2;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.41 | 02-May-2016 |
christos | branches: 1.41.2;
move scsipi_strvis -> libkern:strnvisx()
change the prototype to match userland
fix sizes of strings passed to it
|
| 1.40 | 21-Aug-2010 |
pgoyette | branches: 1.40.18; 1.40.36;
Update the various xxx_verbose modules to conform to the module subsystem's
new locking protocol.
|
| 1.39 | 25-Jul-2010 |
pgoyette | Move setting of scsi_verbose_loaded flag into the module's init routine.
This ensures that the flag is set even if the module was manually loaded
by the user rather than just auto-loaded.
|
| 1.38 | 07-Jun-2010 |
pgoyette | Update scsiverbose module to use module_autoload() rather than module_load().
Load the module right before each attempt to use its features, and let the
module subsystem handle unloading.
|
| 1.37 | 30-May-2010 |
pgoyette | Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.36 | 28-Apr-2008 |
martin | branches: 1.36.20; 1.36.22;
Remove clause 3 and 4 from TNF licenses
|
| 1.35 | 09-Jul-2007 |
ad | branches: 1.35.28; 1.35.30; 1.35.32;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.34 | 11-Dec-2005 |
christos | branches: 1.34.30; 1.34.32;
merge ktrace-lwp.
|
| 1.33 | 30-May-2005 |
christos | branches: 1.33.2;
- remove bogus casts
- add more const
|
| 1.32 | 29-May-2005 |
christos | - Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.31 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.30 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.29 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.28 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.27 | 18-Sep-2004 |
mycroft | branches: 1.27.4; 1.27.6;
Minor rearrangement. Whitespace and #include cleanup.
|
| 1.26 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.25 | 17-Sep-2004 |
mycroft | Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.24 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.23 | 21-Aug-2004 |
thorpej | Oops, missed one in last.
|
| 1.22 | 21-Aug-2004 |
thorpej | Const poison scsipi_dtype().
|
| 1.21 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.20 | 18-Jan-2003 |
martin | branches: 1.20.2; 1.20.4;
Merge botch.
|
| 1.19 | 09-Nov-2002 |
thorpej | Fix signed/unsigned comparison warnings.
|
| 1.18 | 04-Oct-2002 |
soren | Slightly simplicy the SCSI device attachment message:
- Use the plain words 'disk' and 'tape' instead of 'direct' and 'sequential'.
- Media status will be printed in the frontend, so don't bother with it.
- Don't bother printing the SCSI version, which is fixed these days, or the
numberic device type.
|
| 1.17 | 12-Jan-2002 |
tsutsui | Call malloc(9) with M_ZERO flag instead of memset() after malloc().
|
| 1.16 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.15 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.14 | 18-Jul-2001 |
thorpej | bcmp -> memcmp
|
| 1.13 | 25-Apr-2001 |
bouyer | branches: 1.13.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.12 | 31-Dec-2000 |
ad | branches: 1.12.2;
PR 6402: some SCSI devices use \377 as blank, so adjust scsipi_strvis().
|
| 1.11 | 28-Jun-2000 |
mrg | <vm/vm.h> -> <uvm/uvm_extern.h>
|
| 1.10 | 09-Jun-2000 |
enami | Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.9 | 15-May-2000 |
dante | branches: 1.9.2;
Add few peripheral device type
Remove "???" from T_IT8_1/2. They actually are pre-press devices for graphic arts as described by ASC IT8
Zeros and blanks scsipi_inquiry_data from byte 58 to byte 74 if additional_length is less than 58
|
| 1.8 | 17-Nov-1998 |
bouyer | branches: 1.8.10;
Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.7 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.6 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.5 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.4 | 03-Oct-1997 |
thorpej | Clean up and comment the device types, and add the "enclosure services",
"storage array", and "IT8" types.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipiconf.c was initially added on branch bouyer-scsipi.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipiconf.c was added on branch marc-pcmcia on 1997-08-27 23:33:29 +0000
|
| 1.8.10.4 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.8.10.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.8.10.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.8.10.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.9.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.12.2.10 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.12.2.9 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.8 | 12-Jul-2002 |
nathanw | No longer need to pull in lwp.h; proc.h pulls it in for us.
|
| 1.12.2.7 | 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).
|
| 1.12.2.6 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.12.2.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.12.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.12.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.12.2.1 | 05-Mar-2001 |
nathanw | Initial commit of scheduler activations and lightweight process support.
|
| 1.13.2.3 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.13.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.13.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.20.4.1 | 11-Sep-2004 |
he | Pull up revision 1.24 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.20.2.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.20.2.4 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.20.2.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.20.2.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.20.2.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.27.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.27.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.33.2.1 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.34.32.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.34.30.1 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.35.32.3 | 09-Oct-2010 |
yamt | sync with head
|
| 1.35.32.2 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.35.32.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.35.30.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.35.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.36.22.2 | 05-Mar-2011 |
rmind | sync with head
|
| 1.36.22.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.36.20.2 | 22-Oct-2010 |
uebayasi | Sync with HEAD (-D20101022).
|
| 1.36.20.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.40.36.2 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.40.36.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.40.18.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.41.2.2 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.41.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.42.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.44.12.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.133 | 29-Oct-2024 |
nat | Revert to previous - scsipi_done_once is not necessary.
A follow up commit will contain a simpler change (from riastradh@) to
ncr5380sbc.
|
| 1.132 | 28-Oct-2024 |
nat | Introduce scsipi_done_once.
This allows for transfers to be sucessfully aborted on the ncr5380sbc(4).
This may be usefull in future for other scsi controllers.
Callers of scsipi_done are not affected by this change.
Part of kern/58452.
As posted to tech-kern:
https://mail-index.netbsd.org/tech-kern/2024/08/02/msg029652.html
Ok thorpej@.
|
| 1.131 | 22-Jun-2024 |
palle | branches: 1.131.2;
Add quirk for devices that does not handle READ_DISCINFO
|
| 1.130 | 28-Mar-2019 |
kardel | Add reading of supported opcodes and their timeouts
at attachment time. Though this information is optional,
it allows to override our fixed timeouts with device
provided timeouts. These timeouts will override the
hardcoded values if the device provided timeouts
exceed the hardcoded values and are less than a day.
Using the device provided timeouts avoids premature
device resets and unreliable operation due to
inadequate timeouts.
Due to the limited implementations of USB
umass devices this feature is disabled for all
umass attached devices.
|
| 1.129 | 07-Oct-2018 |
christos | Handle the SATA to USB external enclosure sold by "Sabrent" and
made by JMicro (vendor=0x152d product=0x0578). This bridge does
not understand FUA, so add a quirk for it.
|
| 1.128 | 01-Sep-2018 |
mlelstv | Wait in detach if the discovery thread is still running. Avoids crashes
when a device is attached/detached rapidly.
|
| 1.127 | 04-Jul-2018 |
kamil | Avoid undefined behavior in scsipiconf.h in _4ltol() and _4btol()
Do not shift (through integer promotion) a signed value in an operation
than can change the bit of signedness.
sys/dev/scsipi/scsipiconf.h:808:17, left shift of 255 by 24 places cannot be represented in type 'int'
Detected with Kernel Undefined Behavior Sanitizer.
Reported by <Harry Pantazis>
|
| 1.126 | 29-Nov-2016 |
mlelstv | branches: 1.126.14; 1.126.16;
reference count adapter mutex possibly shared by multiple channels.
fix error in atapibusdetach, when a child device cannot be detached,
keep atapibus instance alive.
|
| 1.125 | 25-Nov-2016 |
christos | Move the #endif down to include other kernel only routines.
|
| 1.124 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.123 | 02-May-2016 |
christos | branches: 1.123.2;
move scsipi_strvis -> libkern:strnvisx()
change the prototype to match userland
fix sizes of strings passed to it
|
| 1.122 | 13-Jul-2014 |
dholland | branches: 1.122.4;
"peripheral", not "peripherial".
|
| 1.121 | 20-Apr-2012 |
bouyer | branches: 1.121.2; 1.121.12;
Add a bustype_async_event_xfer_mode() callback to scsipi_bustype (which can
be NULL), so that transport-specific details of transfer mode setting/printing
can be handled more easily.
Move scsipi_async_event_xfer_mode() and scsipi_print_xfer_mode() to
scsi_base.c and split in parallel scsi and FC/SAS parts.
size of struct scsipi_bustype has changed, welcome to 6.99.5
|
| 1.120 | 19-Apr-2012 |
bouyer | Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.119 | 06-Apr-2012 |
christos | Add a quirk for the Apple iPod whose mode sense commands fails with not ready.
Seems to work just fine if we send a start command first...
|
| 1.118 | 07-Jun-2010 |
pgoyette | branches: 1.118.8; 1.118.12; 1.118.14;
Update scsiverbose module to use module_autoload() rather than module_load().
Load the module right before each attempt to use its features, and let the
module subsystem handle unloading.
|
| 1.117 | 30-May-2010 |
pgoyette | Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.116 | 21-Oct-2009 |
rmind | branches: 1.116.2; 1.116.4;
Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.115 | 13-May-2009 |
christos | sprinkle #ifdef _KERNEL to make scsictl compile.
|
| 1.114 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.113 | 08-Sep-2008 |
gmcgarry | branches: 1.113.8;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.
|
| 1.112 | 11-May-2008 |
mlelstv | branches: 1.112.4;
Reintroduce the NODOORLOCK quirk. Helps with PR kern/23875.
|
| 1.111 | 28-Apr-2008 |
martin | branches: 1.111.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.110 | 09-Jul-2007 |
ad | branches: 1.110.28; 1.110.30; 1.110.32;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.109 | 04-Mar-2007 |
christos | branches: 1.109.2; 1.109.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.108 | 26-Nov-2006 |
itohy | branches: 1.108.4;
If the block size reported by Read Capacity looks valid, just use it.
Use Request Sense only if Read Capacity succeeded and did not return
valid block size.
Discussed on tech-kern.
Fix the easier part of NetBSD PR kern/26537.
(The harder part is the device hangs on large (>= 8KB) transfer.
Possibly umass BBB problem?)
Remove scsipi_size() and scsipi_validate_secsize() from scsipi_base.c
and add their functions to sd.c since they are used only by sd.c.
Use SCSI term `block' instead of `sector' where applicable.
|
| 1.107 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.106 | 30-Oct-2006 |
skrll | Remove the boot-from-disk hppa hack now that a real fix has been found.
|
| 1.105 | 20-Oct-2006 |
scw | Validate the sector size returned by READ CAPACITY. If it looks bogus
print a warning and fallback to a suitable default.
Fixes a problem on hp700 reported by skrll@
|
| 1.104 | 09-Oct-2006 |
scw | Some removable umass(4) devices don't respond to mode sense page 6, or
simply return zero for logical block size. In either case, use the sector
length reported by READ_CAPACITY instead of defaulting to 512 bytes.
This partially addresses the problems reported in PR port-i386/34707 and
PR kern/34737. Namely the incorrectly reported drive geometry and the
'hanging' issue.
However, since the device in question reports 2048-byte physical sectors
it will remain unusable until DEV_BSIZE is banished.
|
| 1.103 | 16-Feb-2006 |
perry | branches: 1.103.14; 1.103.16;
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@.
|
| 1.102 | 24-Dec-2005 |
perry | branches: 1.102.2; 1.102.4; 1.102.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.
|
| 1.101 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.100 | 30-May-2005 |
christos | branches: 1.100.2;
- remove bogus casts
- add more const
|
| 1.99 | 29-May-2005 |
christos | - Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.98 | 18-May-2005 |
chs | as a stop-gap measure to allow boot-from-disk to work on hppa,
add some padding to the end of struct scsipi_xfer on that platform.
|
| 1.97 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.96 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.95 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.94 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.93 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.92 | 04-Dec-2004 |
thorpej | branches: 1.92.4; 1.92.6;
Use READ/WRITE (16) if the LBA is larger than 32 bits. Another bit
of kern/28514.
|
| 1.91 | 03-Dec-2004 |
thorpej | And thus spake SBC-3:
If the number of logical blocks exceeds the maximum value that is able to be
specified in the RETURNED LOGICAL BLOCK ADDRESS field, the device server
shall set the RETURNED LOGICAL BLOCK ADDRESS field to FFFFFFFFh. The
application client should then issue a READ CAPACITY (16) command (see 5.11)
to retrieve the READ CAPACITY (16) parameter data.
Implement this in scsipi_size().
First issue in kern/28514.
|
| 1.90 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.89 | 17-Sep-2004 |
mycroft | Change the way bustype_cmd is used. Rather than having it be responsible for
calling scsipi_make_xs() and scsipi_execute_xs(), instead push these into
scsipi_command. Make bustype_cmd and PHOLD/PRELE be called from
scsipi_execute_xs(). This allows us to create a xfer structure -- possibly on
the stack -- and call scsipi_execute_xs() directly.
|
| 1.88 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.87 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.86 | 21-Aug-2004 |
thorpej | Const poison scsipi_dtype().
|
| 1.85 | 21-Aug-2004 |
thorpej | - De-__P.
- Use ANSI function decls.
|
| 1.84 | 16-Mar-2004 |
bouyer | branches: 1.84.2;
cbd -> cdb
Command Block Descriptor -> Command Descriptor Block
Pointed out by Allen Briggs.
|
| 1.83 | 15-Mar-2004 |
bouyer | Extract the code printing the CBD from scsipi_print_sense(), so that it's
usable in other context.
Use the new scsipi_print_cbd() to dump the command in case of timeout
in siop/esiop.
|
| 1.82 | 15-Mar-2004 |
bouyer | Indent.
|
| 1.81 | 22-Feb-2004 |
mycroft | Remove PQUIRK_CDROM. It is definitely not correct -- witness that no other
OS needs such a hack, and the same drive works fine on Suns -- and is much
more likely to be a bug in the host adapter driver (which is corroborated by
the PQUIRK_NOLUNS).
|
| 1.80 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.79 | 17-Sep-2003 |
mycroft | Remove PQUIRK_BYTE5_ZERO.
|
| 1.78 | 09-Sep-2003 |
mycroft | Exorcise PQUIRK_NODOORLOCK.
|
| 1.77 | 08-Sep-2003 |
mycroft | Do a START UNIT only if the TEST UNIT READY reports that the device is not
ready. This avoids gratuitously starting the motor on floppy and CD-ROM
drives, and eliminates the need for the audio playing test in cdopen().
Therefore, also remove PQUIRK_NOSTARTUNIT.
|
| 1.76 | 29-Jun-2003 |
fvdl | branches: 1.76.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.75 | 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
|
| 1.74 | 03-May-2003 |
wiz | DMA, not dma nor Dma.
|
| 1.73 | 19-Apr-2003 |
fvdl | Add a PCAP_NODT quirk. I have two older Quantum Atlas drives that work
fine at U2W, but barf at U160, with different controllers. Unfortunately
sometimes just being initialized at U160 does the trick, so setting them
to lower speed in the BIOS may also be needed.
|
| 1.72 | 03-Apr-2003 |
fvdl | Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.71 | 19-Oct-2002 |
mjacob | Fixes kern/18709 (correcting some typos in comments)
|
| 1.70 | 19-Sep-2002 |
jmc | Force the initial probes to happen within the newly forked off kthread.
This eliminates problems where the underlying interrupt handler isn't the
specific layer calling scsipi_complete() for a given scsi transaction.
This avoids deadlocks where the kthread that called the autoconf routines
to configure a scsibus shouldn't be the one put to sleep waiting on a
scsipi_complete (only the scsibus's kthread should be doing that).
To avoid jitter this will force the scsibus's to probe in the order they
run through autoconf (so machines with multiple bus's don't move sd* devices
around on every reboot).
|
| 1.69 | 16-May-2002 |
thorpej | Don't use a 2-dimensional array for the channel's periph table. Instead,
hash the target and lun together and use a hash table. This will be
necessary in order to support very large (64-bit) LUN number spaces.
|
| 1.68 | 15-May-2002 |
bouyer | branches: 1.68.2;
Change PQUIRK_CAP_SYNC to set PERIPH_CAP_SYNC aslo for devices that
improperly claim to only support DT clocking.
Add a PQUIRK_CAP_WIDE16 quirk.
Add a quirk for IBM DXHS36D SCSI disk (claims to only support DT).
|
| 1.67 | 23-Apr-2002 |
bouyer | Report SPC-2 features (DT clocking, quick arbitration and selection,
information unit transfers) though periph_cap.
|
| 1.66 | 01-Apr-2002 |
bouyer | Add a chan_name to struct scsipi_channel, holding the channel's name.
Set this to dv_xname for scsibus and atapibus.
Set the name of the kernel thread to chan_name instead of controller's
name:channel number (so that we can use this name for controller-specific
threads).
|
| 1.65 | 02-Dec-2001 |
bouyer | Clean up attach of wd/atapibus:
kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead
of ata_atapi_attach as attach arch. Create a ata_device, compatible with
scsipi_channel, to attach wd.
|
| 1.64 | 19-Nov-2001 |
tsutsui | Add a new quirk value PQUIRK_CAP_SYNC for SCSI1 devices which support
sync transfers, and set PERIPH_CAP_SYNC for them in scsi_probe_device().
As per discussion on tech-kern.
|
| 1.63 | 18-Nov-2001 |
tsutsui | Fix formatting of scsi_quirk_patterns[] and wrap long lines.
|
| 1.62 | 14-Oct-2001 |
bouyer | Call ADAPTER_REQ_GROW_RESOURCES from the completion thread, if possible.
This allows HBA drivers to call bus_dmamem_map() safely.
|
| 1.61 | 14-Oct-2001 |
bouyer | Split channel flags in chan_flags used for communications between
scsipi and HBA, and chan_tflags used for communications between scsipi
and its kernel thread. No functionnal change.
|
| 1.60 | 27-Sep-2001 |
mjacob | Restore previous definitions.
|
| 1.59 | 21-Sep-2001 |
fvdl | Unfortunately, the previous change seems to make most (all?) configurations
using the ahc driver hang when probing scsi devices. The problem may be in
the ahc driver and not in this change, but back it out until this has been
fixed anyway.
|
| 1.58 | 18-Sep-2001 |
mjacob | Rather than run a periph's queue from scsipi_periph_timed_thaw which is
called via a callout, kick the completion thread to run it for us (uses
a new flag, SCSIPI_CHAN_KICK).
If we've received BUSY status and we haven't started the completion
thread yet, don't freeze do a callout to scsipi_periph_timed_thaw which
then will try and kick the completion thread- instead treat the command
as if it were a polled command and just call delay for 1 second.
If DIAGNOSTIC is defined, and the periph qfreeze count is less than
zero, panic because some HBA has corrupted the periph structure's
accounting.
|
| 1.57 | 02-Sep-2001 |
tsutsui | branches: 1.57.2;
Use be16toh(), be32toh(), le16toh() and le32toh() rather than
ntohs(), ntohl() or homegrown bswap() functions.
XXX Does anyone use drives with PQUIRK_LITTLETOC on big endian machines?
|
| 1.56 | 31-Aug-2001 |
augustss | Add a new quirk, PQUIRK_NOBIGMODESENSE, for devices that need big
SCSI ops, but still fails on big mode sense.
Makes M-Sys DiskOnKey work.
|
| 1.55 | 18-Jul-2001 |
bouyer | Adn scsipi_target_detach() and scsipi_thread_call_callback() as discussed
on tech-kern. scsipi_target_detach() accept wildcard target/lun as requested.
|
| 1.54 | 26-Jun-2001 |
bouyer | branches: 1.54.2;
Add a XS_CTL_SILENT_NODEV flag: if the sense info is "not ready, medium
not present" don't print any message but still return ENODEV.
Use this in cd driver to allow open of character raw partition even if
the drive is empty (older drives fails at LOAD_UNLOAD time, newer
ones fail at TEST_UNIT_READY time).
|
| 1.53 | 13-Jun-2001 |
bjh21 | Add explicit support for IDE and SCSI adaptors which don't support interrupts.
On such adaptors, all transfers are done in polling mode.
OK'ed by Manuel on tech-kern.
|
| 1.52 | 21-May-2001 |
mjacob | Remove reference to adapter_q in scsipi structure. Comment about the
usage of channel_q.
|
| 1.51 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.50 | 02-May-2001 |
bouyer | Use same type for first field of struct ata_atapi_attach and struct
scsipi_channel, so umass has better chances of working on big-endian systems.
|
| 1.49 | 30-Apr-2001 |
lukem | remove some lint
|
| 1.48 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.47 | 20-Mar-2001 |
augustss | Extend quirks to 32 bits. Add two new ATAPI quirks, one to
deal with a bad byte5 of the capacity descriptor, one to
handle devices that lack the flex geometry page.
|
| 1.46 | 22-Sep-2000 |
ad | branches: 1.46.2;
Add a new, optional method to scsipi_adapter (scsipi_accesschk()), and use
it when considering whether to attach devices. This is to facilitate
`non-SCSI' RAID controller drivers that want to provide SCSI pass-through
services to the kernel.
|
| 1.45 | 06-Jul-2000 |
mjacob | Reserve top nibble of status for use by the HBA. It's up to the
HBA to set or clear it.
|
| 1.44 | 09-Jun-2000 |
enami | Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.43 | 31-May-2000 |
augustss | Add a method, scsipi_getgeom, to the adapter struct. If this method is
present it will be used to get the drive geometry if the MODE SENSE fails.
|
| 1.42 | 30-May-2000 |
augustss | Add a quirk, SDEV_ONLYBIG, which implies that the device cannot handle
the 6 byte versions of READ, WRITE, and MODE_SENSE.
This greatly simplifies the UFI (USB Floppy) handling.
|
| 1.41 | 29-May-2000 |
bouyer | ADEV_CDROM -> SDEV_CDROM, for consistency.
|
| 1.40 | 02-Apr-2000 |
augustss | branches: 1.40.2;
Change the initial field in struct ata_atapi_attach and struct scsipi_link
slightly to allow scsibus and atapibus to attach to the same device.
Furthermore, only attach a scsibus when the bus type is BUS_SCSI.
|
| 1.39 | 30-Mar-2000 |
augustss | Get rid of register declarations.
|
| 1.38 | 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.
|
| 1.37 | 20-Jan-2000 |
mjacob | Nobody said no to adding a pointer to original scsi inquiry data
to the scsibus attach args. Make sure it's nulled for ATAPI. Also,
for scsiconf.c, modify SENA's quirk entry.
|
| 1.36 | 17-Jan-2000 |
bouyer | Use SCSIPIRETRIES instead of hard-coded value '2' for number of retries in
common routines. Define SCSIPIRETRIES as 4, so that the command will finally
succeed after several consecutive downgrades from UDMA2 to PIO4.
|
| 1.35 | 11-Jan-2000 |
pk | <machine/cpu.h> is not needed here.
|
| 1.34 | 23-Dec-1999 |
leo | Implement the DIOCKLABEL ioctl. Through this ioctl it it possible keep the
disklabel on the last close or to drop it.
|
| 1.33 | 20-Oct-1999 |
enami | Cancel active transfers on aic/wdc detach.
Also makes LS-120 drive works for me again.
|
| 1.32 | 30-Sep-1999 |
thorpej | branches: 1.32.2; 1.32.4; 1.32.6;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.31 | 11-Sep-1999 |
thorpej | Implement a function to kill off all commands pending for a given
scsipi_link.
|
| 1.30 | 08-Apr-1999 |
bouyer | Fix a long-standing bug with the ACAP_* flags:
- the cap field is a u_int8_t, so none of the defined flags would fit in.
Looks like nobody had a drive using 16 bytes commands.
- the ACAP_DRQ_* flags are all wrong. Just remove them and use the definitions
from ata/atareg.h, there's no need to duplicate theses. The effect of this
was that we were always polling for the command phase, even for drives
with interrupt DRQ. This didn't break until the code was changed to support
shared interrupts.
Should fix the lookup problems or 'boot hangs' reported by some users, and
kern/7111.
|
| 1.29 | 19-Jan-1999 |
bouyer | branches: 1.29.2;
Move test for SDEV_NOSTARTUNIT quirk from sd.c to scsipi_start().
Add a SDEV_NOSTARTUNIT quirk entry for BCD-16X 1997-04-25", "", "VER 2.2"
CD-rom (from Michael Santos).
|
| 1.28 | 17-Dec-1998 |
bouyer | Add a new quirk flags, "ADEV_NOSENSE", for devices that don't
handle properly the request sense command. Add <FX320S, , q01> as being
sense-unfriendly in the quirk table.
|
| 1.27 | 08-Dec-1998 |
thorpej | - Remove the "free_list" member from scsipi_xfer. Add "adapter_q" and
"device_q" TAILQ entries. The former is for use by the adapter driver,
as it sees fit. The latter is for the scsipi middle layer to track
pending xfers per device.
- Add a pending_xfers queue to scsipi_link, to track pending xfers per
device.
- Grow scsipi_link's flags to int, and add SDEV_WAITDRAIN, to indicate that
we're waiting for the pending_xfers queue to empty.
|
| 1.26 | 05-Dec-1998 |
mjacob | Add a max_lun property to match with the max_target property. Shorten
max_target to int16.
|
| 1.25 | 19-Nov-1998 |
thorpej | Add support for reference counting and enabling/disabling SCSI and
ATAPI controllers. This will eventually be used for power management
of e.g. PCMCIA SCSI and IDE controller cards.
|
| 1.24 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.23 | 15-Oct-1998 |
thorpej | Some const poisioning.
|
| 1.22 | 15-Oct-1998 |
thorpej | Protect kernel prototypes w/ #ifdef _KERNEL.
|
| 1.21 | 15-Oct-1998 |
thorpej | Forward decl of struct proc.
|
| 1.20 | 12-Oct-1998 |
bouyer | Merge bouyer-ide
|
| 1.19 | 10-Oct-1998 |
thorpej | Define an additional host adapter return code, XS_RESET, indicating
that the bus was reset while the command was pending. The middle
layer may choose to retry the command.
|
| 1.18 | 10-Oct-1998 |
thorpej | Garbage-collect the open_target_lu and close_target_lu entry points from
struct scsipi_adapter; they were not used.
Add a scsipi_ioctl entry point to struct scsipi_adapter. This will be
used to issue ioctl commands to the host adapters.
Inspired by PR #6090, from Matt Jacob.
|
| 1.17 | 08-Oct-1998 |
thorpej | Add a "NOSYNCCACHE" quirk, as suggested by Matt Jacob in PR #6027.
|
| 1.16 | 18-Sep-1998 |
scottr | Define a SCSIPI_XFER_ASYNC macro to test whether a request will be
completed asynchronously. Eliminate SCSI_ASYNCREQ, which is actually
redundant information.
|
| 1.15 | 14-Sep-1998 |
scottr | Fix a problem uncovered when we started to use the pool allocator to manage
scsipi_xfer structures.
When scsipi_execute_xs() calls the driver's scsi_cmd function, it assumes
that it can still dereference a pointer to the scsipi_xfer struct. Since
scsipi_done() has already been called, which in turn has called
scsipi_free_xs(), the struct has already been returned the structure to
the pool! In other words, xs->flags has been compromised, but we are still
testing it.
These changes resolve the problem by doing the following:
- In scsipi_execute_xs(), if the hardware driver's scsi_cmd function
returns SUCCESSFULLY_QUEUED, set a new flag (SCSI_ASYNCREQ) in xs->flags.
Since the request will be handled asynchronously, we will need the
scsipi_xfer struct to be freed in scsipi_done().
If the hardware driver's scsi_cmd function returns COMPLETE, we now
simply return any actual errors, or 0 if none occurred. (Previously,
we may have returned EJUSTRETURN, of which the sole effect was to
avoid freeing the scsipi_xfer struct in our caller.)
- In scsipi_done(), only free the scsipi_xfer struct for async requests.
The contents of the struct will otherwise remain valid until the
function that initiated the transfer frees it.
With this change, responsibility for freeing the struct now lies in two
places, depending on the type of the request:
- For synchronous requests, the routine calling scsipi_execute_xs()
must clean up.
- For asynchronous requests, scsipi_done() cleans up (as it always has).
[Note: this change also corrects a problem with sddump(): scsipi_done()
was attempting to return a static scsipi_xfer struct to the pool! Since
dumps are performed synchronously, we now handle this correctly.]
This solution was provided by Jason Thorpe, after I got him to look at
some related (but insufficient) attempts of my own.
|
| 1.14 | 08-Sep-1998 |
mjacob | Clarify quirk flags somwhat- split SDEV_NOSYNCWIDE into SDEV_NOSYNC and
SDEV_NOWIDE and add SDEV_NOTAG.
|
| 1.13 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.12 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.11 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.10 | 31-Jul-1998 |
thorpej | Use the pool allocator for scsipi_xfer structures.
|
| 1.9 | 15-Jul-1998 |
mjacob | branches: 1.9.2;
1) Part of fix for kern/3835: add in enumerated return values for target
drivers' sense handlers to return. Coincidentally one of them ends up
being ERESTARTSYS.
2) Add a SCSI_URGENT flag to xs structure- this allows host adapters that
do command tagging to do the right thing wrt a tag.
|
| 1.8 | 13-Feb-1998 |
enami | branches: 1.8.2;
Fix or add protection for mutiple inclusion.
|
| 1.7 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.6 | 15-Dec-1997 |
scottr | Fix a typo in a comment.
|
| 1.5 | 20-Nov-1997 |
thorpej | Declare the scsipi_xfer's "flags" member to be volatile, to force it
to be reloaded every time it is checked. This avoids a condition where
it can be cached in a register in such a way that updates to the flags in
an interrupt handler to not be noticed, which in turn causes the process
doing the i/o to sleep forever. Bug report and suggested fix from
Hiroshi HORIMOTO <horimoto@cs-yuugao.cs.sist.ac.jp>, PR $4460.
|
| 1.4 | 18-Oct-1997 |
thorpej | branches: 1.4.2;
Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.3 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.2 | 27-Aug-1997 |
bouyer | branches: 1.2.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.1 | 01-Jul-1997 |
bouyer | branches: 1.1.2;
file scsipiconf.h was initially added on branch bouyer-scsipi.
|
| 1.1.2.3 | 17-Jul-1997 |
bouyer | Update from trunck: quirk entries, __attribute__ ((unused)) and
probe message of scsi disks.
|
| 1.1.2.2 | 01-Jul-1997 |
thorpej | Purely cosmetic formatting changes: some spaces -> tabs, indenting
consistency in prototypes.
|
| 1.1.2.1 | 01-Jul-1997 |
bouyer | New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
New merged scsi/atapi code. Tested on i386 and sparc. Commiting to a branch
for now, so that it can be tested on other ports too.
The config sheme is as follow:
scsibus at aha
sd at scsibus
cd at scsibus
atapibus at wdc
cd at atapibus
cd has bus-specific config and function front-end in scsicd.c and atapicd.c
The call to theses functions from cd.c is conditionned to
NSCSICD and NATAPICD (all defined in cd.h by config).
|
| 1.2.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.2.2.1 | 27-Aug-1997 |
thorpej | file scsipiconf.h was added on branch marc-pcmcia on 1997-08-27 23:33:30 +0000
|
| 1.4.2.1 | 20-Nov-1997 |
thorpej | Pull up from trunk: declare scsipi_xfer's "flags" to be volatile.
|
| 1.8.2.7 | 12-Oct-1998 |
bouyer | Sync with HEAD.
|
| 1.8.2.6 | 20-Sep-1998 |
bouyer | sync with HEAD
|
| 1.8.2.5 | 11-Sep-1998 |
bouyer | Sync with HEAD.
|
| 1.8.2.4 | 21-Aug-1998 |
bouyer | Sync with -current
|
| 1.8.2.3 | 13-Aug-1998 |
bouyer | - sync with HEAD
- better error detection and report for wd drives
- better wddump() support
- wdcwait() now takes a timout argument, so that we can honnor the
timeout argument of scsipi cmds.
|
| 1.8.2.2 | 27-Jul-1998 |
bouyer | Sync with -current
|
| 1.8.2.1 | 04-Jun-1998 |
bouyer | Commit changes to the IDE system in a branch. This allows a better separation
between higth-level and low-level (i.e. registers read/write) and generalize
the queue for all commands. This also add supports for IDE DMA.
|
| 1.9.2.1 | 08-Aug-1998 |
eeh | Revert cdevsw mmap routines to return int.
|
| 1.29.2.3 | 23-Jan-2000 |
he | Pull up revision 1.36 (via patch, requested by bouyer):
Use SCSIPIRETRIES instead of hard-coded value '2' for number of
retries in common routines, and define it as 4 so that ATAPI
command will succeed after several downgrade.
|
| 1.29.2.2 | 16-Jan-2000 |
he | Apply patch (requested by ad):
When probing a SCSI target, do not report an error if the target
indicates that LUNs are not supported.
|
| 1.29.2.1 | 08-Apr-1999 |
bouyer | branches: 1.29.2.1.2;
Pull up atapi_wdc.c rev 1.21 and scsipiconf.h rev 1.30:
Fix botch with atapi capability flags, preventing 16 bytes commands to be ever
sent, and interrupt DRQ capable devices to effectively use interrupts for
the CMD_OUT phase. Didn't break until the code was changed to handle shared
interrupts (and I guess nobody has a device that requires 16 bytes commands
yet). Should fix the hangs reported by some users at first access of an ATAPI
device.
|
| 1.29.2.1.2.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.32.6.1 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.32.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.32.2.15 | 23-Apr-2001 |
ad | If SCSIPI_CHAN_NOSETTLE is set in the channel's flags, then don't bother
with SCSI_DELAY.
|
| 1.32.2.14 | 22-Apr-2001 |
bouyer | Keep track of the number of commands per periph really sent to the adapter.
In scsipi_run_queue(), compare this to periph_openings instead of periph_active
When a drive reject all queued commands with QUEUE FULL (my IBM DDRS34560D
does this on large writes) we end up with all commands in the queue, and
periph_active > periph_openings.
While doing this I found a bug in scsipi_async_event_channel_reset(): calling
scsipi_done() would corrupt accounting.
|
| 1.32.2.13 | 11-Apr-2001 |
mjacob | Make scsipi_channel_init a function returning an int- non-zero means
it failed to initialize the channel (this should be acceptable)- in
which case we complain and don't schedule bus probing for later.
We make the internal memory allocations for the periph and the chan_periphs
array M_NOWAIT- this way we have a hope of booting instead of silently hanging
during boot if we've run out of memory.
|
| 1.32.2.12 | 03-Apr-2001 |
bouyer | When we send a untagged command, block the queue until this command is
complete.
|
| 1.32.2.11 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.32.2.10 | 22-Jan-2001 |
bouyer | Don't forget to remove xfer from the channel's queue in
scsipi_async_event_channel_reset()
thaw the periph when request_sense failed(), and when calling user done,
as we're not going though the usual path.
Implement XS_REQUEUE, for drivers needing to requeue a xfer.
|
| 1.32.2.9 | 15-Jan-2001 |
bouyer | Handle CHECK CONDITION status in mid-layer:
the REQUEST_SENSE command is generated from scsipi_complete() so it can
tsleep() (we use a regular scsipi_command() call for this). Add a new
periph_flag, PERIPH_SENSE, and a new xs_control, XS_CTL_REQSENSE.
When PERIPH_SENSE is set only xfer with XS_CTL_REQSENSE may be sent
(same logic as PERIPH_RECOVERING but with higther priority). xfer with
XS_CTL_REQSENSE need to have XS_CTL_URGENT urgent too.
XS_CTL_USERCMD xfers are now handled in scsipi_complete().
We need to pay special attention to SCSI resets, as we may have:
- an aborted REQUEST_SENSE. In this case we need to requeue the original
command, not the request sense.
- sense pending for a command no longer in the queue but for which a
request sense has not yet been queued. In this case we should not issue
the request sense but requeue the original command instead.
For this:
- the xfer with the CHECK CONDITION status is stored in the scsipi_periph
(periph_xscheck); the CHECK CONDITION is tested in scsipi_done, PERIPH_SENSE
and periph_xscheck is set here.
- XS_CTL_REQSENSE xfers are not allowed to be requeued, and are terminated
with EINTR in case of reset.
- we have a new async event, "ASYNC_EVENT_RESET", which cleanup the
xfer in periph_xscheck.
- appropriate splbio/splx to avoid race condition (especially,
scsipi_request_sense() runs at splbio. Should not be a real problem as it
doesn't happen often.
While I'm there kill req_sense_length from struct scsipi_xfer, it's always set
to 0 and only checked by a few drivers.
|
| 1.32.2.8 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.32.2.7 | 04-Feb-2000 |
thorpej | Make sure the channel's completion thread exits when the channel
is detached, and completely encapsulate the way periph's are stored
by the channel.
|
| 1.32.2.6 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.32.2.5 | 26-Oct-1999 |
thorpej | Deal a little more gracefully with the fact that xfer mode parameters
are for the I_T Nexus, and make all xfer mode updates `async events'.
|
| 1.32.2.4 | 20-Oct-1999 |
thorpej | Sync w/ trunk.
|
| 1.32.2.3 | 20-Oct-1999 |
thorpej | Add channel freeze/thaw, and when a timed thaw occurs on either a
channel or periph, make sure to kick the channel queue.
|
| 1.32.2.2 | 19-Oct-1999 |
thorpej | Add an asynchronous event for the adapter to notify the midlayer that
xfer mode parameters have changed for an I_T Nexus.
|
| 1.32.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.40.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.46.2.11 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.46.2.10 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.46.2.9 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.46.2.8 | 17-Apr-2002 |
nathanw | Catch up to -current.
|
| 1.46.2.7 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.46.2.6 | 22-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.46.2.5 | 08-Oct-2001 |
nathanw | Catch up to -current.
|
| 1.46.2.4 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.46.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.46.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.46.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.54.2.5 | 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
|
| 1.54.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.54.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.54.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.54.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.57.2.2 | 01-Oct-2001 |
fvdl | Catch up with -current.
|
| 1.57.2.1 | 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).
|
| 1.68.2.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.76.2.8 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.76.2.7 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.76.2.6 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.76.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.76.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.76.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.76.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.76.2.1 | 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.
|
| 1.84.2.1 | 11-Sep-2004 |
he | Pull up revision 1.87 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.92.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.92.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.100.2.3 | 03-Sep-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.
|
| 1.102.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.102.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.102.2.1 | 18-Feb-2006 |
yamt | sync with head.
|
| 1.103.16.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.103.16.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.103.14.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.103.14.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.108.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.109.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.109.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.109.2.1 | 09-Apr-2007 |
ad | - Add two new arguments to kthread_create1: pri_t pri, bool mpsafe.
- Fork kthreads off proc0 as new LWPs, not new processes.
|
| 1.110.32.5 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.110.32.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.110.32.3 | 16-May-2009 |
yamt | sync with head
|
| 1.110.32.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.110.32.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.110.30.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.110.28.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.110.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.111.2.2 | 24-Sep-2008 |
wrstuden | Merge in changes between wrstuden-revivesa-base-2 and
wrstuden-revivesa-base-3.
|
| 1.111.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.112.4.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.113.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.116.4.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.116.2.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.118.14.1 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.118.12.1 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.118.8.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.118.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.121.12.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.121.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.121.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.122.4.2 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.122.4.1 | 29-May-2016 |
skrll | Sync with HEAD
|
| 1.123.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.126.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.126.14.3 | 20-Oct-2018 |
pgoyette | Sync with head
|
| 1.126.14.2 | 06-Sep-2018 |
pgoyette | Sync with HEAD
Resolve a couple of conflicts (result of the uimin/uimax changes)
|
| 1.126.14.1 | 28-Jul-2018 |
pgoyette | Sync with HEAD
|
| 1.131.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.347 | 11-Oct-2025 |
mlelstv | Make probe command silent.
|
| 1.346 | 11-Oct-2025 |
mlelstv | For SPC-3 devices try READ_CAPACITY_16 first and fall back to older
READ_CAPACITY_10.
This fixes SCSI drivers like mpt(4) that announce SPC-3 conformity but
don't support READ_CAPACITY_16 in all configurations.
|
| 1.345 | 13-Apr-2025 |
jakllsch | Add physical sector and alignment info to struct disk_geom and the
geometry plist, and handle in partutil.
Bump version for disk_geom addition.
Collect DIOCGSECTORALIGN handling into one place.
|
| 1.344 | 12-Apr-2025 |
mlelstv | Use WRITE_16 for dumps when necessary.
|
| 1.343 | 05-Mar-2025 |
jakllsch | sd(4): print physical sector information
|
| 1.342 | 02-Mar-2025 |
riastradh | sd(4): Break overlong lines.
No functional change intended.
|
| 1.341 | 27-Feb-2025 |
jakllsch | sd(4): Translate discard to UNMAP (10) command
|
| 1.340 | 27-Feb-2025 |
jakllsch | Implement DIOCGSECTORALIGN for sd(4).
Use only READ CAPACITY 16 if peripheral claims conformance to SPC-3.
|
| 1.339 | 27-Feb-2025 |
jakllsch | whitespace fixes
|
| 1.338 | 11-Feb-2025 |
mlelstv | Handle unit startup issues also before "media is loaded".
|
| 1.337 | 28-Sep-2024 |
mlelstv | Don't artificially limit block size to 4096 bytes, use MAXPHYS.
|
| 1.336 | 24-Feb-2024 |
mlelstv | branches: 1.336.2;
Don't try to discover wedges when the unit isn't online.
|
| 1.335 | 28-Aug-2022 |
mlelstv | branches: 1.335.4;
Don't fetch data beyond end of inquiry buffer, which, here, is not
NUL-terminated.
Reduce target buffer to needed size (product name + NUL terminator).
|
| 1.334 | 28-Mar-2022 |
riastradh | sd(4): Use d_cfdriver/devtounit to avoid open/detach races.
|
| 1.333 | 27-Jan-2022 |
jakllsch | use powerof2() in sd_validate_blksize()
|
| 1.332 | 30-May-2021 |
dholland | typo in comment
|
| 1.331 | 10-Feb-2021 |
christos | branches: 1.331.4; 1.331.6;
PR/55986: Ryo Onodera: DK_BUSY must have mask as second argument. Make cd.c
consistent by also using __BIT()
|
| 1.330 | 25-Sep-2020 |
jakllsch | branches: 1.330.2;
spelling fixes in sd.c comments
|
| 1.329 | 13-Apr-2020 |
maxv | constify
|
| 1.328 | 01-Mar-2020 |
riastradh | branches: 1.328.4;
Add a flag to dk_dump for virtual disk devices.
If a disk is backed by a physical medium other than itself, such as
cgd(4), then it passes DK_DUMP_RECURSIVE to disable the recursion
detection for dk_dump.
If, however, a device represents a physical medium on its own, such
as wd(4), then it passes 0 instead.
With this, I can now dump to dk on cgd on dk on wd.
|
| 1.327 | 19-Mar-2019 |
mlelstv | branches: 1.327.4;
Set disk product name as disk type. The information can be queried through
drvctl(4).
|
| 1.326 | 07-Oct-2018 |
christos | Handle the SATA to USB external enclosure sold by "Sabrent" and
made by JMicro (vendor=0x152d product=0x0578). This bridge does
not understand FUA, so add a quirk for it.
|
| 1.325 | 17-Jun-2017 |
mlelstv | branches: 1.325.4; 1.325.6;
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.324 | 10-Apr-2017 |
jdolecek | branches: 1.324.4; 1.324.6;
execute the cache page MODE SENSE with XS_CTL_SILENT; it's pretty normal
for e.g. USB sticks thus showing error is not really useful, and the pretty
printing triggers autoload of scsiverbose module and immediate deadlock when
the DIOCGCACHE call is made by WAPBL during root mount
adresses PR kern/52147 by Michael van Elst
|
| 1.323 | 05-Apr-2017 |
jdolecek | expose disk device FUA/DPO support via DIOCGCACHE, and allow the flags
to be set for I/O; implement support in sd(4) and nvme(4)
discussed on tech-kern
|
| 1.322 | 21-Dec-2016 |
mlelstv | branches: 1.322.2;
really keep request queued after resource shortage
|
| 1.321 | 16-Dec-2016 |
mlelstv | add comment about "missing" dk_start.
|
| 1.320 | 10-Dec-2016 |
mlelstv | Refactored sd and cd to use common disk subroutines.
|
| 1.319 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.318 | 20-Nov-2016 |
pgoyette | Avoid calling bufq_free() from critical code sections.
|
| 1.317 | 24-Aug-2015 |
pooka | branches: 1.317.2;
would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.316 | 26-Apr-2015 |
mlelstv | Use C99-style initializers for struct dkdriver.
|
| 1.315 | 13-Apr-2015 |
riastradh | Convert sys/dev to use <sys/rndsource.h>.
|
| 1.314 | 02-Jan-2015 |
christos | We have three sets of DTYPE_ constants in the kernel:
altq Drop Type
disklabel Disk Type
file Descriptor Type
(not to mention constants that contain the string DTYPE).
Let's make them two, by changing the disklabel one to be DisK TYPE since the
other disklabel constants seem to do that. Not many userland programs use
these constants (and the ones that they do are mostly in ifdefs). They will
be fixed shortly.
|
| 1.313 | 31-Dec-2014 |
christos | make more drivers use disk_ioctl, and add a dev parameter to it so that
we can merge the "easy" disklabel ioctls to it. Ultimately all this will
go do dk_ioctl once all the drivers have been converted.
|
| 1.312 | 31-Dec-2014 |
christos | Centralize wedge ioctls in disk_ioctl.
|
| 1.311 | 31-Dec-2014 |
mlelstv | disk_blocksize and disk_set_info relay the same information
to the disk subsystem.
Make disk_set_info also set blocksize shift values.
Remove every call to disk_blocksize.
Keep disk_blocksize for ABI compatibility, make it also set dg_secsize.
|
| 1.310 | 04-Nov-2014 |
mlelstv | branches: 1.310.2;
support DIOCMWEDGES ioctl.
|
| 1.309 | 05-Sep-2014 |
matt | Don't use new as a variable name.
|
| 1.308 | 10-Aug-2014 |
tls | branches: 1.308.2;
Merge tls-earlyentropy branch into HEAD.
|
| 1.307 | 25-Jul-2014 |
dholland | Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.306 | 25-Jul-2014 |
dholland | Add d_discard to all struct bdevsw instances I could find.
I've set them all to nodiscard. Some of them (wd, dk, vnd, ld,
raidframe, maybe cgd) should be implemented for real.
|
| 1.305 | 16-Mar-2014 |
dholland | branches: 1.305.2;
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.
|
| 1.304 | 25-Oct-2013 |
martin | Mark a diagnostic-only variable
|
| 1.303 | 28-Sep-2013 |
skrll | Remove unnecessary initialisation of error
|
| 1.302 | 26-Sep-2013 |
kiyohara | Reset variable error to 0 before handle each ioctl cmd. Or returns EPASSTHROUGH
when some cmds. (e.g. ODIOCGDINFO)
|
| 1.301 | 13-Jun-2013 |
tls | branches: 1.301.2;
Convert the entropy pool framework from pseudo-callout-driven to
soft interrupt driven operation.
Add a polling mode of operation -- now we can ask hardware random number
generators to top us up just when we need it (bcm2835_rng and amdpm
converted as examples).
Fix a stall noticed with repeated reads from /dev/random while testing.
|
| 1.300 | 29-May-2013 |
christos | phase 1 of disk geometry cleanup:
- centralize the geometry -> plist code so that we don't have
n useless copies of it.
|
| 1.299 | 16-Apr-2013 |
jakllsch | Clamp sectors per unit in the default disklabel to the maximum value,
instead of allowing it to wrap, as is already done in wd(4).
|
| 1.298 | 19-Apr-2012 |
bouyer | branches: 1.298.2;
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.297 | 06-Apr-2012 |
christos | Add a quirk for the Apple iPod whose mode sense commands fails with not ready.
Seems to work just fine if we send a start command first...
|
| 1.296 | 02-Feb-2012 |
tls | branches: 1.296.2;
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.
|
| 1.295 | 25-Nov-2011 |
joerg | branches: 1.295.2;
Replace memset with incorrect size argument with M_ZERO.
|
| 1.294 | 30-Jul-2011 |
jmcneill | branches: 1.294.2;
Add an FSILENT flag and use it to suppress "Medium Not Present" scsipi
spam when trying to access offline drives at boot.
|
| 1.293 | 14-Apr-2010 |
jakllsch | Implement DIOC[GS]STRATEGY for sd(4).
|
| 1.292 | 24-Feb-2010 |
dyoung | branches: 1.292.2;
A pointer typedef entails trading too much flexibility to declare const
and non-const types, and the kernel uses both const and non-const
PMF qualifiers and device suspensors, so change the pmf_qual_t and
device_suspensor_t typedefs from "pointers to const" to non-pointer,
non-const types.
|
| 1.291 | 08-Jan-2010 |
dyoung | branches: 1.291.2;
Expand PMF_FN_* macros.
|
| 1.290 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.289 | 03-Aug-2009 |
jnemeth | Fix previous where the call of sd_set_properties was moved from
the end of sdattach() to the end of sd_get_parms(). The code at
the bottom of sd_get_parms() was only used for optical drives or
in the case where the drive geometry couldn't be retrived for some
reason, to create a fake geometry. The case for setting up the
real geometry was handled above that code.
|
| 1.288 | 21-Jun-2009 |
jakllsch | Move call of sd_set_properties() to end of sd_get_parms(), rather than
sdattach(). This allows DIOCGDISKINFO to do the right thing even when
the media has changed. Note that drvctl -p will only DTRT if disk has
been opened since the most recent media chenged.
|
| 1.287 | 05-Jun-2009 |
haad | Add support for DIOCGDISKINFO to disk like device drivers. Change
partutil.c::getdiskinfo to use it to get disk geometry info.
Use DIOCGWEDGEINFO ioctl to get information about partition size, if disk
driver doesn't support it use old DIOCGDINFO. This patch adds support for
wedge like devices(lvm logical volumes, ZFS zvol partitions) to newfs and
other tools.
No objections on tech-userlevel@.
|
| 1.286 | 20-May-2009 |
dyoung | On second thought, let's call disk_predetach() disk_begindetach().
Verbs are good.
|
| 1.285 | 19-May-2009 |
dyoung | Use disk_predetach().
|
| 1.284 | 19-May-2009 |
dyoung | Delete wdactivate() and sdactivate(). They were no-ops, but as a
side-effect of registering them, config_detach(9) cleared DVF_ACTIVE
before it called wddetach() or sddetach(). Even though sd(4)'s
detachment may have subsequently failed with EBUSY, we could not
begin new disk I/O on sd(4) because the device had been deactivated.
By analogy to sdstrategy(), test device_is_active() in wdstrategy()
and if it is false set b_errno to EIO instead of initiating new
disk I/O.
XXX We should decline more politely to start new I/O, since
XXX !device_is_active() may mean simply that the device is suspended.
XXX I suppose that EIO is safe as long as system suspension is
XXX all-or-nothing.
|
| 1.283 | 17-May-2009 |
dyoung | Reverse the sense of the DETACH_FORCE check so that we return EBUSY in
the correct circumstances.
|
| 1.282 | 16-May-2009 |
dyoung | Do not detach a sd(4) instance if it is open, but return EBUSY,
instead. If the detach is forced, do what sdclose() does on the
last close: flush the cache, delete our reference to the SCSI bus
device, et cetera.
|
| 1.281 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.280 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.279 | 10-Apr-2009 |
dyoung | Use a PMF shutdown hook instead of shutdownhook_establish(). This ought
to stop the crash during shutdown that yamt@ reports in kern/41182.
|
| 1.278 | 07-Apr-2009 |
dyoung | Detach atapibus(4), scsibus(4), cd(4), and sd(4) during shutdown.
Destroy sd->sc_callout in sddetach(). Delete some dead code in
cddetach().
|
| 1.277 | 13-Jan-2009 |
yamt | branches: 1.277.2;
g/c BUFQ_FOO() macros and use bufq_foo() directly.
|
| 1.276 | 11-Jan-2009 |
cegger | make this compile
|
| 1.275 | 16-Jul-2008 |
drochner | branches: 1.275.2; 1.275.4;
split device/softc for sd (tested with a USB stick)
|
| 1.274 | 08-Jun-2008 |
tsutsui | branches: 1.274.2; 1.274.4;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.273 | 28-Apr-2008 |
martin | branches: 1.273.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.272 | 05-Apr-2008 |
cegger | branches: 1.272.2; 1.272.4;
use aprint_*_dev and device_xname
|
| 1.271 | 29-Feb-2008 |
dyoung | Use PMF_FN_ARGS, PMF_FN_PROTO.
|
| 1.270 | 21-Feb-2008 |
drochner | put back the old shutdown handlers to flush the disk cache,
the pmf API can't deal with all the different suspend/resume/reboot
cases well yet, so better keep suspend/resume and reboot/halt/poweroff
clearly seperated
|
| 1.269 | 25-Dec-2007 |
perry | branches: 1.269.2; 1.269.6;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h
|
| 1.268 | 09-Dec-2007 |
jmcneill | branches: 1.268.2;
Merge jmcneill-pm branch.
|
| 1.267 | 08-Oct-2007 |
ad | branches: 1.267.4; 1.267.6; 1.267.8;
Merge disk init changes from the vmlocking branch. These seperate init /
destroy of 'struct disk' from attach / detach.
|
| 1.266 | 01-Oct-2007 |
bouyer | Properly initialize xs_callout in sddump(). We're using a static scsipi_xfer
here which didn't get through scsipi_get_xs().
Pointed out by Greg Oster.
|
| 1.265 | 03-Aug-2007 |
tsutsui | branches: 1.265.2; 1.265.4; 1.265.6; 1.265.8;
Don't allocate data buffer on stack in sd_read_capacity().
If data buffer is allocated on stack at the end of CPU cacheline,
the same cacheline region might be accessed via kernel stack even
after bus_dmamap_sync(9) against the data buffer in lower scsi drivers
so that read data via DMA could be lost.
Fixes occasional readdisklabel(9) failure at boot on ews4800mips
with osiop(4) driver and 128byte cacheline.
XXX: Other functions/drivers might have the similar problem.
|
| 1.264 | 29-Jul-2007 |
ad | It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.263 | 21-Jul-2007 |
ad | Replace some uses of lockmgr().
|
| 1.262 | 09-Jul-2007 |
ad | branches: 1.262.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.261 | 04-Mar-2007 |
christos | branches: 1.261.2; 1.261.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.260 | 03-Mar-2007 |
jnemeth | add support devctl properties
|
| 1.259 | 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.
|
| 1.258 | 30-Nov-2006 |
christos | branches: 1.258.2; 1.258.4; 1.258.6;
Prevent kernel crashes caused by umass devices that return bogus mode sense
data, such as the "Cooler Master". From Rhialto. While there, clean up some
debugging code and make the offset unsigned.
|
| 1.257 | 29-Nov-2006 |
drochner | restore a nice little feature which got lost in some previous commit:
the d_typename field of the default disklabel is initialized with
text from the SCSI inquiry, so that one gets at least a vague idea which
disk he is dealing with
|
| 1.256 | 26-Nov-2006 |
itohy | If the block size reported by Read Capacity looks valid, just use it.
Use Request Sense only if Read Capacity succeeded and did not return
valid block size.
Discussed on tech-kern.
Fix the easier part of NetBSD PR kern/26537.
(The harder part is the device hangs on large (>= 8KB) transfer.
Possibly umass BBB problem?)
Remove scsipi_size() and scsipi_validate_secsize() from scsipi_base.c
and add their functions to sd.c since they are used only by sd.c.
Use SCSI term `block' instead of `sector' where applicable.
|
| 1.255 | 25-Nov-2006 |
scw | - Call disk_blocksize(9) once we know the disk's physical block size.
- bounds_check_with_label() has been fixed, so no need to fake the
raw partition's size in terms of DEV_BSIZE units.
|
| 1.254 | 16-Nov-2006 |
christos | __unused removal on arguments; approved by core.
|
| 1.253 | 20-Oct-2006 |
scw | Validate the sector size returned by READ CAPACITY. If it looks bogus
print a warning and fallback to a suitable default.
Fixes a problem on hp700 reported by skrll@
|
| 1.252 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.251 | 09-Oct-2006 |
scw | Some removable umass(4) devices don't respond to mode sense page 6, or
simply return zero for logical block size. In either case, use the sector
length reported by READ_CAPACITY instead of defaulting to 512 bytes.
This partially addresses the problems reported in PR port-i386/34707 and
PR kern/34737. Namely the incorrectly reported drive geometry and the
'hanging' issue.
However, since the device in question reports 2048-byte physical sectors
it will remain unusable until DEV_BSIZE is banished.
|
| 1.250 | 14-Sep-2006 |
reinoud | branches: 1.250.2;
Fix same panic problem as with cd.c when a scsi error happends on
read/write action. It would panic on a diagnosic assertion. When a scsi
read/write command fails, the whole extent is gone bad so mark all as not
done.
|
| 1.249 | 09-Jul-2006 |
drochner | branches: 1.249.2; 1.249.4;
some fixes for removable media (USB floppies in particular):
-allow to open RAW_PART, even if readdisklabel() doesn't like what it sees
-fix PERIPH_MEDIA_LOADED toggling on every other open attempt
-Don't reset PERIPH_REMOVABLE if prevent/allow fails - this is normal
for floppy drives. The flag is required for media change detection.
Honour XS_CTL_SILENT for the "no door lock" printouts, and set that flag
where scsipi_prevent() is called during open/close.
Tested with a USB floppy and a USB cardreader.
|
| 1.248 | 01-Jun-2006 |
nathanw | branches: 1.248.2;
Change return type of sdgetdisklabel() from void to int; return an error
if readdisklabel() returned an error.
sdopen(): Return an error if sdgetdisklabel() returns an error. This prevents
a crash in spec_open() (and possibly elsewhere) if a sd device is detached in
the middle of sdopen().
|
| 1.247 | 30-Mar-2006 |
thorpej | branches: 1.247.2;
Use device_private().
|
| 1.246 | 28-Mar-2006 |
thorpej | Use device_unit().
|
| 1.245 | 20-Feb-2006 |
thorpej | branches: 1.245.2; 1.245.4; 1.245.6;
Use device_is_active() rather than testing dv_flags for DVF_ACTIVE
directly.
|
| 1.244 | 11-Dec-2005 |
christos | branches: 1.244.2; 1.244.4; 1.244.6;
merge ktrace-lwp.
|
| 1.243 | 09-Dec-2005 |
bouyer | Adapt previous change to recent scsipi. Pointed out by Simon Burge.
|
| 1.242 | 08-Dec-2005 |
bouyer | Some USB devices reports themselves as removable, but have no door and
so don't support the SCSI_PREVENT_ALLOW_MEDIUM_REMOVAL command.
When an "Illegal field in CDB" is reported for this command, mark the
device as non-removable (which is always true for USB keys from the SCSI
point of view), print a message and ignore the error.
For DIOCLOCK, return ENOTTY if the device is not removable instead of
trying a command which will fail.
Fix a problem reported by Hubert Feyrer for some USB umass devices, patch
tested by him.
|
| 1.241 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.240 | 29-May-2005 |
christos | branches: 1.240.2;
- Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.239 | 25-Apr-2005 |
drochner | Revert the part if 1.236 which comprises a functional change.
While I didn't observe a misbehaviour yet, this _must_ be
wrong because SSD_RCODE_VALID is used incorrectly.
|
| 1.238 | 31-Mar-2005 |
yamt | introduce a function to drain bufq and use it where appropriate.
|
| 1.237 | 27-Feb-2005 |
perry | branches: 1.237.2;
nuke trailing whitespace
|
| 1.236 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.235 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.234 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.233 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.232 | 07-Dec-2004 |
thorpej | branches: 1.232.2; 1.232.4;
SYNCHRONIZE_CACHE -> SYNCHRONIZE_CACHE_10
|
| 1.231 | 07-Dec-2004 |
thorpej | USe more appropriate macro/struct names for READ/WRITE (6) and
READ/WRITE (10).
|
| 1.230 | 04-Dec-2004 |
thorpej | Use READ/WRITE (16) if the LBA is larger than 32 bits. Another bit
of kern/28514.
|
| 1.229 | 28-Oct-2004 |
yamt | move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.228 | 26-Sep-2004 |
dogcow | Fix debug message output args to match yamt's src/sys/sys/buf.h changes.
|
| 1.227 | 25-Sep-2004 |
thorpej | Add support for wedges to the SCSI disk driver.
|
| 1.226 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.225 | 17-Sep-2004 |
mycroft | In places where we've already called scsipi_make_xs(), call scsipi_execute_xs()
directly rather than going through scsipi_command().
|
| 1.224 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.223 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.222 | 06-Sep-2004 |
bouyer | Fix comment: xxstart() can also be called from xxrestart()
|
| 1.221 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.220 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.219 | 28-Jun-2004 |
martin | Add prototypes for the new functions added in previous commit.
|
| 1.218 | 28-Jun-2004 |
mycroft | Separate page 4/5 reading into separate functions. For "removable" devices,
try page 5 (flex geometry) first. This fixes problems with some broken USB
devices that do not handle page 4 correctly (rather than returning nothing,
or stalling, they return the CSW where there should be data).
|
| 1.217 | 21-May-2004 |
bouyer | If READ_FORMAT_CAPACITIES fails with "Medium Format Corrupted" ASC/ASQ error,
handle it as if the drive was not formatted, so that "scsictl sdx format"
can open the raw device.
|
| 1.216 | 14-Mar-2004 |
thorpej | branches: 1.216.2;
Use scsipi_adapter_minphys().
|
| 1.215 | 28-Feb-2004 |
yamt | change the way to handle NEW_BUFQ_STRATEGY option.
instead of putting #ifdefs into each drivers,
use a global variable to indicate default strategy.
XXX should have a way to specify other strategies.
|
| 1.214 | 23-Dec-2003 |
pk | Some older devices do not understand the `disable block descriptors' bit in
the mode sense request. So fall back on mode sense data including a block
descriptor section.
XXX this applies to the `page 4' case. The `page 5' didn't consider the
possible presence of a block descriptor at all, though it did allow the
device to return one. While that's fixed now, the `page 5' mode sense
does not use the `disable block descriptors' bit.
I'm not sure we should bother with this at all..
|
| 1.213 | 29-Oct-2003 |
mycroft | Back out the bogus initializer -- the compiler bug is fixed.
|
| 1.212 | 28-Oct-2003 |
cl | note 'm68k {u,}int64_t used uninitialized' bug.
add reference to gcc bug report.
mark all (known) occurrences.
|
| 1.211 | 27-Oct-2003 |
fredb | Appease gcc3 -Wno-uninitialized (even though blkno is clearly assigned in
each branch of a conditional).
|
| 1.210 | 18-Sep-2003 |
mycroft | Whoops, make sure 6-byte commands work again.
|
| 1.209 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.208 | 13-Sep-2003 |
mycroft | Don't be silent if there is no media present.
|
| 1.207 | 13-Sep-2003 |
mycroft | Make sure the "raw partition" can always be opened again.
|
| 1.206 | 09-Sep-2003 |
mycroft | In the test for whether to start the unit, I used the wrong error code.
ENODEV is only returned when we get "medium not present," which we can fail
immediately on. All other "not ready" cases return EIO.
|
| 1.205 | 08-Sep-2003 |
mycroft | Fix an error path where we might scsipi_prevent() on a non-removable device.
|
| 1.204 | 08-Sep-2003 |
mycroft | Do a START UNIT only if the TEST UNIT READY reports that the device is not
ready. This avoids gratuitously starting the motor on floppy and CD-ROM
drives, and eliminates the need for the audio playing test in cdopen().
Therefore, also remove PQUIRK_NOSTARTUNIT.
|
| 1.203 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.202 | 29-Jun-2003 |
fvdl | branches: 1.202.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.201 | 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
|
| 1.200 | 13-May-2003 |
thorpej | Use aprint*().
|
| 1.199 | 10-May-2003 |
thorpej | Change bounds_check_with_label() to take a pointer to the disk structure,
rather than the label itself. This paves the way for some future changes.
|
| 1.198 | 02-May-2003 |
dsl | Change return type of readdisklabel() to const char *
I hope I've found all the correct places!
|
| 1.197 | 03-Apr-2003 |
fvdl | Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.196 | 20-Mar-2003 |
dbj | use PRId64 to printf bp->b_blkno, which is of type daddr_t
|
| 1.195 | 03-Feb-2003 |
thorpej | Test callout_pending(), not callout_active(), and eliminate now-unnecessary
callout_deactivate() calls.
|
| 1.194 | 23-Jan-2003 |
bad | Being able to make image backups of your whole disk, not to mention not
causing certain ata disks to lock up by reading beyond the last block,
beats catering to broken devices.
bounds_check_with_label() RAW_PART too.
|
| 1.193 | 06-Jan-2003 |
wiz | writable, not writeable.
|
| 1.192 | 15-Dec-2002 |
fvdl | Don't waste too much kernel stack space on (infrequent) ioctl operations,
use malloc instead for temp space.
|
| 1.191 | 01-Nov-2002 |
mrg | implement separate read/write disk statistics:
- disk_unbusy() gets a new parameter to tell the IO direction.
- struct disk_sysctl gets 4 new members for read/write bytes/transfers.
when processing hw.diskstats, add the read&write bytes/transfers for
the old combined stats to attempt to keep backwards compatibility.
unfortunately, due to multiple bugs, this will cause new kernels and old
vmstat/iostat/systat programs to fail. however, the next time this is
change it will not fail again.
this is just the kernel portion.
|
| 1.190 | 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
|
| 1.189 | 18-Oct-2002 |
junyoung | Add option NEW_BUFQ_STRATEGY, a new buffer queue strategy for better
system responsiness under high disk I/O load.
OK'ed by Juergen Hannken-Illjes.
|
| 1.188 | 18-Sep-2002 |
chs | remove all vesitages of dk_establish().
|
| 1.187 | 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.
|
| 1.186 | 30-Aug-2002 |
hannken | Remove the old device buffer queue interface.
Approved by: Jason R. Thorpe <thorpej@wasabisystems.com>
|
| 1.185 | 21-Jul-2002 |
hannken | Rename bufq_init() to bufq_alloc().
Add bufq_free() to remove a buffer queue.
Avoid MALLOC while holding a spinlock.
From Chuck Silvers.
|
| 1.184 | 16-Jul-2002 |
hannken | Implement a new device buffer queue interface.
One basic struct, a function to setup a queue with a specific strategy and
three macros to put buf's into the queue, get and remove the next buf or
get the next buf without removal.
The BUFQ_XXX interface will be removed in the future.
The B_ORDERED flag is not longer supported.
Approved by: Jason R. Thorpe <thorpej@wasabisystems.com>
|
| 1.183 | 15-May-2002 |
bouyer | branches: 1.183.2;
The number of sector is unsigned long, so use %lu to print it.
|
| 1.182 | 05-May-2002 |
bouyer | If periph->periph_callout is already active, don't freeze the periph again:
scispi_periph_timed_thaw() will be called only one time anyway.
|
| 1.181 | 09-Jan-2002 |
thorpej | Add ioctls to get (DIOCGCACHE) and set (DIOCSCACHE) cache enables
on disks in a generic way. Implement these ioctls for SCSI disks.
This is not fully fleshed-out yet, but it allows people to experiment
with disk caches more easily.
|
| 1.180 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.179 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.178 | 18-Jul-2001 |
thorpej | branches: 1.178.2;
bzero -> memset
|
| 1.177 | 01-Jun-2001 |
bouyer | branches: 1.177.2;
Fix typo in sddump() which caused core dumps to fail with ENXIO.
|
| 1.176 | 29-May-2001 |
chs | fix some problems in sddump() with uninitialized variables.
|
| 1.175 | 22-May-2001 |
mjacob | Make SDRETRIES an option you can dig out of opt_scsi.h.
Add SD_IO_TIMEOUT (an option in opt_scsi.h) also, which defaults to the
normal 60 * 1000 timeout for normal read/write commamds. This allows you
to shorten or lengthen the timeout as needed.
|
| 1.174 | 06-May-2001 |
drochner | Since __HAVE_OLD_DISKLABEL, DISKMINOR(u, p) isn't continuous over "p"
anymore. So we can't pass a range to vdevgone() but have to loop over the
partitions explicitely.
(otherwise the next device unit will be killed)
|
| 1.173 | 28-Apr-2001 |
tsutsui | Don't forget to set XS_CTL_DATA_OUT (that was removed in previous).
|
| 1.172 | 28-Apr-2001 |
thorpej | Use simple tags for read and write, unless B_ORDERED is set, in
which case we use an ordered tag.
|
| 1.171 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.170 | 09-Mar-2001 |
explorer | Only lock the pack in if the disk device is a removable one. This fixes a
problem with Hitachi SCA drives
|
| 1.169 | 08-Jan-2001 |
fvdl | branches: 1.169.2;
Return error in the case of using ODIOCGDINFO or ODIOCGDEFLABEL when
the number of partitions is > OLDMAXPARTITIONS. This is better
than silently truncating the label (don't want to silently throw
away partitions when using an old disklabel binary on a label with
> 8 partitions). From Enami Tsugutomo.
|
| 1.168 | 07-Jan-2001 |
fvdl | Adapt all disk devices in MI directories to handle ODIOC* calls
for ports that have bumped MAXPARTITIONS (and thus define
__HAVE_OLD_DISKLABEL).
|
| 1.167 | 08-Dec-2000 |
augustss | Don't access (in core) disk label of a removable disk that is not there.
|
| 1.166 | 24-Nov-2000 |
chs | avoid division and remainder operations for devices where
the sector size is DEV_BSIZE (ie. practically everything).
|
| 1.165 | 22-Nov-2000 |
soren | Remove obsolete comment.
|
| 1.164 | 05-Jul-2000 |
thorpej | Garbage-collect __BDEVSW_DUMP_OLD_TYPE.
|
| 1.163 | 09-Jun-2000 |
enami | branches: 1.163.2;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.162 | 30-May-2000 |
augustss | Add a quirk, SDEV_ONLYBIG, which implies that the device cannot handle
the 6 byte versions of READ, WRITE, and MODE_SENSE.
This greatly simplifies the UFI (USB Floppy) handling.
|
| 1.161 | 23-May-2000 |
bouyer | branches: 1.161.2;
sd_interpret_sense() can be called from interrupt context; call scsipi_start()
with XS_CTL_ASYNC if the failed command was called with XS_CTL_ASYNC.
Add a SDF_RESTART flag to keep state, cleared in sddone().
A mounted disk can now spin down, it will propely spin up at the next access.
|
| 1.160 | 16-May-2000 |
thorpej | Nuke dk_establish() from orbit except from those ports which still use
it to determine the boot device: mvme68k, pc532, macppc, ofppc. Those
platforms should be changed to use device_register(). In the mean time,
those ports defined __BROKEN_DK_ESTABLISH.
|
| 1.159 | 30-Mar-2000 |
augustss | Get rid of register declarations.
|
| 1.158 | 13-Mar-2000 |
soren | Fix doubled 'the's in comments.
|
| 1.157 | 07-Feb-2000 |
thorpej | Fix a bug in disksort_*() which caused non-optimal ordering when multiple
active partitions were on a single spindle. Add a b_rawblkno member to
struct buf which contains the non-partition-relative block number to sort
by.
|
| 1.156 | 25-Jan-2000 |
drochner | try to dtrt for sector sizes <DEV_BSIZE (at least, don't divide by 0)
|
| 1.155 | 21-Jan-2000 |
thorpej | Update for sys/buf.h/disksort_*() changes.
|
| 1.154 | 23-Dec-1999 |
leo | Implement the DIOCKLABEL ioctl. Through this ioctl it it possible keep the
disklabel on the last close or to drop it.
|
| 1.153 | 03-Nov-1999 |
matt | use __vax__ and __i386__ instead of vax and i386
|
| 1.152 | 30-Oct-1999 |
bouyer | In sddump(), check sd->sc_dev.dv_flags once sd has been initialised.
Patch supplied by tsutsui@ceres.dti.ne.jp in kern/8708.
|
| 1.151 | 17-Oct-1999 |
ragge | branches: 1.151.2; 1.151.4;
Don't call dk_establish() on vax either.
|
| 1.150 | 17-Oct-1999 |
enami | Cosmetic changes.
|
| 1.149 | 30-Sep-1999 |
thorpej | branches: 1.149.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.148 | 21-Sep-1999 |
enami | Calculate the correct device minor of the first partition.
|
| 1.147 | 11-Sep-1999 |
thorpej | Implement detaching SCSI and ATAPI disks.
|
| 1.146 | 26-Aug-1999 |
hannken | Make `sc_ops->sdo_flush' synchronous. The flags `SDF_FLUSHING' and
`SDF_DIRTY' were never reset because `sddone' doesn't get called from
synchronous scsi commands.
|
| 1.145 | 31-May-1999 |
lukem | * use format_bytes() to display the disk size
* wrap a couple of printf()s
|
| 1.144 | 28-Feb-1999 |
explorer | branches: 1.144.4;
Update to slightly altered rnd_attach_source() api
|
| 1.143 | 10-Feb-1999 |
bouyer | Abort transfer if b_blkno is negative. Closes PR kern/5553 by Johan Danielsson.
|
| 1.142 | 08-Feb-1999 |
bouyer | Change DIOCEJECT to do what's needed to eject a device before the eject
command (unlock for sd and cd) if no other partitions are open, return
EBUSY otherwise. DIOCEJECT will have the old semantic if its argument is not
0. The old ioctl has been renamed to ODIOCEJECT for binary compatibility.
|
| 1.141 | 29-Jan-1999 |
bouyer | Return ENODEV instead of EIO when we are trying to open a device without media
in the drive. restrict "opening of empty drive" to character devices only
(reading a block device returns a short read instead of ENODEV, which can lead
to confusion).
|
| 1.140 | 26-Jan-1999 |
bouyer | Allows the raw partition to be open()'ed, even when scsipi_start() fails
(no media or other ...) so that we can always send ioctl's to the device.
|
| 1.139 | 19-Jan-1999 |
bouyer | Move test for SDEV_NOSTARTUNIT quirk from sd.c to scsipi_start().
Add a SDEV_NOSTARTUNIT quirk entry for BCD-16X 1997-04-25", "", "VER 2.2"
CD-rom (from Michael Santos).
|
| 1.138 | 08-Dec-1998 |
thorpej | When closing, wait for pending xfers to drain before unlocking the door,
and wait again before deleting the reference to the adapter.
|
| 1.137 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.136 | 17-Nov-1998 |
bouyer | The ATAPI-specific hack is not needed any more.
|
| 1.135 | 17-Aug-1998 |
thorpej | Only mark the disk dirty if we write.
|
| 1.134 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.133 | 15-Aug-1998 |
thorpej | Should not use SCSI_POLL if the system is not cold.
|
| 1.132 | 15-Aug-1998 |
thorpej | Keep track if the disk has had done i/o, making it dirty. On last close,
or at shutdown, flush the cache if the disk is still dirty and it support
cache flushing. Fixes PRs #5588 (Jim Bernard) and #5589 (Justin Gibbs).
|
| 1.131 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.130 | 31-Jul-1998 |
thorpej | Typo.
|
| 1.129 | 30-Jul-1998 |
thorpej | From cd.c: call disk_unbusy() if we are unable to enqueue our command.
|
| 1.128 | 30-Jul-1998 |
mjacob | branches: 1.128.2;
opt_scsiverbose.h now is opt_scsi.h and contains the defopted
SCSIVERBOSE and the new ST_ENABLE_EARLYWARN.
|
| 1.127 | 15-Jul-1998 |
mjacob | Fix for kern/3835: add an sd sense handler. If we get a check condition/sense
data ASC/ASCQ of 0x04/0x01 (logical unit not ready, initialization in progress),
hang out for 5 seconds and return a RETRY THE OPERATION command. If we
get a check condition/sense data ASC/ASCQ of 0x04/0x02 (logical unit not
ready, initialization command required), send a polled/nosleep START UNIT
command and return a RETRY THE OPERATION command if that succeeds.
Don't send a START UNIT to a disk quirked as SDEV_NO_START. Don't send
a START UNIT to removable media. The reason for the latter is to not
just blindly spin up new (maybe changed) media.
I should note that I've successfully made this work with the ISP host
adapter so far. Other host adapters will need some work to be able
to manage or reasonably fail NOSLEEP/POLL commands while in this
state. Alternatively, the internal SCSI midlayer structure has got
to allow for more controlled error recovery (e.g., restart queues
controlled by the target driver).
|
| 1.126 | 03-Mar-1998 |
cgd | in the ATAPI bug workaround code ("HACK"), don't forget to set
SCSI_AUTOCONF when executing SCSI commands.
|
| 1.125 | 15-Jan-1998 |
cgd | #if some bug workaround code which only needs to be compiled if
ATAPI 'sd's are present.
|
| 1.124 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.123 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.122 | 18-Oct-1997 |
thorpej | Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.121 | 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.
|
| 1.120 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.119 | 08-Oct-1997 |
thorpej | Implement DIOCGDEFLABEL.
|
| 1.118 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.117 | 20-Sep-1997 |
mjacob | Allow SDOUTSTANDING to be defined elsewhere (e.g., config file).
|
| 1.116 | 27-Aug-1997 |
bouyer | branches: 1.116.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.115 | 04-Aug-1997 |
fair | trivial change for PR#3815
|
| 1.114 | 02-Jul-1997 |
phil | branches: 1.114.2;
Report actual disksize on finding a disk.
(From: der Mouse <mouse@Rodents.Montreal.QC.CA>)
|
| 1.113 | 24-Jun-1997 |
thorpej | foosize()'s return value is in DEV_BSIZE units; adjust the size obtained
from the disklabel accordingly.
|
| 1.112 | 18-Jun-1997 |
pk | sdsize(): only call sdopen()/sdclose() if not already open, per the
current openmask.
|
| 1.111 | 02-Apr-1997 |
mycroft | Push the buffer cleanup code into scsi_done(), and split it so that biodone()
is called *after* the driver `done' routine. This fixes disk I/O statistics
on SCSI devices.
Also, calling the `done' routine with a `complete' argument of 0 and actually
having it do anything meaningful loses in at least 3 ways, so just nuke the
argument altogether and don't call it this way. If the driver needs to do
some error handling, that's what `err_handler' is for.
|
| 1.110 | 04-Mar-1997 |
mikel | garbage collect unused variable (was this struct declaration orginally
a typedef?); from Enami Tsugutomo in PR kern/3282.
|
| 1.109 | 21-Feb-1997 |
thorpej | If posting an error condition because the media has been unloaded,
make sure to set the residual count to reflect that no data was
transfered.
From Naofumi HONDA / MINOURA Makoto, PR #3007.
|
| 1.108 | 21-Feb-1997 |
thorpej | Handle optical disks a little better, from Naofumi HONDA / MINOURA Makoto
in PR #3009, with some slight KNF from me.
|
| 1.107 | 05-Dec-1996 |
cgd | branches: 1.107.6;
update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.106 | 29-Nov-1996 |
thorpej | Start a mode sense with a clean sense buffer.
Fixes PR #2953, from David M. Stanhope <dms@celtech.com>.
|
| 1.105 | 23-Oct-1996 |
matthias | * At least the Teac FC-1 Shugart-SCSI bridgeboard does motor on/off
management by itself. But when it gets a start unit request, it keeps
the floppy motor running all the time. This adds code for dealing with
yet another quirk (SDEV_NOSTARTUNIT) that prevents sd.c from sending
start unit requests. A entry for the Teac FC-1 is added to the quirk
table.
|
| 1.104 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.103 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.102 | 05-Jul-1996 |
christos | - Apply PR/2535: Add support for flex scsi disks.
- Add a quirk called SCSI_NOMODESENSE for drives like the iomega jaz,
that don't support mode sense.
|
| 1.101 | 04-Jun-1996 |
thorpej | In sddump(), don't check the SDEV_MEDIA_LOADED bit of the device, since
that bit might have been cleared by successful unmounting of all filesystems
on the dump device, e.g. in the case of "reboot -d".
|
| 1.100 | 14-May-1996 |
leo | branches: 1.100.4;
Clear SDEV_MEDIA_LOADED bit on the last close of a device. This causes the
geometry info & the disklabel to be reloaded on each 'first-open' of the
device.
|
| 1.99 | 14-May-1996 |
thorpej | Print the error code in the "not queued" message, to make debugging
a bit easier, and make sure a newline gets printed.
|
| 1.98 | 05-May-1996 |
christos | Cleanup the rest of the SCSIDEBUG printfs. From Bernd Ernesti.
|
| 1.97 | 03-May-1996 |
christos | - SCSIDEBUG printf() fixes.
|
| 1.96 | 22-Apr-1996 |
christos | remove include of <sys/cpu.h>
|
| 1.95 | 30-Mar-1996 |
christos | Eliminate scsi_conf.h.
|
| 1.94 | 26-Mar-1996 |
mycroft | Put back dk_establish() for now.
|
| 1.93 | 26-Mar-1996 |
mycroft | Increase *OUTSTANDING.
Remove old dk_establish() garbage.
Support the `ancient' bit in both drivers.
|
| 1.92 | 24-Mar-1996 |
mycroft | Fix trivial oversight in last.
|
| 1.91 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.90 | 17-Mar-1996 |
pk | Parentheses are your friends.
|
| 1.89 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.88 | 05-Mar-1996 |
thorpej | Normalize SCSI autoconfiguration output.
From Chris Demetriou <cgd@NetBSD.ORG>. Fixes PR #1958.
|
| 1.87 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.86 | 30-Jan-1996 |
thorpej | Implement DIOCLOCK and DIOCEJECT. DIOCEJECT is limited to removable
media. Fixes PR #1975.
|
| 1.85 | 12-Jan-1996 |
thorpej | Handle cases like the following:
- controller calls scsi_done() with error XS_TIMEOUT
- scsi_done() calls sddone()
- sddone() calls disk_unbusy()
- scsi_done() calls controller to retry command (missing the
call to disk_busy())
- controller calls scsi_done()
- scsi_done() calls sddone()
- sddone() calls disk_busy(), which panics because of the imbalance.
Bug noticed by Leo Weppleman, who also suggested this fix; pass an additional
boolean argument ("complete") to the device's "done" routine, with a
value of `0' passed from the previous call to "done", and add an additional
call to "done" when the xfer resources are freed.
|
| 1.84 | 07-Jan-1996 |
thorpej | New generic disk framework. Highlights:
- New metrics handling. Metrics are now kept in the new
`struct disk'. Busy time is now stored as a timeval, and
transfer count in bytes.
- Storage for disklabels is now dynamically allocated, so that
the size of the disk structure is not machine-dependent.
- Several new functions for attaching and detaching disks, and
handling metrics calculation.
Old-style instrumentation is still supported in drivers that did it before.
However, old-style instrumentation is being deprecated, and will go away
once the userland utilities are updated for the new framework.
For usage and architectural details, see the forthcoming disk(9) manual
page.
|
| 1.83 | 07-Dec-1995 |
thorpej | Remove a couple of now unneeded assignments.
|
| 1.82 | 07-Dec-1995 |
thorpej | In sdminphys(), add a comment about the semantics of the "length" field
in a 6-byte read/write cdb (namely, length == 0 really means 256), and
that we handle this conservatively by limiting the transfer to 255 blocks.
*sigh* One just never knows how a pre-scsi-1 device is going to behave.
|
| 1.81 | 07-Dec-1995 |
thorpej | In both cd.c and sd.c:
If the read or write request can fit into a 6-byte cdb, then
use a 6-byte cdb, otherwise use the 10-byte as before.
In sd.c:
In sdattach(), make a note if the device is "ancient" (i.e.
inqbuf.version & SID_ANSII == 0).
Implement sdminphys(): if the device is "ancient", shorten the
transfer so it will fit into a 6-byte cdb.
These changes have eliminated the rejected read/write requests on my
Sun 3/60 with 2 ESDI disks behind an Emulex MD21.
|
| 1.80 | 10-Oct-1995 |
mycroft | branches: 1.80.2;
Return EINVAL if something other than a whole number of blocks is requested.
|
| 1.79 | 30-Sep-1995 |
pk | Specify SCSI_DATA_OUT in sddump() (from John Kohl; PR#1482).
|
| 1.78 | 26-Sep-1995 |
thorpej | Don't declare Debugger(). It's handled in <sys/systm.h>.
#include <sys/systm.h> where necessary, as suggested by Jonathan Stone.
Fixes PR #1511.
|
| 1.77 | 12-Aug-1995 |
mycroft | opri --> s
|
| 1.76 | 12-Aug-1995 |
mycroft | Fix oversight in previous.
|
| 1.75 | 12-Aug-1995 |
mycroft | minphys() functions really should return void.
|
| 1.74 | 05-Aug-1995 |
mycroft | Use an intermediate variable to shorten label initialization code.
|
| 1.73 | 24-Jul-1995 |
cgd | update SCSI minphys routines' definitions to match standard minphys()
definition and usage.
|
| 1.72 | 04-Jul-1995 |
mycroft | Make each disk and tape driver define its own read and write functions.
Deprecate rawread() and rawwrite() completely. Remove d_strategy from cdevsw to
force the abstraction barrier.
|
| 1.71 | 26-Jun-1995 |
cgd | implement new-style block device dump routine if __BDEVSW_DUMP_OLD_TYPE
not defined, otherwise use an "not implemented" stub.
|
| 1.70 | 03-May-1995 |
cgd | looks like some extra ')'s were left in here.
|
| 1.69 | 03-May-1995 |
mycroft | Make the byte-shifting code consistent.
|
| 1.68 | 15-Apr-1995 |
mycroft | Don't boundary check I/O to the `raw' partition.
|
| 1.67 | 10-Apr-1995 |
mycroft | kernel_pmap --> pmap_kernel()
|
| 1.66 | 01-Apr-1995 |
mycroft | Add a missing unlock.
|
| 1.65 | 29-Mar-1995 |
mycroft | Revamp the locking mechanism slightly.
|
| 1.64 | 25-Mar-1995 |
mycroft | Remove the write protect check altogether, and rely on the drive to do it.
|
| 1.63 | 23-Mar-1995 |
mycroft | Check for read-only media in open(), not write().
|
| 1.62 | 23-Mar-1995 |
mycroft | Rearrange DIOCWDINFO a little; closer to wd.c.
|
| 1.61 | 23-Mar-1995 |
mycroft | Allow DIOCWDINFO even if label is not `writable'.
|
| 1.60 | 07-Mar-1995 |
mycroft | Correct spelling of `fictitious'. Add patterns for optical memory devices.
From Alistair Crooks.
|
| 1.59 | 30-Jan-1995 |
mycroft | Some drives don't grok START with LoEj=1, either.
|
| 1.58 | 30-Jan-1995 |
mycroft | Set the LoadEject bit when issuing a START.
|
| 1.57 | 30-Jan-1995 |
mycroft | Some devices really do require the START before the PREVENT. This makes no
sense.
|
| 1.56 | 26-Jan-1995 |
mycroft | Update copyrights.
|
| 1.55 | 26-Jan-1995 |
mycroft | Redo the `wait for spin up' code. The Mach 3 method clearly doesn't work.
Also, rearrange the first open sequences a bit; SDEV_OPEN is no longer magic.
|
| 1.54 | 23-Jan-1995 |
mycroft | Do the PREVENT before the START.
|
| 1.53 | 16-Jan-1995 |
mycroft | If the disk reports a block size of 0, assume it really meant 512.
|
| 1.52 | 13-Jan-1995 |
mycroft | Always boundary check I/O.
|
| 1.51 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.50 | 16-Dec-1994 |
mycroft | Remove DIOCSBAD handling. It's not actually handled, so why recognize it?
|
| 1.49 | 14-Dec-1994 |
mycroft | Remove dkbad.h.
|
| 1.48 | 23-Nov-1994 |
mycroft | There's no point in bothering to unlock what we didn't lock.
|
| 1.47 | 22-Nov-1994 |
mycroft | Add still more locking.
|
| 1.46 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.45 | 20-Nov-1994 |
mycroft | Add some missing locking, and some general cleanup.
|
| 1.44 | 15-Nov-1994 |
deraadt | Enough Chinese torture already
|
| 1.43 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.42 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.41 | 20-Oct-1994 |
mycroft | Back out change to print geometry from label at boot; can't do this reliably.
|
| 1.40 | 20-Oct-1994 |
mycroft | openpart -> openmask
|
| 1.39 | 20-Oct-1994 |
mycroft | Trivial stylistic changes.
|
| 1.38 | 20-Oct-1994 |
mycroft | Use the block size in the label in most cases.
|
| 1.37 | 20-Oct-1994 |
mycroft | Various changes:
* Get rid of unused `partflags'.
* Use dk_openpart, and remove `openparts'. Also, allow an open on both
the block and characters devices at the same time.
* If there was a label, print the geometry from the label at boot time.
* Turn `wlabel' into a flag, and *don't* set it automatically in
DIOCWLABEL; disklabel(8) does that.
Other minor changes.
|
| 1.36 | 14-Oct-1994 |
cgd | move arch dependence to archs. clean up a bit. deal with weird MAXPARTITONS.
|
| 1.35 | 26-Jul-1994 |
mycroft | Normalize label handling.
|
| 1.34 | 04-Jul-1994 |
chopps | branches: 1.34.2;
change to support other MAXPARTITIONS vals aside from 8
|
| 1.33 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.32 | 16-Jun-1994 |
chopps | resolve confusion over who owns the buf after calling scsi_scsi_cmd()
|
| 1.31 | 16-Jun-1994 |
mycroft | b_un.b_addr -> b_data
|
| 1.30 | 09-May-1994 |
chopps | remove union's from sense_data struct, conditionaly define RAW_PART
|
| 1.29 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.28 | 11-Apr-1994 |
mycroft | Combine scsi_start_unit() and scsi_stop_unit(), and increase the start
timeout (again).
|
| 1.27 | 06-Apr-1994 |
mycroft | Redo the last change; it was compltely bogus.
|
| 1.26 | 06-Apr-1994 |
mycroft | Try to detect garbage data from page 4 sense.
|
| 1.25 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.24 | 24-Mar-1994 |
cgd | print even more SCSI sense data; this should be in a function...
|
| 1.23 | 24-Mar-1994 |
cgd | print out lots of info on non-media hardware error...
|
| 1.22 | 06-Feb-1994 |
mycroft | Use b_actf, not av_forw.
|
| 1.21 | 25-Jan-1994 |
deraadt | print nice numbers for disks >2G. from <adrian@lemmings.cursci.com>
|
| 1.20 | 11-Jan-1994 |
mycroft | *strategy functions return void.
|
| 1.19 | 17-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.18 | 01-Aug-1993 |
mycroft | branches: 1.18.2;
Add RCS identifiers (this time on the correct side of the branch), and
incorporate recent changes in netbsd-0-9 branch.
|
| 1.17 | 19-Jul-1993 |
cgd | branches: 1.17.2;
patches from allen briggs to fix a minor bug in *attach()
|
| 1.16 | 04-Jul-1993 |
andrew | Reject non block-aligned transfers (where a block is 512 bytes in length);
they were causing scsi controller lockups. Replace an occurrence of "512"
with SECSIZE, which is #defined to be 512.
|
| 1.15 | 27-Jun-1993 |
andrew | ANSIfications.
|
| 1.14 | 16-Jun-1993 |
deraadt | whoops. typo.
|
| 1.13 | 16-Jun-1993 |
deraadt | fix to intuit the maximum number of scsi units available on a device
driver. this piece at least, should be safe from changing sizeof(dev_t)
|
| 1.12 | 25-May-1993 |
deraadt | patch00149 by Julian Elischer <julian@jules.dialix.oz.au> & Rodney Grimes.
When an error was encountered, the sd/cd drivers printed blockno&ff0000
rather that blockno.
|
| 1.11 | 21-May-1993 |
davidb | Changed when disks are started, to get DEC disks working.
|
| 1.10 | 20-May-1993 |
deraadt | more disklabel changes
|
| 1.9 | 20-May-1993 |
cgd | add rcsids and clean up file headers
|
| 1.8 | 04-May-1993 |
deraadt | support for making dev->id_alive be set, this is for iostat to
find disk devices. wee bit of a kludge. sub-device attach()
routines must now return 1 for successful attach(), 0 otherwise.
Other bsd's do this too..
|
| 1.7 | 26-Apr-1993 |
deraadt | cosmetic change
|
| 1.6 | 20-Apr-1993 |
mycroft | Display more meaningful message on SCSI `unit attention'.
|
| 1.5 | 12-Apr-1993 |
deraadt | fixed various bugs like cdattach() returning garbage.
|
| 1.4 | 12-Apr-1993 |
deraadt | small fixes
|
| 1.3 | 12-Apr-1993 |
deraadt | new scsi subsystem.
changes also in config/mkioconf.c
i386/isa/wd.c, fd.c, and all scsi drivers.
|
| 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 | after 0.2.2 "stable" patches applied
|
| 1.17.2.1 | 31-Jul-1993 |
cgd | give names, err, wmesg's, to my "pain" -- i.e. convert sleep() to tsleep()
|
| 1.18.2.12 | 16-Feb-1994 |
mycroft | More KNF-like.
|
| 1.18.2.11 | 06-Feb-1994 |
mycroft | Use b_actf, not av_forw.
|
| 1.18.2.10 | 01-Feb-1994 |
mycroft | Add scsi_stop_unit(), from Mac code. Other minor nits.
|
| 1.18.2.9 | 29-Nov-1993 |
mycroft | Use dk_establish().
|
| 1.18.2.8 | 25-Nov-1993 |
mycroft | Fix silly typo.
|
| 1.18.2.7 | 25-Nov-1993 |
mycroft | (Hopefully) make cd work. Make cd and sd more similar. Implement variable
block sizes in sd.
|
| 1.18.2.6 | 24-Nov-1993 |
mycroft | Final tweaks to get it going.
|
| 1.18.2.5 | 24-Nov-1993 |
mycroft | Fix unitialized variable.
|
| 1.18.2.4 | 24-Nov-1993 |
mycroft | More construction...
|
| 1.18.2.3 | 24-Nov-1993 |
mycroft | Still under construction...
|
| 1.18.2.2 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.18.2.1 | 24-Sep-1993 |
mycroft | cd.c, sd.c, st.c: strategy functions return void.
|
| 1.34.2.1 | 26-Jul-1994 |
cgd | from branch.
|
| 1.80.2.1 | 04-Feb-1996 |
gwr | Pull in rev 1.82 (Jason's fix to use 6-byte command blocks).
|
| 1.100.4.3 | 04-Mar-1997 |
mycroft | Pull up support for fetching geometry info on optical disks.
|
| 1.100.4.2 | 29-Nov-1996 |
thorpej | Start a mode sense with a clean sense buffer.
From David M. Stanhope <dms@celtech.com>, PR #2953.
|
| 1.100.4.1 | 04-Jun-1996 |
thorpej | Update from trunk:
In sddump(), don't check the SDEV_MEDIA_LOADED bit of the device, since
that bit might have been cleared by successful unmounting of all filesystems
on the dump device, e.g. in the case of "reboot -d".
|
| 1.107.6.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.114.2.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.116.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.116.2.3 | 22-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.116.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.116.2.1 | 27-Aug-1997 |
thorpej | file sd.c was added on branch marc-pcmcia on 1997-08-27 23:33:32 +0000
|
| 1.128.2.2 | 08-Aug-1998 |
eeh | Revert cdevsw mmap routines to return int.
|
| 1.128.2.1 | 30-Jul-1998 |
eeh | file sd.c was added on branch eeh-paddr_t on 1998-08-08 03:06:52 +0000
|
| 1.144.4.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.149.2.2 | 27-Dec-1999 |
wrstuden | Pull up to last week's -current.
|
| 1.149.2.1 | 21-Dec-1999 |
wrstuden | Initial commit of recent changes to make DEV_BSIZE go away.
Runs on i386, needs work on other arch's. Main kernel routines should be
fine, but a number of the stand programs need help.
cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512
byte block devices. vnd, raidframe, and lfs need work.
Non 2**n block support is automatic for LKM's and conditional for kernels
on "options NON_PO2_BLOCKS".
|
| 1.151.4.1 | 15-Nov-1999 |
fvdl | Sync with -current
|
| 1.151.2.9 | 12-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.151.2.8 | 18-Jan-2001 |
bouyer | Sync with head (for UBC+NFS fixes, mostly).
|
| 1.151.2.7 | 13-Dec-2000 |
bouyer | Make it compile again.
|
| 1.151.2.6 | 13-Dec-2000 |
bouyer | Sync with HEAD (for UBC fixes).
|
| 1.151.2.5 | 08-Dec-2000 |
bouyer | Sync with HEAD.
|
| 1.151.2.4 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.151.2.3 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.151.2.2 | 20-Oct-1999 |
thorpej | Remove an unnecessary comment regarding XS_CTL_NOSLEEP.
|
| 1.151.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.161.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.163.2.3 | 06-May-2001 |
he | Pull up revision 1.174 (requested by fvdl):
Compensate for the fact that after __HAVE_OLD_DISKLABEL,
DISKMINOR(u,p) is no longer contiguous over ``p'', so ranges
cannot be passed to vdevgone().
|
| 1.163.2.2 | 01-May-2001 |
he | Pull up revisions 1.167-1.168 (requested by fvdl):
Increase the number of BSD disklabel partitions on i386 to 16.
|
| 1.163.2.1 | 11-Mar-2001 |
he | Pull up revision 1.170 (requested by explorer):
Fix a problem with some Hitachi drives by only issuing the ``lock
media'' and ``unlock media'' commands to removable media.
|
| 1.169.2.14 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.169.2.13 | 19-Dec-2002 |
thorpej | Sync with HEAD.
|
| 1.169.2.12 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.169.2.11 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.169.2.10 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.169.2.9 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.169.2.8 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.169.2.7 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.169.2.6 | 11-Jan-2002 |
nathanw | More catchup.
|
| 1.169.2.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.169.2.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.169.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.169.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.169.2.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 1.177.2.5 | 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
|
| 1.177.2.4 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.177.2.3 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.177.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.177.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.178.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.178.2.1 | 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).
|
| 1.183.2.4 | 31-Aug-2002 |
gehenna | catch up with -current.
|
| 1.183.2.3 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.183.2.2 | 20-Jul-2002 |
gehenna | catch up with -current.
|
| 1.183.2.1 | 16-May-2002 |
gehenna | Add the block/character device switches.
Replace the direct-access to devsw table with calling devsw API.
|
| 1.202.2.13 | 11-Dec-2005 |
christos | Sync with head.
|
| 1.202.2.12 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.202.2.11 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.202.2.10 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.202.2.9 | 18-Dec-2004 |
skrll | Sync with HEAD.
|
| 1.202.2.8 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.202.2.7 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.202.2.6 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.202.2.5 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.202.2.4 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.202.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.202.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.202.2.1 | 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.
|
| 1.216.2.3 | 11-Sep-2004 |
he | Pull up revisions 1.221-1.223 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.216.2.2 | 02-Jul-2004 |
he | Pull up revisions 1.218-1.219 (requested by mycroft in ticket #572):
Several fixes mostly related to USB:
o Separate page 4/5 reading into separate functions. For
``removeable'' devices, try page 5 first. Fixes problem
with some broken USB devices which do not handle page 4
correctly.
o Add prototypes for recently added functions.
|
| 1.216.2.1 | 29-May-2004 |
tron | Pull up revision 1.217 (requested by bouyer in ticket #399):
If READ_FORMAT_CAPACITIES fails with "Medium Format Corrupted" ASC/ASQ error,
handle it as if the drive was not formatted, so that "scsictl sdx format"
can open the raw device.
|
| 1.232.4.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.232.2.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.237.2.3 | 29-Dec-2005 |
riz | Pull up following revision(s) (requested by bouyer in ticket #1045):
sys/dev/scsipi/sd.c: revision 1.243
Adapt previous change to recent scsipi. Pointed out by Simon Burge.
|
| 1.237.2.2 | 29-Dec-2005 |
riz | Pull up following revision(s) (requested by bouyer in ticket #1045):
sys/dev/scsipi/sd.c: revision 1.242
Some USB devices reports themselves as removable, but have no door and
so don't support the SCSI_PREVENT_ALLOW_MEDIUM_REMOVAL command.
When an "Illegal field in CDB" is reported for this command, mark the
device as non-removable (which is always true for USB keys from the SCSI
point of view), print a message and ignore the error.
For DIOCLOCK, return ENOTTY if the device is not removable instead of
trying a command which will fail.
Fix a problem reported by Hubert Feyrer for some USB umass devices, patch
tested by him.
|
| 1.237.2.1 | 06-Apr-2005 |
tron | Pull up revision 1.238 (requested by yamt in ticket #112):
introduce a function to drain bufq and use it where appropriate.
|
| 1.240.2.8 | 17-Mar-2008 |
yamt | sync with head.
|
| 1.240.2.7 | 27-Feb-2008 |
yamt | sync with head.
|
| 1.240.2.6 | 21-Jan-2008 |
yamt | sync with head
|
| 1.240.2.5 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.240.2.4 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.240.2.3 | 26-Feb-2007 |
yamt | sync with head.
|
| 1.240.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.240.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.244.6.2 | 01-Jun-2006 |
kardel | Sync with head.
|
| 1.244.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.244.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.244.2.1 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.245.6.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.245.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.245.2.3 | 11-Aug-2006 |
yamt | sync with head
|
| 1.245.2.2 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.245.2.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.247.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.248.2.1 | 13-Jul-2006 |
gdamore | Merge from HEAD.
|
| 1.249.4.2 | 12-Jan-2007 |
ad | Sync with head.
|
| 1.249.4.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.249.2.1 | 22-Sep-2006 |
riz | Pull up following revision(s) (requested by reinoud in ticket #168):
sys/dev/scsipi/sd.c: revision 1.250
Fix same panic problem as with cd.c when a scsi error happends on
read/write action. It would panic on a diagnosic assertion. When a scsi
read/write command fails, the whole extent is gone bad so mark all as not
done.
|
| 1.250.2.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.250.2.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.258.6.1 | 03-Sep-2007 |
wrstuden | Sync w/ NetBSD-4-RC_1
|
| 1.258.4.2 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.258.4.1 | 27-Feb-2007 |
yamt | - sync with head.
- move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
|
| 1.258.2.1 | 31-Jul-2007 |
liamjfoy | Pull up following revision(s) (requested by jnemeth in ticket #803):
sys/dev/scsipi/sd.c: revision 1.260
add support devctl properties
|
| 1.261.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.261.2.5 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.261.2.4 | 20-Aug-2007 |
ad | - Alter disk attach/detach to fix a panic when closing a vnd device.
- Sync with HEAD.
|
| 1.261.2.3 | 19-Aug-2007 |
ad | - Back out the biodone() changes.
- Eliminate B_ERROR (from HEAD).
|
| 1.261.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.261.2.1 | 13-May-2007 |
ad | - Pass the error number and residual count to biodone(), and let it handle
setting error indicators. Prepare to eliminate B_ERROR.
- Add a flag argument to brelse() to be set into the buf's flags, instead
of doing it directly. Typically used to set B_INVAL.
- Add a "struct cpu_info *" argument to kthread_create(), to be used to
create bound threads. Change "bool mpsafe" to "int flags".
- Allow exit of LWPs in the IDL state when (l != curlwp).
- More locking fixes & conversion to the new API.
|
| 1.262.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.265.8.2 | 03-Aug-2007 |
tsutsui | Don't allocate data buffer on stack in sd_read_capacity().
If data buffer is allocated on stack at the end of CPU cacheline,
the same cacheline region might be accessed via kernel stack even
after bus_dmamap_sync(9) against the data buffer in lower scsi drivers
so that read data via DMA could be lost.
Fixes occasional readdisklabel(9) failure at boot on ews4800mips
with osiop(4) driver and 128byte cacheline.
XXX: Other functions/drivers might have the similar problem.
|
| 1.265.8.1 | 03-Aug-2007 |
tsutsui | file sd.c was added on branch matt-mips64 on 2007-08-03 13:56:38 +0000
|
| 1.265.6.2 | 14-Oct-2007 |
yamt | sync with head.
|
| 1.265.6.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.265.4.3 | 23-Mar-2008 |
matt | sync with HEAD
|
| 1.265.4.2 | 09-Jan-2008 |
matt | sync with HEAD
|
| 1.265.4.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.265.2.5 | 08-Dec-2007 |
jmcneill | Rename pnp(9) -> pmf(9), as requested by many.
|
| 1.265.2.4 | 06-Nov-2007 |
joerg | Refactor PNP API:
- Make suspend/resume directly a device functionality. It consists of
three layers (class logic, device logic, bus logic), all of them being
optional. This replaces D0/D3 transitions.
- device_is_active returns true if the device was not disabled and was
not suspended (even partially), device_is_enabled returns true if the
device was enabled.
- Change pnp_global_transition into pnp_system_suspend and
pnp_system_resume. Before running any suspend/resume handlers, check
that all currently attached devices support power management and bail
out otherwise. The latter is not done for the shutdown/panic case.
- Make the former bus-specific generic network handlers a class handler.
- Make PNP message like volume up/down/toogle PNP events. Each device
can register what events they are interested in and whether the handler
should be global or not.
- Introduce device_active API for devices to mark themselve in use from
either the system or the device. Use this to implement the idle handling
for audio and input devices. This is intended to replace most ad-hoc
watchdogs as well.
- Fix somes situations in which audio resume would lose mixer settings.
- Make USB host controllers better deal with suspend in the light of
shared interrupts.
- Flush filesystem cache on suspend.
- Flush disk caches on suspend. Put ATA disks into standby on suspend as
well.
- Adopt drivers to use the new PNP API.
- Fix a critical bug in the generic cardbus layer that made D0->D3
break.
- Fix ral(4) to set if_stop.
- Convert cbb(4) to the new PNP API.
- Apply the PCI Express SCI fix on resume again.
|
| 1.265.2.3 | 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.
|
| 1.265.2.2 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.265.2.1 | 03-Aug-2007 |
joerg | file sd.c was added on branch jmcneill-pm on 2007-10-02 18:28:39 +0000
|
| 1.267.8.1 | 11-Dec-2007 |
yamt | sync with head.
|
| 1.267.6.1 | 26-Dec-2007 |
ad | Sync with head.
|
| 1.267.4.1 | 18-Feb-2008 |
mjf | Sync with HEAD.
|
| 1.268.2.1 | 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.269.6.8 | 17-Jan-2009 |
mjf | - Fix a CVS merge breakage
- Catch up with some driver changes
|
| 1.269.6.7 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.269.6.6 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.269.6.5 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.269.6.4 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.269.6.3 | 06-Apr-2008 |
mjf | - after some discussion with agc@ i agreed it would be a good idea to move
device_unregister_* to device_deregister_* to be more like the pmf(9)
functions, especially since a lot of the time the function calls are next
to each other.
- add device_register_name() support for dk(4).
|
| 1.269.6.2 | 03-Apr-2008 |
mjf | Sync with HEAD.
|
| 1.269.6.1 | 29-Mar-2008 |
mjf | - etc/devfsd.conf: Add some rules to give nodes like /dev/tty and
/dev/null better default modes, i.e. 0666.
- sbin/init: Run devfsd -s before going to multiuser.
- sys/arch: Provide arm32, i386, sparc with a mem_init() function to request
device nodes for /dev/null, /dev/zero, etc.
- sys/dev: Convert rnd, wd, agp, raid, cd, sd, wsdisplay, wskbd, wsmouse,
wsmux, tty, bpf, swap to devfs New World Order.
- sys/fs/devfs: Make the visibility attribute of device nodes configurable.
Also provide a function to mount a devfs on boot.
- sys/kern: Add a new boot flag, -n. This disables devfs support. Unless
the -n flag is specified the kernel will mount a devfs file
system on boot.
|
| 1.269.2.1 | 24-Mar-2008 |
keiichi | sync with head.
|
| 1.272.4.8 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.272.4.7 | 11-Mar-2010 |
yamt | sync with head
|
| 1.272.4.6 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.272.4.5 | 18-Jul-2009 |
yamt | sync with head.
|
| 1.272.4.4 | 20-Jun-2009 |
yamt | sync with head
|
| 1.272.4.3 | 16-May-2009 |
yamt | sync with head
|
| 1.272.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.272.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.272.2.2 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.272.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.273.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.273.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.274.4.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.274.2.1 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.275.4.1 | 31-Oct-2012 |
riz | sys/dev/scsipi/sd.c patch
Allow disk strategy to be queried and changed at runtime
for sd(4) disks.
[buhrow, ticket #1807]
|
| 1.275.2.2 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.275.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.277.2.2 | 23-Jul-2009 |
jym | Sync with HEAD.
|
| 1.277.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.291.2.1 | 30-Apr-2010 |
uebayasi | Sync with HEAD.
|
| 1.292.2.1 | 30-May-2010 |
rmind | sync with head
|
| 1.294.2.3 | 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")
|
| 1.294.2.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.294.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.295.2.2 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.295.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.296.2.1 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.298.2.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.298.2.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.298.2.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.298.2.1 | 12-Sep-2012 |
tls | Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for
physio (I/O to raw devices); needs more doing to get it going with the
filesystems, but it shouldn't damage data.
All work's been done on amd64 so far. Not hard to add support to other
ports. If others want to pitch in, one very helpful thing would be to
sort out when and how IDE disks can do 128K or larger transfers, and
adjust the various PCI IDE (or at least ahcisata) drivers and wd.c
accordingly -- it would make testing much easier. Another very helpful
thing would be to implement a smart minphys() for RAIDframe along the
lines detailed in the MAXPHYS-NOTES file.
|
| 1.301.2.1 | 18-May-2014 |
rmind | sync with head
|
| 1.305.2.2 | 10-Aug-2014 |
tls | Rebase.
|
| 1.305.2.1 | 07-Apr-2014 |
tls | Be a little more clear and consistent about harvesting entropy from devices:
1) deprecate RND_FLAG_NO_ESTIMATE
2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE
3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE
4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME|
RND_FLAG_COLLECT_VALUE|RND_FLAG_ESTIMATE_TIME
5) Make entropy harvesting from environmental sensors a little more generic
and remove it from individual sensor drivers.
6) Remove individual open-coded delta-estimators for values from a few
places in the tree (uvm, environmental drivers).
7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers
that had stubbed out code, other minor cleanups.
|
| 1.308.2.1 | 11-Nov-2014 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #201):
sbin/dkctl/dkctl.8: revision 1.24
sbin/dkctl/dkctl.8: revision 1.25
sys/dev/scsipi/sd.c: revision 1.310
sys/dev/ata/wd.c: revision 1.415
sbin/dkctl/dkctl.c: revision 1.21
sys/dev/raidframe/rf_netbsdkintf.c: revision 1.315
sys/dev/ld.c: revision 1.78
sys/dev/vnd.c: revision 1.234
sys/dev/dksubr.c: revision 1.54
sys/sys/dkio.h: revision 1.20
sys/dev/dkwedge/dk.c: revision 1.74
Add ioctl to autodiscover wedges.
Implement DIOCMWEDGES ioctl that triggers wedge autodiscovery.
Also fix a reference counting bug and clean up some code.
support DIOCMWEDGES ioctl.
Add 'makewedges' option to autodiscover wedges from a changed label.
New sentence, new line. Bump date for previous.
|
| 1.310.2.6 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.310.2.5 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.310.2.4 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.310.2.3 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.310.2.2 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.310.2.1 | 06-Apr-2015 |
skrll | Sync with HEAD
|
| 1.317.2.5 | 26-Apr-2017 |
pgoyette | Sync with HEAD
|
| 1.317.2.4 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.317.2.3 | 26-Jul-2016 |
pgoyette | Rename LOCALCOUNT_INITIALIZER to DEVSW_MODULE_INIT. This better describes
what we're doing, and why.
|
| 1.317.2.2 | 19-Jul-2016 |
pgoyette | Instead of repeatedly typing the conditional initialization of the
.d_localcount members in the various {b,c}devsw, define an initializer
macro and use it. This also removes the need for defining new symbols
for each 'struct localcount'.
As suggested by riastradh@
|
| 1.317.2.1 | 18-Jul-2016 |
pgoyette | Rump drivers are always installed via devsw_attach() so we need to
always allocate a 'struct localcount' for these drivers whenever they
are built as modules.
|
| 1.322.2.1 | 21-Apr-2017 |
bouyer | Sync with HEAD
|
| 1.324.6.2 | 11-Feb-2021 |
martin | Pull up following revision(s) (requested by ryoon in ticket #1652):
sys/dev/scsipi/cd.c: revision 1.350
sys/dev/scsipi/sd.c: revision 1.331
PR 55986: Ryo Onodera: DK_BUSY must have mask as second argument. Make cd.c
consistent by also using __BIT()
|
| 1.324.6.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.324.4.1 | 27-Apr-2017 |
pgoyette | Restore all work from the former pgoyette-localcount branch (which is
now abandoned doe to cvs merge botch).
The branch now builds, and installs via anita. There are still some
problems (cgd is non-functional and all atf tests time-out) but they
will get resolved soon.
|
| 1.325.6.3 | 21-Apr-2020 |
martin | Sync with HEAD
|
| 1.325.6.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.325.6.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.325.4.1 | 20-Oct-2018 |
pgoyette | Sync with head
|
| 1.327.4.3 | 29-Aug-2022 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1509):
sys/dev/scsipi/sdvar.h: revision 1.40
sys/dev/scsipi/sd.c: revision 1.335
Don't fetch data beyond end of inquiry buffer, which, here, is not
NUL-terminated.
Reduce target buffer to needed size (product name + NUL terminator).
|
| 1.327.4.2 | 11-Feb-2021 |
martin | Pull up following revision(s) (requested by ryoon in ticket #1203):
sys/dev/scsipi/cd.c: revision 1.350
sys/dev/scsipi/sd.c: revision 1.331
PR 55986: Ryo Onodera: DK_BUSY must have mask as second argument. Make cd.c
consistent by also using __BIT()
|
| 1.327.4.1 | 21-Mar-2020 |
martin | Pull up following revision(s) (requested by riastradh in ticket #795):
sys/dev/dksubr.c: revision 1.112
sys/arch/xen/xen/xbd_xenbus.c: revision 1.95
sys/dev/scsipi/sd.c: revision 1.328
sys/dev/dkvar.h: revision 1.32
sys/dev/ld.c: revision 1.109
sys/dev/cgd.c: revision 1.120
sys/dev/raidframe/rf_netbsdkintf.c: revision 1.380
sys/dev/ata/wd.c: revision 1.458
Add a flag to dk_dump for virtual disk devices.
If a disk is backed by a physical medium other than itself, such as
cgd(4), then it passes DK_DUMP_RECURSIVE to disable the recursion
detection for dk_dump.
If, however, a device represents a physical medium on its own, such
as wd(4), then it passes 0 instead.
With this, I can now dump to dk on cgd on dk on wd.
|
| 1.328.4.1 | 20-Apr-2020 |
bouyer | Sync with HEAD
|
| 1.330.2.1 | 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.331.6.1 | 31-May-2021 |
cjep | sync with head
|
| 1.331.4.1 | 17-Jun-2021 |
thorpej | Sync w/ HEAD.
|
| 1.335.4.1 | 12-Sep-2024 |
martin | Pull up following revision(s) (requested by rin in ticket #838):
sys/dev/scsipi/sd.c: revision 1.336
Don't try to discover wedges when the unit isn't online.
|
| 1.336.2.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.21 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.20 | 17-Sep-2003 |
mycroft | Remove PQUIRK_BYTE5_ZERO.
|
| 1.19 | 08-Sep-2003 |
mycroft | Use the const shaker.
|
| 1.18 | 03-Apr-2003 |
fvdl | branches: 1.18.2;
Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.17 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.16 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.15 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.14 | 09-Jan-2002 |
drochner | -don't apply scsipi_strvis() twice, it tries to octal-print the string
terminating zero and uninitialized dada following
-use use the "vendor" field for the name which is used in the default
disklabel, the "product" field is meaningless or not existing for ATAPI
(serial number)
|
| 1.13 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.12 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.11 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.10 | 14-May-2001 |
bouyer | branches: 1.10.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.9 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.8 | 20-Mar-2001 |
augustss | Extend quirks to 32 bits. Add two new ATAPI quirks, one to
deal with a bad byte5 of the capacity descriptor, one to
handle devices that lack the flex geometry page.
|
| 1.7 | 09-Jun-2000 |
enami | branches: 1.7.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.6 | 29-Mar-2000 |
simonb | branches: 1.6.2;
Don't need to include <sys/conf.h> here.
|
| 1.5 | 30-Sep-1999 |
thorpej | branches: 1.5.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.4 | 11-Sep-1999 |
thorpej | Implement detaching SCSI and ATAPI disks.
|
| 1.3 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.2 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.1 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.5.2.4 | 27-Mar-2001 |
bouyer | Sync with HEAD.
|
| 1.5.2.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.5.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.5.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.6.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.7.4.8 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.7.4.7 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.7.4.6 | 11-Jan-2002 |
nathanw | More catchup.
|
| 1.7.4.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.7.4.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.7.4.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.7.4.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.7.4.1 | 09-Apr-2001 |
nathanw | Catch up with -current.
|
| 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 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.10.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.18.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.37 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.36 | 17-Sep-2003 |
mycroft | If READ CAPACITY fails, try a READ FORMAT CAPACITIES. Separate this into
another function, always doing a page 0 MODE SENSE to get the block
descriptor if we use READ CAPACITY, and use SMS_DBD on the page 4/5 MODE
SENSE. This does one extra command in some cases, but it separates and
simplifies the code a little.
Why do we prefer READ CAPACITY over READ FORMAT CAPACITIES? Two reasons:
1) It's much older and is much less likely to have had its command code
abused, and is thus "safer" to try first. 2) ALL of my USB flash readers
and pen drives screw up their capacity descriptors -- mostly off-by-one
errors in the size (they return the maximum LBA number instead, a la READ
CAPACITY, which has *never* been how READ FORMAT CAPACITIES was documented
in the MMC spec), and one returns the "no media" code on slots that have
media inserted (despite returning almost-correct data otherwise)!
F*** me with a chainsaw.
|
| 1.35 | 09-Sep-2003 |
mycroft | Make similar changes as with cd_mode_{sense,select}().
|
| 1.34 | 08-Sep-2003 |
mycroft | Rework sd_scsibus_get_parms() a little. Try to get a block descriptor even
if our other MODE SENSEs fail. Use this code for the optical device case,
at least for now. (We could query the optical media type and do a table
lookup for the geometry, but why bother? Actually, why bother with geometry
at all, but I digress...)
|
| 1.33 | 08-Sep-2003 |
mycroft | RBC devices do in fact requre using 6-byte MODE SENSE/SELECT. This is totally
the wrong way to deal with that, but... I'll get back to it later.
|
| 1.32 | 08-Sep-2003 |
mycroft | Use the const shaker.
|
| 1.31 | 07-Sep-2003 |
mycroft | Set PQUIRK_ONLYBIG in the wdc-atapi frontend, obviating the need to ever test
the "bus type" for this.
Merge all the code in the SCSI and ATAPI backends for "cd" devices. All of
the mode page handling and whatnot is general to SCSI MMC devices, and should
never have been separated to begin with. This fixes a variety of problems,
and adds load/unload support for SCSI-attached devices.
|
| 1.30 | 05-Sep-2003 |
mycroft | In addition to skipping a variable-length list of block descriptors, we also
have to adjust for whether he have a "small" or "large" mode sense response.
sd_scsibus_getcache(), sd_scsibus_setache(): Set SMS_DBD so that we don't
have to deal with block descriptors here.
sd_scsibus_mode_sense(): Add a new parameter, "int *big", which tells the
caller whether it's a small or large header.
sd_scsibus_get_parms(), sd_scsibus_getcache(), sd_scsibus_setcache(): Use
the "big" value to find the data following the mode sense header correctly.
sd_scsibus_mode_sense(): Disable PQUIRK_NOBIGMODESENSE for now, as the
reasons for it were dubious and most likely fixed now.
sd_scsibus_get_parms(): Now that we're actually locating the right data,
check the returned page code to see if it's correct before using the mode
page info.
sd_scsibus_get_parms(): Get the rotational rate from the flex geometry page
too, if present.
XXX sd_scsibus_get_{opt,simplifiedparms}() need some work.
|
| 1.29 | 05-Sep-2003 |
mycroft | sd_scsibus_get_parms():
* Use the "block descriptor length" to find the beginning of the mode page
data. This is critical, since some devices do not return a block descriptor
at all. (XXX There is disgusting structure hackery here.)
* Always use scsipi_size() (i.e. READ CAPACITY) to get the total disk size.
* Get the rotational rate from the rigid geometry page.
* Don't read the block size from the block descriptor if there was none.
* Get the block size from the flexible geometry page if possible.
* If we don't like the way page 4 smells, try page 5, rather than just punting.
|
| 1.28 | 03-Apr-2003 |
fvdl | branches: 1.28.2;
Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.27 | 07-Mar-2003 |
drochner | Add support for "simplified direct" (aka RBC) devices.
This reveals some ugly points in the sd driver, as
the PQUIRK_ONLYBIG flag (there are 6, 10, 12 and 16 byte
CDBs for r/w) and the handling of small/big mode
select/sense. This should be cleaned up.
Also the "long" struct disk_parms.disksize is worth
a thought...
|
| 1.26 | 04-Oct-2002 |
soren | Pass on indication of absent media in the general case as in the optical
device case.
|
| 1.25 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.24 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.23 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.22 | 09-Jan-2002 |
thorpej | Add ioctls to get (DIOCGCACHE) and set (DIOCSCACHE) cache enables
on disks in a generic way. Implement these ioctls for SCSI disks.
This is not fully fleshed-out yet, but it allows people to experiment
with disk caches more easily.
|
| 1.21 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.20 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.19 | 31-Aug-2001 |
augustss | Add a new quirk, PQUIRK_NOBIGMODESENSE, for devices that need big
SCSI ops, but still fails on big mode sense.
Makes M-Sys DiskOnKey work.
|
| 1.18 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.17 | 14-May-2001 |
bouyer | branches: 1.17.2;
Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.16 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.15 | 09-Jun-2000 |
enami | branches: 1.15.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.14 | 31-May-2000 |
augustss | Add a method, scsipi_getgeom, to the adapter struct. If this method is
present it will be used to get the drive geometry if the MODE SENSE fails.
|
| 1.13 | 30-May-2000 |
augustss | Add a quirk, SDEV_ONLYBIG, which implies that the device cannot handle
the 6 byte versions of READ, WRITE, and MODE_SENSE.
This greatly simplifies the UFI (USB Floppy) handling.
|
| 1.12 | 29-Mar-2000 |
simonb | branches: 1.12.2;
Don't need to include <sys/conf.h> here.
|
| 1.11 | 30-Sep-1999 |
thorpej | branches: 1.11.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.10 | 11-Sep-1999 |
thorpej | Implement detaching SCSI and ATAPI disks.
|
| 1.9 | 26-Aug-1999 |
hannken | Make `sc_ops->sdo_flush' synchronous. The flags `SDF_FLUSHING' and
`SDF_DIRTY' were never reset because `sddone' doesn't get called from
synchronous scsi commands.
|
| 1.8 | 08-Oct-1998 |
thorpej | Handle devices that have the SDEV_NOSYNCCACHE quirk.
|
| 1.7 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.6 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.5 | 15-Aug-1998 |
thorpej | Keep track if the disk has had done i/o, making it dirty. On last close,
or at shutdown, flush the cache if the disk is still dirty and it support
cache flushing. Fixes PRs #5588 (Jim Bernard) and #5589 (Justin Gibbs).
|
| 1.4 | 12-Aug-1998 |
thorpej | Run the shutdown hook w/ SCSI_AUTOCONF.
|
| 1.3 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.2 | 10-Jun-1998 |
thorpej | Add a shutdown hook for SCSI-2 and higher disks that issues a SYNCHRONIZE CACHE
operation with address 0 length 0, which, according to the SCSI-2 spec, should
be interpreted as "synchronize all remaining blocks beginning at address 0".
|
| 1.1 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.11.2.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.11.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.11.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.12.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.15.4.8 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.15.4.7 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.15.4.6 | 11-Jan-2002 |
nathanw | More catchup.
|
| 1.15.4.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.15.4.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.15.4.3 | 21-Sep-2001 |
nathanw | Catch up to -current.
|
| 1.15.4.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.15.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.17.2.4 | 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
|
| 1.17.2.3 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.17.2.2 | 13-Sep-2001 |
thorpej | Update the kqueue branch to HEAD.
|
| 1.17.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.28.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.42 | 27-Feb-2025 |
jakllsch | sd(4): Translate discard to UNMAP (10) command
|
| 1.41 | 27-Feb-2025 |
jakllsch | Implement DIOCGSECTORALIGN for sd(4).
Use only READ CAPACITY 16 if peripheral claims conformance to SPC-3.
|
| 1.40 | 28-Aug-2022 |
mlelstv | branches: 1.40.10;
Don't fetch data beyond end of inquiry buffer, which, here, is not
NUL-terminated.
Reduce target buffer to needed size (product name + NUL terminator).
|
| 1.39 | 19-Mar-2019 |
mlelstv | branches: 1.39.4;
Set disk product name as disk type. The information can be queried through
drvctl(4).
|
| 1.38 | 10-Dec-2016 |
mlelstv | branches: 1.38.16;
Refactored sd and cd to use common disk subroutines.
|
| 1.37 | 24-Aug-2015 |
pooka | branches: 1.37.2;
would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.36 | 13-Apr-2015 |
riastradh | Convert sys/dev to use <sys/rndsource.h>.
|
| 1.35 | 05-Sep-2014 |
matt | branches: 1.35.2;
Don't nest structure definitions.
|
| 1.34 | 02-Feb-2012 |
tls | branches: 1.34.6;
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.
|
| 1.33 | 19-Nov-2011 |
tls | branches: 1.33.2;
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.
|
| 1.32 | 10-Apr-2009 |
dyoung | branches: 1.32.12;
Use a PMF shutdown hook instead of shutdownhook_establish(). This ought
to stop the crash during shutdown that yamt@ reports in kern/41182.
|
| 1.31 | 16-Jul-2008 |
drochner | branches: 1.31.2; 1.31.8;
split device/softc for sd (tested with a USB stick)
|
| 1.30 | 28-Apr-2008 |
martin | branches: 1.30.2; 1.30.4; 1.30.6;
Remove clause 3 and 4 from TNF licenses
|
| 1.29 | 09-Jul-2007 |
ad | branches: 1.29.28; 1.29.30; 1.29.32;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.28 | 11-Dec-2005 |
christos | branches: 1.28.30; 1.28.32;
merge ktrace-lwp.
|
| 1.27 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.26 | 01-Feb-2005 |
reinoud | branches: 1.26.6;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.25 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.24 | 01-Oct-2004 |
thorpej | branches: 1.24.4;
Don't need the lock in the softc anymore.
|
| 1.23 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.22 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.21 | 24-Feb-2004 |
wiz | branches: 1.21.2;
parameter with two es. From Peter Postma.
|
| 1.20 | 18-Sep-2003 |
mycroft | Merge the geometry and cache handling code for all direct access and optical
devices, as it's general to all SCSI MMC devices. In the process, remove
PQUIRK_NO_FLEX_PAGE.
|
| 1.19 | 03-Apr-2003 |
fvdl | branches: 1.19.2;
Check RAW_PART against the media size instead of the disklabel.
Add the media size in 512-byte sectors to the softc, to avoid
some 64 bit computations. Bump the capacity stored in softcs
for disks to 64 bits.
|
| 1.18 | 06-Jan-2003 |
matt | Add multiple inclusion protection.
|
| 1.17 | 16-Jul-2002 |
hannken | Implement a new device buffer queue interface.
One basic struct, a function to setup a queue with a specific strategy and
three macros to put buf's into the queue, get and remove the next buf or
get the next buf without removal.
The BUFQ_XXX interface will be removed in the future.
The B_ORDERED flag is not longer supported.
Approved by: Jason R. Thorpe <thorpej@wasabisystems.com>
|
| 1.16 | 09-Jan-2002 |
thorpej | branches: 1.16.8;
Add ioctls to get (DIOCGCACHE) and set (DIOCSCACHE) cache enables
on disks in a generic way. Implement these ioctls for SCSI disks.
This is not fully fleshed-out yet, but it allows people to experiment
with disk caches more easily.
|
| 1.15 | 23-May-2001 |
chs | branches: 1.15.2;
fix typo in previous revision so that the scsi code compiles again.
|
| 1.14 | 22-May-2001 |
mjacob | Make SDRETRIES an option you can dig out of opt_scsi.h.
Add SD_IO_TIMEOUT (an option in opt_scsi.h) also, which defaults to the
normal 60 * 1000 timeout for normal read/write commamds. This allows you
to shorten or lengthen the timeout as needed.
|
| 1.13 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.12 | 23-May-2000 |
bouyer | branches: 1.12.6;
sd_interpret_sense() can be called from interrupt context; call scsipi_start()
with XS_CTL_ASYNC if the failed command was called with XS_CTL_ASYNC.
Add a SDF_RESTART flag to keep state, cleared in sddone().
A mounted disk can now spin down, it will propely spin up at the next access.
|
| 1.11 | 22-Jan-2000 |
drochner | include "rnd.h" explicitely since the size of sd_softc depends on it
(sd_scsi.c and sd_atapi.c missed it, leading to random memory corruption)
|
| 1.10 | 21-Jan-2000 |
thorpej | Update for sys/buf.h/disksort_*() changes.
|
| 1.9 | 11-Sep-1999 |
thorpej | branches: 1.9.2;
Implement detaching SCSI and ATAPI disks.
|
| 1.8 | 26-Aug-1999 |
hannken | Make `sc_ops->sdo_flush' synchronous. The flags `SDF_FLUSHING' and
`SDF_DIRTY' were never reset because `sddone' doesn't get called from
synchronous scsi commands.
|
| 1.7 | 17-Aug-1998 |
mycroft | branches: 1.7.6;
Assign my copyrights to TNF.
|
| 1.6 | 15-Aug-1998 |
thorpej | Keep track if the disk has had done i/o, making it dirty. On last close,
or at shutdown, flush the cache if the disk is still dirty and it support
cache flushing. Fixes PRs #5588 (Jim Bernard) and #5589 (Justin Gibbs).
|
| 1.5 | 05-Aug-1998 |
drochner | Improve generation of default disklabels:
-store printable product ID in cd's and sd's softc, use it as "typename"
-for this, add a "destination buffer length" argument to scsipi_strvis()
-return ATAPI device type for ATAPI devices
|
| 1.4 | 10-Jun-1998 |
thorpej | Add a shutdown hook for SCSI-2 and higher disks that issues a SYNCHRONIZE CACHE
operation with address 0 length 0, which, according to the SCSI-2 spec, should
be interpreted as "synchronize all remaining blocks beginning at address 0".
|
| 1.3 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.2 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.1 | 15-Jan-1998 |
cgd | add support for an ATAPI attachment for 'sd'.
fix 'cd' driver's NCD_SCSI bogosity (was using testing wrong macro!)
clean up in various ways:
* make common atapi_mode_{sense,select}() functions.
* put ATAPI data structures in more sensible headers, split up by
device type.
* include headers a bit more carefully.
* pass flags to attachment-specific cd functions, and use them.
* get rid of SCSI bits in scsipi_base.h's scsipi_make_xs(), move
them into the correct place in scsi_base.c.
* fix minor typo in struct name in scsipiconf.h (which was apparently
never used except in a #define later in the same file).
* use __attribute__ to force 4-byte alignment for xs command store,
so that architectures trying to bus_space_write_multi_N() (where
N > 1) that data to a controller won't lose.
* clean up a few comments in typos, and make a few #defines easier to
understand/maintain.
* rename cd_link.h to cdvar.h (via repository copy). This is exactly
what a 'var' file is supposed to be.
|
| 1.7.6.1 | 23-Jan-2000 |
he | Pull up revision 1.11 (requested by drochner):
Fix a size inconsistency of "struct sd_softc" if the "random"
pseudo-device is enabled, leading to data corruption.
|
| 1.9.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.9.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.12.6.5 | 07-Jan-2003 |
thorpej | Sync with HEAD.
|
| 1.12.6.4 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.12.6.3 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.12.6.2 | 11-Jan-2002 |
nathanw | More catchup.
|
| 1.12.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.15.2.2 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.15.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.16.8.1 | 20-Jul-2002 |
gehenna | catch up with -current.
|
| 1.19.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.19.2.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.19.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.19.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.19.2.2 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.19.2.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.21.2.1 | 11-Sep-2004 |
he | Pull up revision 1.22 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.24.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.26.6.2 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.26.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.28.32.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.28.30.1 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.29.32.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.29.32.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.29.30.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.29.28.2 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.29.28.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.30.6.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.30.4.1 | 18-Jul-2008 |
simonb | Sync with head.
|
| 1.30.2.1 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.31.8.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.31.2.1 | 28-Apr-2009 |
skrll | Sync with HEAD.
|
| 1.32.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.33.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.34.6.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.35.2.3 | 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.35.2.2 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.35.2.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.37.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.38.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.39.4.1 | 29-Aug-2022 |
martin | Pull up following revision(s) (requested by mlelstv in ticket #1509):
sys/dev/scsipi/sdvar.h: revision 1.40
sys/dev/scsipi/sd.c: revision 1.335
Don't fetch data beyond end of inquiry buffer, which, here, is not
NUL-terminated.
Reduce target buffer to needed size (product name + NUL terminator).
|
| 1.40.10.1 | 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.52 | 09-Sep-2021 |
riastradh | sys/dev: Memset zero before copyout.
Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.
I think the iwi(4), mcd(4), and ses(4) changes actually plug leaks;
the raidframe(4) change probably doesn't (but doesn't hurt).
|
| 1.51 | 08-Mar-2019 |
msaitoh | branches: 1.51.4;
s/ are are / are /
s/ a a / a /
|
| 1.50 | 20-Nov-2016 |
mlelstv | branches: 1.50.8; 1.50.16;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.49 | 14-Jul-2016 |
msaitoh | branches: 1.49.2;
- Use aprint*() more in xxx_attach().
- Add missing aprint_naive("\n").
- Remove extra spaces and tabs.
- KNF.
|
| 1.48 | 24-Aug-2015 |
pooka | would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.47 | 25-Jul-2014 |
dholland | branches: 1.47.4;
Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.46 | 16-Mar-2014 |
dholland | branches: 1.46.2;
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.
|
| 1.45 | 07-Apr-2013 |
wiz | branches: 1.45.4;
Add detach support. From martin@.
"Works for me."
Addresses PR 44283.
|
| 1.44 | 27-Oct-2012 |
chs | split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.
|
| 1.43 | 17-Jul-2011 |
joerg | branches: 1.43.2; 1.43.8; 1.43.12; 1.43.14;
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.
|
| 1.42 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.41 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.40 | 08-Jun-2008 |
tsutsui | branches: 1.40.12;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.39 | 05-Apr-2008 |
cegger | branches: 1.39.2; 1.39.4; 1.39.6;
use aprint_*_dev and device_xname
|
| 1.38 | 04-Mar-2007 |
christos | branches: 1.38.36;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.37 | 16-Nov-2006 |
christos | branches: 1.37.4;
__unused removal on arguments; approved by core.
|
| 1.36 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.35 | 03-Sep-2006 |
christos | branches: 1.35.2; 1.35.4;
add missing initializer
|
| 1.34 | 14-Apr-2006 |
christos | Coverity CID 1100: Avoid NULL pointer dereferences.
|
| 1.33 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.32 | 11-Dec-2005 |
christos | branches: 1.32.4; 1.32.6; 1.32.8; 1.32.10; 1.32.12;
merge ktrace-lwp.
|
| 1.31 | 30-May-2005 |
christos | branches: 1.31.2;
- remove bogus casts
- add more const
|
| 1.30 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.29 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.28 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.27 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.26 | 18-Sep-2004 |
mycroft | branches: 1.26.4; 1.26.6;
Minor rearrangement. Whitespace and #include cleanup.
|
| 1.25 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.24 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.23 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.22 | 29-Jun-2003 |
fvdl | branches: 1.22.2; 1.22.4;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.21 | 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
|
| 1.20 | 07-Feb-2003 |
mjacob | kern/20231- pass the correct argument (arg_addr) on thru to scsipi_do_ioctl,
not the locally dereferenced fetch.
|
| 1.19 | 20-Jan-2003 |
simonb | The Double-Semi-Colon Police.
|
| 1.18 | 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
|
| 1.17 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.16 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.15 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.14 | 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.
|
| 1.13 | 15-Nov-2001 |
lukem | branches: 1.13.8; 1.13.10;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.12 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.11 | 18-Jul-2001 |
thorpej | branches: 1.11.2;
bcopy -> memcpy
|
| 1.10 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.9 | 25-Apr-2001 |
bouyer | branches: 1.9.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.8 | 08-Aug-2000 |
mjacob | branches: 1.8.2; 1.8.4;
Don't attempt to actually read SAF-TE temperature objects- nobody seems
to be obeying the original spec as to what the numeric value means.
Temperature flags are unaffected- these are still the 'pseudo-thermometers'
and overtemp/undertemp warnings will be caught and translated to SES objects
here.
|
| 1.7 | 08-Jul-2000 |
sommerfeld | printf format paranoia
|
| 1.6 | 22-May-2000 |
thorpej | Remove an used variable and add a missing `}' so that this compiles.
|
| 1.5 | 14-May-2000 |
dante | Change scsipi_inquiry_data strucure to be ANSI SPC-2 rev16 compliant
|
| 1.4 | 20-Feb-2000 |
mjacob | Oops...I forgot that the GEM-2 chip is so lame that if you don't tell
it exactly the right amount of data to transfer for WRITE BUFFER/Global
it will stay stuck in command phase (causing a command phase overrun).
|
| 1.3 | 21-Jan-2000 |
mjacob | slight change for platform independence
|
| 1.2 | 21-Jan-2000 |
mjacob | cleanup SAF-TE attachment code
|
| 1.1 | 20-Jan-2000 |
mjacob | add SES driver
|
| 1.8.4.7 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.8.4.6 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.8.4.5 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.8.4.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.8.4.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.8.4.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.8.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.8.2.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.8.2.1 | 08-Aug-2000 |
bouyer | file ses.c was added on branch thorpej_scsipi on 2000-11-20 09:59:28 +0000
|
| 1.9.2.3 | 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
|
| 1.9.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.9.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.11.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.11.2.1 | 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).
|
| 1.13.10.2 | 16-Jun-2003 |
grant | Pull up revision 1.20 (requested by mjacob in ticket #1162):
kern/20231- pass the correct argument (arg_addr) on thru to scsipi_do_ioctl,
not the locally dereferenced fetch.
|
| 1.13.10.1 | 16-Jun-2003 |
grant | Pull up revision 1.19 (requested by mjacob in ticket #1162):
The Double-Semi-Colon Police.
|
| 1.13.8.1 | 16-May-2002 |
gehenna | Add the character device switch.
|
| 1.22.4.1 | 11-Sep-2004 |
he | Pull up revision 1.24 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.22.2.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.22.2.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.22.2.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.22.2.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.22.2.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.22.2.1 | 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.
|
| 1.26.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.26.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.31.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.31.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.31.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.32.12.2 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.32.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.32.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.32.8.3 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.32.8.2 | 24-May-2006 |
yamt | sync with head.
|
| 1.32.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.32.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.32.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.35.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.35.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.35.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.37.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.38.36.3 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.38.36.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.38.36.1 | 05-Apr-2008 |
mjf | - add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing
as these are always needed.
- convert many, many drivers over to the New Devfs World Order. For a
list of device drivers yet to be converted see,
http://www.netbsd.org/~mjf/devfs-todo.html.
- add a new device_unregister_all(device_t) function to remove all device
names associated with a device_t, which saves us having to construct
device names when the driver is detached.
- add a DEV_AUDIO type for devices.
|
| 1.39.6.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.39.4.2 | 16-May-2009 |
yamt | sync with head
|
| 1.39.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.39.2.1 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.40.12.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.43.14.1 | 29-Apr-2013 |
riz | Pull up following revision(s) (requested by jakllsch in ticket #878):
sys/dev/scsipi/ses.c: revision 1.45
Add detach support. From martin@.
"Works for me."
Addresses PR 44283.
|
| 1.43.12.4 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.43.12.3 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.43.12.2 | 23-Jun-2013 |
tls | resync from head
|
| 1.43.12.1 | 20-Nov-2012 |
tls | Resync to 2012-11-19 00:00:00 UTC
|
| 1.43.8.1 | 29-Apr-2013 |
riz | Pull up following revision(s) (requested by jakllsch in ticket #878):
sys/dev/scsipi/ses.c: revision 1.45
Add detach support. From martin@.
"Works for me."
Addresses PR 44283.
|
| 1.43.2.2 | 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")
|
| 1.43.2.1 | 30-Oct-2012 |
yamt | sync with head
|
| 1.45.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.46.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.47.4.3 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.47.4.2 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.47.4.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.49.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.50.16.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.50.8.1 | 21-Jun-2023 |
martin | Pull up following revision(s) (requested by riastradh in ticket #1835):
sys/dev/pci/if_iwi.c: revision 1.117
sys/dev/raidframe/rf_netbsdkintf.c: revision 1.401
sys/dev/scsipi/ses.c: revision 1.52
sys/dev/isa/mcd.c: revision 1.121
(all via patch)
sys/dev: Memset zero before copyout.
Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.
I think the iwi(4), mcd(4), and ses(4) changes actually plug leaks;
the raidframe(4) change probably doesn't (but doesn't hurt).
|
| 1.51.4.1 | 03-Aug-2022 |
martin | Pull up following revision(s) (requested by riastradh in ticket #1485):
sys/dev/pci/if_iwi.c: revision 1.117
sys/dev/raidframe/rf_netbsdkintf.c: revision 1.401
sys/dev/scsipi/ses.c: revision 1.52
sys/dev/isa/mcd.c: revision 1.121
sys/dev: Memset zero before copyout.
Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.
I think the iwi(4), mcd(4), and ses(4) changes actually plug leaks;
the raidframe(4) change probably doesn't (but doesn't hurt).
|
| 1.4 | 06-Sep-2015 |
dholland | More on PR 41200: headers that declare ioctls should include sys/ioccom.h.
This covers (I think) all the MI headers outside of external/ (and dist/).
|
| 1.3 | 11-Dec-2005 |
christos | branches: 1.3.120; 1.3.140;
merge ktrace-lwp.
|
| 1.2 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.1 | 20-Jan-2000 |
mjacob | branches: 1.1.6; 1.1.30; 1.1.38; 1.1.40;
add SES driver
|
| 1.1.40.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.1.38.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.1.30.1 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.1.6.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.1.6.1 | 20-Jan-2000 |
bouyer | file ses.h was added on branch thorpej_scsipi on 2000-11-20 09:59:28 +0000
|
| 1.3.140.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.3.120.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.91 | 23-Feb-2022 |
andvar | fix various typos in comments, mainly immediatly/immediately/,
as well shared and recently fixed typos in OpenBSD code by Jonathan Grey.
|
| 1.90 | 29-Sep-2020 |
msaitoh | s/occurence/occurrence/
|
| 1.89 | 17-Jun-2017 |
mlelstv | The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.88 | 20-Nov-2016 |
mlelstv | branches: 1.88.8;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.87 | 20-Nov-2016 |
pgoyette | Avoid calling bufq_free() from critical code sections.
|
| 1.86 | 25-Jul-2014 |
dholland | branches: 1.86.4; 1.86.8;
Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.85 | 16-Mar-2014 |
dholland | branches: 1.85.2;
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.
|
| 1.84 | 28-Feb-2012 |
mbalmer | branches: 1.84.2; 1.84.4;
one line comments on one line.
|
| 1.83 | 28-Feb-2012 |
mbalmer | Unbreak the build.
|
| 1.82 | 28-Feb-2012 |
mbalmer | Remove a pointless #if 0, compiler will eliminate the code anways.
Noticed by skrll.
|
| 1.81 | 28-Feb-2012 |
mbalmer | Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.80 | 06-Dec-2009 |
dyoung | branches: 1.80.12; 1.80.16; 1.80.18;
Delete do-nothing device-activation hooks.
|
| 1.79 | 23-Nov-2009 |
rmind | Remove some unecessary includes sys/user.h header.
|
| 1.78 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.77 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.76 | 13-Jan-2009 |
yamt | branches: 1.76.2;
g/c BUFQ_FOO() macros and use bufq_foo() directly.
|
| 1.75 | 11-Jan-2009 |
cegger | make this compile
|
| 1.74 | 08-Jun-2008 |
tsutsui | branches: 1.74.6;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.73 | 05-Apr-2008 |
cegger | branches: 1.73.2; 1.73.4; 1.73.6;
use aprint_*_dev and device_xname
|
| 1.72 | 29-Jul-2007 |
ad | branches: 1.72.26; 1.72.28;
It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.71 | 09-Jul-2007 |
ad | branches: 1.71.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.70 | 04-Mar-2007 |
christos | branches: 1.70.2; 1.70.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.69 | 16-Nov-2006 |
christos | branches: 1.69.4;
__unused removal on arguments; approved by core.
|
| 1.68 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.67 | 03-Sep-2006 |
christos | branches: 1.67.2; 1.67.4;
add missing initializer
|
| 1.66 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.65 | 28-Mar-2006 |
thorpej | Use device_unit().
|
| 1.64 | 20-Feb-2006 |
thorpej | branches: 1.64.2; 1.64.4; 1.64.6;
Use device_is_active() rather than testing dv_flags for DVF_ACTIVE
directly.
|
| 1.63 | 11-Dec-2005 |
christos | branches: 1.63.2; 1.63.4; 1.63.6;
merge ktrace-lwp.
|
| 1.62 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.61 | 30-May-2005 |
christos | branches: 1.61.2;
- remove bogus casts
- add more const
|
| 1.60 | 31-Mar-2005 |
yamt | introduce a function to drain bufq and use it where appropriate.
|
| 1.59 | 01-Feb-2005 |
reinoud | branches: 1.59.4;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.58 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.57 | 28-Oct-2004 |
yamt | branches: 1.57.4;
move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.56 | 17-Oct-2004 |
bouyer | Adjust some debug printfs after b_bcount was changed from long to int.
From Patrick Welche on current-users@
|
| 1.55 | 17-Sep-2004 |
mycroft | Oops; fix ssrestart().
|
| 1.54 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.53 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.52 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.51 | 14-Mar-2004 |
thorpej | branches: 1.51.2;
Use scsipi_adapter_minphys().
|
| 1.50 | 08-Sep-2003 |
mycroft | Use the const shaker.
|
| 1.49 | 29-Jun-2003 |
fvdl | branches: 1.49.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.48 | 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
|
| 1.47 | 20-Mar-2003 |
dbj | use PRId64 to printf bp->b_blkno, which is of type daddr_t
|
| 1.46 | 21-Jan-2003 |
simonb | Remove a line of unreachable code. It seems to be a remnant of starting
off with a different driver, and the way this code is structured the
offending line isn't needed.
|
| 1.45 | 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
|
| 1.44 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.43 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.42 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.41 | 14-Sep-2002 |
chs | recognize HP ScanJet 4c.
add a catch-all case for other HP scanners.
remove a local version of atoi() in favor of strtoul() from libkern.
|
| 1.40 | 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.
|
| 1.39 | 22-Jul-2002 |
hannken | Convert to new device buffer queue interface.
|
| 1.38 | 15-Nov-2001 |
lukem | branches: 1.38.8;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.37 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.36 | 18-Jul-2001 |
thorpej | branches: 1.36.2;
bcopy -> memcpy
|
| 1.35 | 26-Jun-2001 |
bouyer | branches: 1.35.2;
Make it compile with SCSIPI_DEBUG
|
| 1.34 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.33 | 01-Jan-2001 |
augustss | branches: 1.33.2;
Call vdevgone() with correct minor range.
|
| 1.32 | 01-Jan-2001 |
augustss | Support device detach.
Don't call scanjet_attach() for all HP scanners.
Also, ANSIfy while I'm here.
|
| 1.31 | 31-May-2000 |
phil | Add an entry for the HP ScanJet 6300C.
|
| 1.30 | 30-Mar-2000 |
augustss | branches: 1.30.2;
Get rid of register declarations.
|
| 1.29 | 21-Jan-2000 |
thorpej | Update for sys/buf.h/disksort_*() changes.
|
| 1.28 | 19-Jan-2000 |
abs | newline terminate ss: atatch message - from Roger Brooks
|
| 1.27 | 17-Jan-2000 |
abs | allow ioctl() on all scanner device nodes - now SANE can work with ss* driver
|
| 1.26 | 30-Sep-1999 |
thorpej | branches: 1.26.2; 1.26.8;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.25 | 10-Feb-1999 |
bouyer | branches: 1.25.2;
Abort transfer if b_blkno is negative. Closes PR kern/5553 by Johan Danielsson.
|
| 1.24 | 08-Dec-1998 |
thorpej | When closing, wait for pending xfers to drain before deleting the reference
to the adapter.
|
| 1.23 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.22 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.21 | 19-Aug-1998 |
explorer | ss->special could be NULL, so sane would crash the kernel.
|
| 1.20 | 22-Apr-1998 |
pk | Set a flag during auto-configuration, so SCSI_AUTOCONF can be passed
when doing additional I/O from the scanner attach functions.
|
| 1.19 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.18 | 19-Nov-1997 |
augustss | Recognize HP ScanJet 5p.
|
| 1.17 | 01-Oct-1997 |
enami | branches: 1.17.2;
Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.16 | 27-Aug-1997 |
bouyer | branches: 1.16.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.15 | 05-Dec-1996 |
cgd | update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.14 | 29-Nov-1996 |
thorpej | Add an inquiry pattern for the HP ScanJet 4p.
From Valtteri Vuorikoski <vuori@notcom.org>, PR #2957.
|
| 1.13 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.12 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.11 | 12-Jul-1996 |
is | Some MUSTEK scanners fill the vendor string in the SCSI identify
string with 0xFF instead of 0x20. Shorten the compare to 6 bytes to
take care of this.
Fixes PR #2617 by myself, which is based on a private problem report
by Neil Brewitt.
|
| 1.10 | 05-May-1996 |
christos | Cleanup the rest of the SCSIDEBUG printfs. From Bernd Ernesti.
|
| 1.9 | 30-Mar-1996 |
christos | - Eliminate scsi_conf.h
- Fix prototypes
- Fix PR/2248 [Problems with HP scanjet...]
|
| 1.8 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.7 | 05-Mar-1996 |
thorpej | Normalize SCSI autoconfiguration output.
From Chris Demetriou <cgd@NetBSD.ORG>. Fixes PR #1958.
|
| 1.6 | 19-Feb-1996 |
mycroft | Handle EOF a little differently.
|
| 1.5 | 18-Feb-1996 |
mycroft | Change variable name to avoid GCC warning.
|
| 1.4 | 18-Feb-1996 |
mycroft | Minor tweak; make sure window size is 0 on close.
|
| 1.3 | 18-Feb-1996 |
mycroft | Missed some prototyping foo.
|
| 1.2 | 18-Feb-1996 |
mycroft | Truncate to the window size in ssminphys(), not ssread().
|
| 1.1 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.16.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.16.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.16.2.1 | 27-Aug-1997 |
thorpej | file ss.c was added on branch marc-pcmcia on 1997-08-27 23:33:34 +0000
|
| 1.17.2.1 | 22-Nov-1997 |
mellon | Pull rev 1.18 up from trunk (augustss)
|
| 1.25.2.2 | 20-Jan-2000 |
he | Pull up revision 1.28 (requested by abs):
Newline terminate 'ss:' attach message.
|
| 1.25.2.1 | 17-Jan-2000 |
he | Pull up revision 1.27 (requested by abs):
Allow ioctl() on all scanner device nodes - allows SANE to work.
|
| 1.26.8.1 | 21-Dec-1999 |
wrstuden | Initial commit of recent changes to make DEV_BSIZE go away.
Runs on i386, needs work on other arch's. Main kernel routines should be
fine, but a number of the stand programs need help.
cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512
byte block devices. vnd, raidframe, and lfs need work.
Non 2**n block support is automatic for LKM's and conditional for kernels
on "options NON_PO2_BLOCKS".
|
| 1.26.2.6 | 21-Apr-2001 |
bouyer | Get rid of references to sc_link/scsipi_link in debug or comments.
|
| 1.26.2.5 | 06-Jan-2001 |
bouyer | Make compile.
|
| 1.26.2.4 | 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.26.2.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.26.2.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.26.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.30.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.33.2.8 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.33.2.7 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.6 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.5 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.33.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.33.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.33.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.35.2.4 | 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
|
| 1.35.2.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.35.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.35.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.36.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.36.2.1 | 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).
|
| 1.38.8.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.38.8.1 | 16-May-2002 |
gehenna | Add the character device switch.
Replace the direct-access to devsw table with calling devsw API.
|
| 1.49.2.10 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.49.2.9 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.49.2.8 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.49.2.7 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.49.2.6 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.49.2.5 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.49.2.4 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.49.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.49.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.49.2.1 | 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.
|
| 1.51.2.1 | 11-Sep-2004 |
he | Pull up revision 1.53 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.57.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.59.4.1 | 06-Apr-2005 |
tron | Pull up revision 1.60 (requested by yamt in ticket #112):
introduce a function to drain bufq and use it where appropriate.
|
| 1.61.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.61.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.61.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.63.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.63.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.63.2.1 | 01-Mar-2006 |
yamt | sync with head.
|
| 1.64.6.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.64.4.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.64.2.2 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.64.2.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.67.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.67.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.67.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.69.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.70.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.70.2.3 | 19-Aug-2007 |
ad | - Back out the biodone() changes.
- Eliminate B_ERROR (from HEAD).
|
| 1.70.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.70.2.1 | 13-May-2007 |
ad | - Pass the error number and residual count to biodone(), and let it handle
setting error indicators. Prepare to eliminate B_ERROR.
- Add a flag argument to brelse() to be set into the buf's flags, instead
of doing it directly. Typically used to set B_INVAL.
- Add a "struct cpu_info *" argument to kthread_create(), to be used to
create bound threads. Change "bool mpsafe" to "int flags".
- Allow exit of LWPs in the IDL state when (l != curlwp).
- More locking fixes & conversion to the new API.
|
| 1.71.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.72.28.2 | 29-Jul-2007 |
ad | It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.72.28.1 | 29-Jul-2007 |
ad | file ss.c was added on branch matt-mips64 on 2007-07-29 12:50:24 +0000
|
| 1.72.26.5 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.72.26.4 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.72.26.3 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.72.26.2 | 06-Apr-2008 |
mjf | - after some discussion with agc@ i agreed it would be a good idea to move
device_unregister_* to device_deregister_* to be more like the pmf(9)
functions, especially since a lot of the time the function calls are next
to each other.
- add device_register_name() support for dk(4).
|
| 1.72.26.1 | 05-Apr-2008 |
mjf | - add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing
as these are always needed.
- convert many, many drivers over to the New Devfs World Order. For a
list of device drivers yet to be converted see,
http://www.netbsd.org/~mjf/devfs-todo.html.
- add a new device_unregister_all(device_t) function to remove all device
names associated with a device_t, which saves us having to construct
device names when the driver is detached.
- add a DEV_AUDIO type for devices.
|
| 1.73.6.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.73.4.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.73.4.2 | 16-May-2009 |
yamt | sync with head
|
| 1.73.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.73.2.1 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.74.6.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.76.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.80.18.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #65):
sys/dev/scsipi/ss_scanjet.c: revision 1.52
sys/dev/scsipi/ss.c: revision 1.81
sys/dev/scsipi/ssvar.h: revision 1.19
sys/dev/scsipi/ss_mustek.c: revision 1.41
Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.80.16.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.80.16.2 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.80.16.1 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.80.12.2 | 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")
|
| 1.80.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.84.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.84.2.3 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.84.2.2 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.84.2.1 | 12-Sep-2012 |
tls | Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for
physio (I/O to raw devices); needs more doing to get it going with the
filesystems, but it shouldn't damage data.
All work's been done on amd64 so far. Not hard to add support to other
ports. If others want to pitch in, one very helpful thing would be to
sort out when and how IDE disks can do 128K or larger transfers, and
adjust the various PCI IDE (or at least ahcisata) drivers and wd.c
accordingly -- it would make testing much easier. Another very helpful
thing would be to implement a smart minphys() for RAIDframe along the
lines detailed in the MAXPHYS-NOTES file.
|
| 1.85.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.86.8.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.86.4.2 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.86.4.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.88.8.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.43 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.42 | 25-Oct-2013 |
martin | branches: 1.42.6; 1.42.10;
Mark a diagnostic-only variable
|
| 1.41 | 28-Feb-2012 |
mbalmer | branches: 1.41.2; 1.41.4;
Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.40 | 23-Nov-2009 |
rmind | branches: 1.40.12; 1.40.16; 1.40.18;
Remove some unecessary includes sys/user.h header.
|
| 1.39 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.38 | 13-Jan-2009 |
yamt | g/c BUFQ_FOO() macros and use bufq_foo() directly.
|
| 1.37 | 05-Apr-2008 |
cegger | branches: 1.37.4; 1.37.12;
use aprint_*_dev and device_xname
|
| 1.36 | 04-Mar-2007 |
christos | branches: 1.36.36;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.35 | 16-Nov-2006 |
christos | branches: 1.35.4;
__unused removal on arguments; approved by core.
|
| 1.34 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.33 | 03-Sep-2006 |
christos | branches: 1.33.2; 1.33.4;
comment out unreachable code.
|
| 1.32 | 11-Dec-2005 |
christos | branches: 1.32.4; 1.32.8;
merge ktrace-lwp.
|
| 1.31 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.30 | 01-Feb-2005 |
reinoud | branches: 1.30.6;
Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.29 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.28 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.27 | 28-Oct-2004 |
yamt | branches: 1.27.4;
move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.26 | 17-Oct-2004 |
bouyer | Adjust some debug printfs after b_bcount was changed from long to int.
From Patrick Welche on current-users@
|
| 1.25 | 18-Sep-2004 |
mycroft | Minor rearrangement. Whitespace and #include cleanup.
|
| 1.24 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.23 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.22 | 17-Sep-2004 |
mycroft | In places where we've already called scsipi_make_xs(), call scsipi_execute_xs()
directly rather than going through scsipi_command().
|
| 1.21 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.20 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.19 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.18 | 15-Nov-2001 |
lukem | branches: 1.18.16; 1.18.18;
don't need <sys/types.h> when including <sys/param.h>
|
| 1.17 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.16 | 18-Jul-2001 |
thorpej | bcmp -> memcmp
|
| 1.15 | 18-Jul-2001 |
thorpej | bcopy -> memcpy
|
| 1.14 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.13 | 25-Apr-2001 |
bouyer | branches: 1.13.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.12 | 09-Jun-2000 |
enami | branches: 1.12.4;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.11 | 30-Sep-1999 |
thorpej | branches: 1.11.2; 1.11.10;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.10 | 05-Apr-1999 |
mycroft | If scsipi_command() fails, always print out the error code.
|
| 1.9 | 18-Oct-1997 |
thorpej | branches: 1.9.12;
Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.8 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.7 | 27-Aug-1997 |
bouyer | branches: 1.7.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.6 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.5 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.4 | 05-May-1996 |
christos | Cleanup the rest of the SCSIDEBUG printfs. From Bernd Ernesti.
|
| 1.3 | 30-Mar-1996 |
christos | - Eliminate scsi_conf.h
- Fix prototypes
- Fix PR/2248 [Problems with HP scanjet...]
|
| 1.2 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.1 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.7.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.7.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.7.2.1 | 27-Aug-1997 |
thorpej | file ss_mustek.c was added on branch marc-pcmcia on 1997-08-27 23:33:35 +0000
|
| 1.9.12.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.11.10.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.11.2.4 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.11.2.3 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.11.2.2 | 20-Oct-1999 |
thorpej | Remove an unnecessary comment regarding XS_CTL_NOSLEEP.
|
| 1.11.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.12.4.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.12.4.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.12.4.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.12.4.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.13.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.13.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.18.18.1 | 11-Sep-2004 |
he | Pull up revisions 1.20-1.21 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.18.16.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.18.16.6 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.18.16.5 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.18.16.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.18.16.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.18.16.2 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.18.16.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.27.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.30.6.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.30.6.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.30.6.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.32.8.1 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.32.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.33.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.33.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.33.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.35.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.36.36.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.36.36.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.37.12.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.37.4.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.37.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.40.18.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #65):
sys/dev/scsipi/ss_scanjet.c: revision 1.52
sys/dev/scsipi/ss.c: revision 1.81
sys/dev/scsipi/ssvar.h: revision 1.19
sys/dev/scsipi/ss_mustek.c: revision 1.41
Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.40.16.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.40.16.2 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.40.16.1 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.40.12.2 | 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")
|
| 1.40.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.41.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.41.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.41.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.42.10.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.42.6.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.9 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.8 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.7 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.6 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.5 | 13-Feb-1998 |
enami | branches: 1.5.48; 1.5.56; 1.5.58;
Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.4 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.3 | 27-Aug-1997 |
bouyer | branches: 1.3.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.2 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.1 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.3.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.3.2.1 | 27-Aug-1997 |
thorpej | file ss_mustek.h was added on branch marc-pcmcia on 1997-08-27 23:33:36 +0000
|
| 1.5.58.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.5.56.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.5.48.2 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.5.48.1 | 04-Feb-2005 |
skrll | Sync with HEAD.
|
| 1.54 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.53 | 25-Oct-2013 |
martin | branches: 1.53.6; 1.53.10;
Mark a diagnostic-only variable
|
| 1.52 | 28-Feb-2012 |
mbalmer | branches: 1.52.2; 1.52.4;
Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.51 | 23-Nov-2009 |
rmind | branches: 1.51.12; 1.51.16; 1.51.18;
Remove some unecessary includes sys/user.h header.
|
| 1.50 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.49 | 13-Jan-2009 |
yamt | g/c BUFQ_FOO() macros and use bufq_foo() directly.
|
| 1.48 | 05-Apr-2008 |
cegger | branches: 1.48.4; 1.48.12;
use aprint_*_dev and device_xname
|
| 1.47 | 16-Nov-2006 |
christos | branches: 1.47.48;
__unused removal on arguments; approved by core.
|
| 1.46 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.45 | 03-Sep-2006 |
christos | branches: 1.45.2; 1.45.4;
comment out unreachable code
|
| 1.44 | 11-Dec-2005 |
christos | branches: 1.44.4; 1.44.8;
merge ktrace-lwp.
|
| 1.43 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.42 | 30-May-2005 |
christos | branches: 1.42.2;
- remove bogus casts
- add more const
|
| 1.41 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.40 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.39 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.38 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.37 | 28-Oct-2004 |
yamt | branches: 1.37.4; 1.37.6;
move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.36 | 18-Sep-2004 |
mycroft | Minor rearrangement. Whitespace and #include cleanup.
|
| 1.35 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.34 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.33 | 17-Sep-2004 |
mycroft | In places where we've already called scsipi_make_xs(), call scsipi_execute_xs()
directly rather than going through scsipi_command().
|
| 1.32 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.31 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.30 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.29 | 22-Apr-2004 |
itojun | sprintf -> snprintf
|
| 1.28 | 14-Sep-2002 |
chs | branches: 1.28.6; 1.28.8;
recognize HP ScanJet 4c.
add a catch-all case for other HP scanners.
remove a local version of atoi() in favor of strtoul() from libkern.
|
| 1.27 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.26 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.25 | 18-Jul-2001 |
thorpej | bcmp -> memcmp
|
| 1.24 | 18-Jul-2001 |
thorpej | bcopy -> memcpy
|
| 1.23 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.22 | 25-Apr-2001 |
bouyer | branches: 1.22.2;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.21 | 08-Jul-2000 |
sommerfeld | branches: 1.21.2;
printf format paranoia
|
| 1.20 | 09-Jun-2000 |
enami | Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.19 | 12-Mar-2000 |
augustss | branches: 1.19.2;
Remove extra \n in attach message.
|
| 1.18 | 30-Sep-1999 |
thorpej | branches: 1.18.2;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.17 | 05-Apr-1999 |
mycroft | If scsipi_command() fails, always print out the error code.
|
| 1.16 | 22-Apr-1998 |
pk | branches: 1.16.10;
Set a flag during auto-configuration, so SCSI_AUTOCONF can be passed
when doing additional I/O from the scanner attach functions.
|
| 1.15 | 19-Nov-1997 |
augustss | Recognize HP ScanJet 5p.
|
| 1.14 | 18-Oct-1997 |
thorpej | branches: 1.14.2;
Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.13 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.12 | 27-Aug-1997 |
bouyer | branches: 1.12.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.11 | 18-Jan-1997 |
cgd | kill use of strchr(); it's moving to libkern
|
| 1.10 | 22-Dec-1996 |
cgd | note in a comment (XXX) that strchr() (what's it doing here, anyway?)
causes 'const' to be discarded.
|
| 1.9 | 03-Dec-1996 |
thorpej | Recognize the HP ScanJet 4p, per PR #2663 (Kenneth Stailey).
|
| 1.8 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.7 | 10-Oct-1996 |
christos | - printf -> kprintf, sprintf -> ksprintf
- replace sequences of sprintf(p, ...); p += strlen(p) with p += sprintf(p,
|
| 1.6 | 18-May-1996 |
christos | Added prototypes for inline functions atoi() and strchr().
|
| 1.5 | 11-May-1996 |
mycroft | Bug fixes from PR 2329, plus a few from me.
|
| 1.4 | 05-May-1996 |
christos | Cleanup the rest of the SCSIDEBUG printfs. From Bernd Ernesti.
|
| 1.3 | 30-Mar-1996 |
christos | - Eliminate scsi_conf.h
- Fix prototypes
- Fix PR/2248 [Problems with HP scanjet...]
|
| 1.2 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.1 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.12.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.12.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.12.2.1 | 27-Aug-1997 |
thorpej | file ss_scanjet.c was added on branch marc-pcmcia on 1997-08-27 23:33:37 +0000
|
| 1.14.2.1 | 22-Nov-1997 |
mellon | Pull rev 1.15 up from trunk (augustss)
|
| 1.16.10.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.18.2.4 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.18.2.3 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.18.2.2 | 20-Oct-1999 |
thorpej | Remove an unnecessary comment regarding XS_CTL_NOSLEEP.
|
| 1.18.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.19.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.21.2.5 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.21.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.21.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.21.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.21.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.22.2.3 | 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
|
| 1.22.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.22.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.28.8.1 | 11-Sep-2004 |
he | Pull up revisions 1.31-1.32 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.28.6.8 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.28.6.7 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.28.6.6 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.28.6.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.28.6.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.28.6.3 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.28.6.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.28.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.37.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.37.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.42.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.42.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.44.8.1 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.44.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.45.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.45.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.45.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.47.48.2 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.47.48.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.48.12.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.48.4.2 | 11-Mar-2010 |
yamt | sync with head
|
| 1.48.4.1 | 04-May-2009 |
yamt | sync with head.
|
| 1.51.18.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #65):
sys/dev/scsipi/ss_scanjet.c: revision 1.52
sys/dev/scsipi/ss.c: revision 1.81
sys/dev/scsipi/ssvar.h: revision 1.19
sys/dev/scsipi/ss_mustek.c: revision 1.41
Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.51.16.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.51.16.2 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.51.16.1 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.51.12.2 | 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")
|
| 1.51.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.52.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.52.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.52.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.53.10.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.53.6.1 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.19 | 28-Feb-2012 |
mbalmer | Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.18 | 30-Jun-2011 |
wiz | branches: 1.18.2; 1.18.6; 1.18.8;
dependant -> dependent
|
| 1.17 | 11-Dec-2005 |
christos | merge ktrace-lwp.
|
| 1.16 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.15 | 27-Feb-2005 |
perry | branches: 1.15.4;
nuke trailing whitespace
|
| 1.14 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.13 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.12 | 27-Aug-2004 |
bouyer | branches: 1.12.4; 1.12.6;
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.11 | 21-Aug-2004 |
thorpej | - De-__P.
- Use ANSI function decls.
|
| 1.10 | 02-Nov-2003 |
wiz | branches: 1.10.2;
boundary, not boundry. Inspired by Tom Cosgrove.
|
| 1.9 | 22-Jul-2002 |
hannken | branches: 1.9.6;
Convert to new device buffer queue interface.
|
| 1.8 | 25-Apr-2001 |
bouyer | branches: 1.8.2; 1.8.14;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.7 | 21-Jan-2000 |
thorpej | branches: 1.7.6;
Update for sys/buf.h/disksort_*() changes.
|
| 1.6 | 22-Apr-1998 |
pk | branches: 1.6.14;
Set a flag during auto-configuration, so SCSI_AUTOCONF can be passed
when doing additional I/O from the scanner attach functions.
|
| 1.5 | 13-Feb-1998 |
enami | Backout previous change, and rather, remove all protection but the one
actually currently required. Suggested by Charles M. Hannum.
|
| 1.4 | 13-Feb-1998 |
enami | Fix or add protection for mutiple inclusion.
|
| 1.3 | 27-Aug-1997 |
bouyer | branches: 1.3.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.2 | 30-Mar-1996 |
christos | - Eliminate scsi_conf.h
- Fix prototypes
- Fix PR/2248 [Problems with HP scanjet...]
|
| 1.1 | 18-Feb-1996 |
mycroft | Add SCSI scanner support by Kenneth Stailey and Joachim Koenig-Baltes,
hacked a bit. Needs more work.
|
| 1.3.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.3.2.1 | 27-Aug-1997 |
thorpej | file ssvar.h was added on branch marc-pcmcia on 1997-08-27 23:33:38 +0000
|
| 1.6.14.2 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.6.14.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.7.6.2 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.7.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.8.14.1 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.8.2.1 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.9.6.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.9.6.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.9.6.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.9.6.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.9.6.3 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.9.6.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.9.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.10.2.1 | 11-Sep-2004 |
he | Pull up revision 1.12 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.12.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.12.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.15.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.18.8.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #65):
sys/dev/scsipi/ss_scanjet.c: revision 1.52
sys/dev/scsipi/ss.c: revision 1.81
sys/dev/scsipi/ssvar.h: revision 1.19
sys/dev/scsipi/ss_mustek.c: revision 1.41
Convert to device_t, aka softc/device_t split, and clean up a bit the code.
|
| 1.18.6.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.18.6.2 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.18.6.1 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.18.2.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.243 | 23-Feb-2022 |
andvar | fix various typos in comments, mainly immediatly/immediately/,
as well shared and recently fixed typos in OpenBSD code by Jonathan Grey.
|
| 1.242 | 05-Dec-2021 |
msaitoh | availabe -> available in comment.
|
| 1.241 | 13-Jun-2021 |
mlelstv | Restore EOM handling.
|
| 1.240 | 27-Dec-2019 |
msaitoh | branches: 1.240.12;
s/transfered/transferred/
|
| 1.239 | 10-Nov-2019 |
chs | 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.
|
| 1.238 | 19-May-2019 |
kardel | Add simple position recovery when positioning to EOM by reading
the position with READ_POSITION.
this allows for
mt eom
mt st
to return the correct file position.
|
| 1.237 | 23-Feb-2019 |
kamil | Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
|
| 1.236 | 12-Feb-2019 |
kardel | Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
|
| 1.235 | 03-Feb-2019 |
mrg | - add or adjust /* FALLTHROUGH */ where appropriate
- add __unreachable() after functions that can return but won't in
this case, and thus can't be marked __dead easily
|
| 1.234 | 24-Mar-2018 |
mlelstv | branches: 1.234.2;
Use separate lock to protect internal state and release locks when
calling biodone.
|
| 1.233 | 24-Mar-2018 |
kamil | Revert previous commit in st.c in order to unbreak build
There are various build errors like:
/usr/src/sys/dev/scsipi/st.c: In function 'stattach':
/usr/src/sys/dev/scsipi/st.c:398:16: error: 'struct st_softc'
has no member named 'buf_defer'; did you mean 'buf_queue'?
bufq_alloc(&st->buf_defer, "fcfs", 0);
Original (reverted) commit:
Use separate lock to protect internal state and release locks when
calling biodone.
|
| 1.232 | 23-Mar-2018 |
mlelstv | Use separate lock to protect internal state and release locks when
calling biodone.
|
| 1.231 | 17-Jun-2017 |
mlelstv | branches: 1.231.4;
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
|
| 1.230 | 20-Nov-2016 |
mlelstv | branches: 1.230.8;
Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.229 | 20-Nov-2016 |
pgoyette | Avoid calling bufq_free() from critical code sections.
|
| 1.228 | 14-Jul-2016 |
msaitoh | branches: 1.228.2;
- Use aprint*() instead of printf() in xxx_attach().
- Add missing aprint_naive("\n");
- KNF
|
| 1.227 | 24-Aug-2015 |
pooka | would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.226 | 10-Aug-2014 |
tls | branches: 1.226.2; 1.226.4; 1.226.6; 1.226.10;
Merge tls-earlyentropy branch into HEAD.
|
| 1.225 | 25-Jul-2014 |
dholland | Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.224 | 25-Jul-2014 |
dholland | Add d_discard to all struct bdevsw instances I could find.
I've set them all to nodiscard. Some of them (wd, dk, vnd, ld,
raidframe, maybe cgd) should be implemented for real.
|
| 1.223 | 16-Mar-2014 |
dholland | branches: 1.223.2;
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.
|
| 1.222 | 25-Oct-2013 |
martin | Mark a diagnostic-only variable
|
| 1.221 | 19-Apr-2012 |
bouyer | branches: 1.221.2; 1.221.4;
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.220 | 28-Feb-2012 |
mbalmer | Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.219 | 25-Feb-2012 |
shattered | Replace magic value 0x7 with constant MT_ISAR.
OK by wiz@
|
| 1.218 | 02-Feb-2012 |
tls | branches: 1.218.2;
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.
|
| 1.217 | 30-May-2010 |
pgoyette | branches: 1.217.8; 1.217.12;
Extract SCSIVERBOSE into a kernel module. The module can be builtin
by defining 'options SCSIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.
The module is built for all architectures, whether or not SCSI or
atapi support exists.
|
| 1.216 | 06-Dec-2009 |
dyoung | branches: 1.216.2; 1.216.4;
Delete do-nothing device-activation hooks.
|
| 1.215 | 05-Dec-2009 |
pooka | Convert tsleep(&lbolt) to kpause(). Make ltsleep/mtsleep on lbolt
illegal. I examined all places where lbolt is referenced to make
sure there were pointer aliases of it passed to tsleep, but put a
KASSERT in m/ltsleep() just to be sure.
|
| 1.214 | 23-Nov-2009 |
rmind | Remove some unecessary includes sys/user.h header.
|
| 1.213 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.212 | 15-Aug-2009 |
pgoyette | 1. Move the mode_select functionality into common code (in st.c) and
invoke the common routine for both scsi and atapi tapes.
2. Replace a numeric constant with some sizeof's when calculating the
size of the mode_select command buffer, clear the entire buffer, and
KASSERT to ensure the page_0_size loaded from quirk table is valid.
3. Add a quirk for my Seagate Travan-40 tape drive.
As discussed on tech-kern@
Addresses my PR kern/34832
|
| 1.211 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.210 | 13-Jan-2009 |
yamt | branches: 1.210.2;
g/c BUFQ_FOO() macros and use bufq_foo() directly.
|
| 1.209 | 11-Jan-2009 |
cegger | make this compile
|
| 1.208 | 17-Nov-2008 |
bouyer | If we return an error make sure we don't return a b_resid of 0; or this
will trigger a KASSERT in physio_done(). Fix issue reported by Todd Kover
on tech-kern@
|
| 1.207 | 02-Sep-2008 |
dholland | branches: 1.207.2; 1.207.4;
Reword some misleading code.
ok bouyer@
|
| 1.206 | 01-Sep-2008 |
dholland | fix typo in comment
|
| 1.205 | 08-Jun-2008 |
tsutsui | branches: 1.205.4;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.204 | 28-Apr-2008 |
martin | branches: 1.204.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.203 | 05-Apr-2008 |
cegger | branches: 1.203.2; 1.203.4;
use aprint_*_dev and device_xname
|
| 1.202 | 11-Oct-2007 |
christos | branches: 1.202.18;
PR/37108: Pauly Goyette: Compute the timeout value, before immediate is
changed for atapi tape drives.
|
| 1.201 | 06-Oct-2007 |
bouyer | Move check for b_resid value to the right place. Should definitively fix
kern/36690.
|
| 1.200 | 01-Oct-2007 |
bouyer | When checking for invalid b_resid valyes, also check negative ones.
Should fix kern/36690.
|
| 1.199 | 29-Sep-2007 |
bouyer | Properly set b_resid in an error case.
Clamp b_resid to b_bcount in case the info field is larger. Fix a KASSERT
in physio() when a media error occurs on my SDLT320 drive (maybe this drive
is bogus for not setting the highter bits to 0 in the INFO field in this
case).
|
| 1.198 | 29-Jul-2007 |
ad | branches: 1.198.4; 1.198.6; 1.198.8; 1.198.10;
It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.197 | 09-Jul-2007 |
ad | branches: 1.197.2;
Merge some of the less invasive changes from the vmlocking branch:
- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements
|
| 1.196 | 04-Mar-2007 |
christos | branches: 1.196.2; 1.196.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.195 | 01-Mar-2007 |
thorpej | TRUE -> true, FALSE -> false
|
| 1.194 | 16-Nov-2006 |
christos | branches: 1.194.2; 1.194.4; 1.194.8;
__unused removal on arguments; approved by core.
|
| 1.193 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.192 | 23-Aug-2006 |
christos | branches: 1.192.2; 1.192.4;
Change iostat_alloc() to take the parent pointer and the name directly, so
that callers are not responsible for initializing the fields. Store the name
inside the struct instead of maintaining a pointer to external storage, or
leaked memory (nfs case).
|
| 1.191 | 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
|
| 1.190 | 20-Apr-2006 |
blymn | branches: 1.190.2;
Prefix iostat structure elements with io_
|
| 1.189 | 14-Apr-2006 |
blymn | Make i/o statistics collection more generic, include tape drives and
nfs mounts in the set of devices that statistics will be reported on.
|
| 1.188 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.187 | 28-Mar-2006 |
thorpej | Use device_unit().
|
| 1.186 | 11-Dec-2005 |
christos | branches: 1.186.4; 1.186.6; 1.186.8; 1.186.10; 1.186.12;
merge ktrace-lwp.
|
| 1.185 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.184 | 13-Aug-2005 |
blymn | Split out stats attach code ready for shifting
Add sysctl hooks here for the time being.
|
| 1.183 | 07-Aug-2005 |
blymn | Gather statistics for tape drives.
|
| 1.182 | 16-Jul-2005 |
rtr | change output from "rogue" to "quirks apply" - resolves pr#25387
|
| 1.181 | 29-May-2005 |
christos | branches: 1.181.2;
- Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.180 | 25-Apr-2005 |
drochner | revert the only part of rev. 1.177 which made a functional
change -- it broke error reporting because st->asc(q) are
not set
(actually, I believe that SSD_RCODE_VALID is misnamed)
|
| 1.179 | 31-Mar-2005 |
yamt | introduce a function to drain bufq and use it where appropriate.
|
| 1.178 | 27-Feb-2005 |
perry | branches: 1.178.2;
nuke trailing whitespace
|
| 1.177 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.176 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.175 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.174 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.173 | 28-Oct-2004 |
yamt | branches: 1.173.4; 1.173.6;
move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.172 | 26-Sep-2004 |
dogcow | Fix debug message output args to match yamt's src/sys/sys/buf.h changes.
|
| 1.171 | 18-Sep-2004 |
mycroft | Minor rearrangement. Whitespace and #include cleanup.
|
| 1.170 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.169 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.168 | 17-Sep-2004 |
mycroft | In places where we've already called scsipi_make_xs(), call scsipi_execute_xs()
directly rather than going through scsipi_command().
|
| 1.167 | 17-Sep-2004 |
mycroft | Do not manipulate xs->bp in "generic" code -- do it only in the psw_done
routine. As part of this, pass down our pre-parsed error code -- though this
interface will probably change later to accomodate better error handling.
|
| 1.166 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.165 | 27-Aug-2004 |
bouyer | Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.164 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.163 | 29-Jun-2003 |
fvdl | branches: 1.163.2; 1.163.4;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.162 | 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
|
| 1.161 | 14-Apr-2003 |
perry | Add a quirk for the Tandberg SLR/5
Fix supplied by Rex McMaster in PR kern/13603
|
| 1.160 | 20-Mar-2003 |
dbj | use PRId64 to printf bp->b_blkno, which is of type daddr_t
|
| 1.159 | 03-Feb-2003 |
thorpej | Test callout_pending(), not callout_active(), and eliminate now-unnecessary
callout_deactivate() calls.
|
| 1.158 | 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
|
| 1.157 | 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.
|
| 1.156 | 22-Jul-2002 |
hannken | Convert to new device buffer queue interface.
|
| 1.155 | 05-May-2002 |
bouyer | branches: 1.155.2;
If periph->periph_callout is already active, don't freeze the periph again:
scispi_periph_timed_thaw() will be called only one time anyway.
|
| 1.154 | 03-May-2002 |
bouyer | Back out previous. It has issues with multiple dumps to one tape using
/dev/nrstx, as pointed out by Matthew Jacob.
My problem is probably related to mt using /dev/nrst0 instead of /dev/enrst0
by default.
|
| 1.153 | 02-May-2002 |
bouyer | If we get a Unit Attention Not Ready To Ready Transition (medium may have
changed) sense, and the periph was not open, then ignore the error.
|
| 1.152 | 22-Mar-2002 |
mjacob | kern/16014- if we do an 'mt offline', make sure we've allowed the tape
to be ejected (PR_ALLOW). This got nuked in a previous change.
|
| 1.151 | 22-Mar-2002 |
mjacob | Forced commit so that some more detail on the previous patches can be
inserted into comments. I really think that more testing was required,
but Christos was eager.
+ Add in tape block/file position reporting support.
+ Change the prevent/allow ejection code so that if the device
is close the tape can be ejected. This allows robots to eject
the tape even if the tape isn't at BOT.
+ Don't do a null filemark write if we're reading block position. This
was such an abysmal performance killer that it had to go. Assume
that the tape will DTRT and tell you what the tape position really is
including any buffered data. You are depending on the tape to do
the right thing to report position at all, so the risk factors don't
increase appreciably here. BTW- we only do this for logical block
position. For 'hardware' block position, we still flush.
+ Some substantial tweaking in st_interpret_sense. This essentially duplicates
the FreeBSD driver's exception handling code. This is to try and correctly
handle EOM cases. I'm not sure that this is now correct with respect to
the ENABLE EARLY WARNING code- I hadn't gotten around to validating that.
Add in full 32 bit residuals- this can then be available to an update
MTIOCGET in the future.
+ Make sure we handle SKEY_VOLUME_OVERFLOW correctly.
|
| 1.150 | 20-Mar-2002 |
christos | Add the ability to report file number/block number. From Matt Jacob.
|
| 1.149 | 12-Jan-2002 |
bouyer | For fixed block drive, for invalid sense error code, set info to
xs->datalen / st->blksize instead of xs->datalen; as code assume later
that info is a block number.
Should fix kern/2727.
|
| 1.148 | 07-Dec-2001 |
yamt | add detach support for st.
|
| 1.147 | 01-Dec-2001 |
bouyer | Hum, no need to print a message each time we're waiting for the drive.
|
| 1.146 | 01-Dec-2001 |
bouyer | Various quirks for the ATAPI OnStream DI-30, mostly from the FreeBSD driver.
Many thanks to Chris Pinnock for giving me remote access to his hardware.
|
| 1.145 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.144 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.143 | 18-Jul-2001 |
thorpej | branches: 1.143.2;
bzero -> memset
|
| 1.142 | 08-Jul-2001 |
wiz | branches: 1.142.2;
Correct various misspellings of 'transfer' and inflected forms.
|
| 1.141 | 18-Jun-2001 |
bouyer | Snapshot of ATAPI tapes support. Known to be able to read/write to tape with:
st0 at atapibus0 drive 1: <Seagate STT8000A, , 5.51> type 1 sequential
removable
Major changes may still happen in order to properly support other ATAPI
tape drives.
|
| 1.140 | 30-May-2001 |
bouyer | uninitialised variable (from Krister Walfridsson):
in st_cmprss(), init flags to 0 (never changed, but this should change soon
anyway, for ATAPI tapes support)
|
| 1.139 | 15-May-2001 |
lukem | delint; use MAX(sizeof(..),sizeof(..)) instead of max(...) for array size
|
| 1.138 | 14-May-2001 |
bouyer | Use SCSI/ATAPI common definition for MODE_{SELECT,SENSE}{,_BIG}. Define
functions to send theses commands in scsipi_base.c and use them instead
of ad-hoc commands setups.
|
| 1.137 | 06-May-2001 |
hannken | Fix a typo from thorpej_scsipi merge. No-sense information with retval == 0
should only print if SCSIPI_DEBUG is defined.
|
| 1.136 | 04-May-2001 |
bouyer | Add an ATAPI front-end to the st driver. Completely untested for now,
but st at scsi should still work :)
|
| 1.135 | 25-Apr-2001 |
bouyer | Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.134 | 18-Jan-2001 |
jdolecek | branches: 1.134.2;
constify
|
| 1.133 | 22-Nov-2000 |
soren | Remove obsolete comment.
|
| 1.132 | 03-Nov-2000 |
pk | st_touch_tape(): Pass XS_CTL_SILENT to st_mode_select().
|
| 1.131 | 03-Nov-2000 |
pk | * st_touch_tape(): start probing at the block size from the sense data if
we have it.
* st_unmount(): reset density to the device default. This prevents using
stale density values after changing to a medium with different density.
Section 9.3.3 of the SCSI specs states that a device shall return the
density value specified in the last succesfull MODE SELECT after an
unload operation, in case it is not able to automatically determine
the density of the new medium.
|
| 1.130 | 03-Nov-2000 |
pk | * Per st(4), retrieving status and setting modes on the control device
should succeed even if no media is loaded.
* In stopen(), check the return value of st_mount_tape().
|
| 1.129 | 02-Nov-2000 |
pk | Introduce ST_Q_ERASE_NOIMM, a quirk indicating that the device rejects
an ERASE command with the `Immed' bit on.
Use it on the VIPER 21247 & 21531.
|
| 1.128 | 02-Nov-2000 |
pk | Prevent printing sense information twice.
|
| 1.127 | 02-Nov-2000 |
pk | Remove useless check for SDEV_MEDIA_LOADED from st_read_block_limits().
This command should work whether or not media is present in the device.
|
| 1.126 | 02-Nov-2000 |
pk | * In st_touch_tape(), don't bail out if the MODE SELECT fails since we are,
after all, probing the device for acceptable parameters.
* In st_loadquirks(), copy mode-specific quirks from the quirk table
to `st_softc'; otherwise all such quirks save ST_Q_FORCE_BLKSIZE are ignored.
|
| 1.125 | 16-Aug-2000 |
matt | Add a NOPREVENT quirk which prevents the sending of PREVENT messages (some
manual 9track tapes don't support it). Add a quirk entry for the NCR H621
9track tape drive.
|
| 1.124 | 09-Jun-2000 |
enami | branches: 1.124.2;
Prevent a process being swapped out during I/O if the data buffer is
allocated on stack. This potential problem is noticed by Noriyuki Soda
and the idea and sample code to fix is given by Jason R. Thorpe.
|
| 1.123 | 06-Jun-2000 |
soren | Density codes are usually given in decimal, so print them that way.
|
| 1.122 | 19-May-2000 |
kleink | branches: 1.122.2;
Add a quirk table entry for the OnStream ADR50 Drive;
from S.P.Zeidler <spz@serpens.swb.de> in kern/10118.
|
| 1.121 | 30-Mar-2000 |
augustss | Get rid of register declarations.
|
| 1.120 | 21-Feb-2000 |
mjacob | Patches from msouth@scruz.net to handle QIC 3220 (HP T20).
|
| 1.119 | 20-Feb-2000 |
mjacob | minor spelling change
|
| 1.118 | 22-Jan-2000 |
mjacob | PR/9271 (from srp@zgi.com)- recognize (indirectly) HP4000s via density
code and do the right QIC type dance (1 FM @EOT, fixed block size).
|
| 1.117 | 21-Jan-2000 |
thorpej | Update for sys/buf.h/disksort_*() changes.
|
| 1.116 | 13-Jan-2000 |
nisimura | Add tweaks for TEAC compact cassette tape drive.
|
| 1.115 | 12-Jan-2000 |
mjacob | Check in the implementation of the ST_MOUNT_DELAY option.
This is an attempt to allow people to change the default configuration
to try harder at 'mounting' a tape. This allows you to specify, in
seconds, the amount of time a non-control unit open will retry
(once per second) the scsipi_test_unit_ready when it tries to mount
the tape. It also turns off the over-verbose error reporting at
this time unless SCSIDEBUG is set.
The reason this is not enabled as a default is that it's a large change
of behaviour. I find it useful to 'try harder' at mounting a tape in
the tape driver, particularly when loaded via a media changer device
rather than specifying the delays in the backup program.
|
| 1.114 | 30-Sep-1999 |
thorpej | branches: 1.114.2; 1.114.8;
Cleanup the scsipi_xfer flags:
- `flags' is now gone, replaced with `xs_control' and `xs_status'.
- Massive cleanup of the control flags. Now we explicitly say that
a job is to complete asynchronously, rather than relying on side-effects,
and use a new flag to now that device discovery is being performed.
- Do SCSI device discovery interrupt-driven.
|
| 1.113 | 04-Sep-1999 |
simonb | Fix tyop in a comment.
|
| 1.112 | 17-Jun-1999 |
mjacob | STK 9490 && SD-3 drives incorrectly quirked as not supporting the LOAD command.
Anyway, just because a drive doesn't support the LOAD (to BOT) command does
not mean that the drive doesn't support the UNLOAD command. Also note and
print errors in rewinds and unloads (and errors in writing closing filemarks
for same).
|
| 1.111 | 05-Apr-1999 |
mycroft | If scsipi_command() fails, always print out the error code.
|
| 1.110 | 28-Feb-1999 |
explorer | branches: 1.110.2; 1.110.4;
Update to slightly altered rnd_attach_source() api
|
| 1.109 | 10-Feb-1999 |
bouyer | Abort transfer if b_blkno is negative. Closes PR kern/5553 by Johan Danielsson.
|
| 1.108 | 10-Jan-1999 |
tron | Make MTIOCRDSPOS, MTIOCSLOCATE, etc. work with write protected tapes.
Patch supplied by Dave Huang in PR kern/5305.
|
| 1.107 | 08-Dec-1998 |
thorpej | When closing, wait for pending xfers to drain before deleting the reference
to the adapter.
|
| 1.106 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.105 | 17-Nov-1998 |
bouyer | Rename scsi_interpret_sense() to scsipi_interpret_sense() and move it from
scsi_base.c to scsipi_base.c. Rename the functions from scsi_verbose.c
too, and rename the file itself. Cleaup includes too (scsi_*.h should not
be #included in scsipi_*.h files, which are supposed to be
common to atapi and scsi).
|
| 1.104 | 11-Sep-1998 |
mjacob | keren/6128: add an entry for the TDC 4200. Full density code set isn't known.
|
| 1.103 | 02-Sep-1998 |
mjacob | typo for non-SCSIVERBOSE case
|
| 1.102 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.101 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.100 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.99 | 10-Aug-1998 |
mjacob | Responding to an issue brought to my attention- when the device is
opened norewind and 2 filemarks are written at the end a phantom file
is left (just what I was afraid of, but I didn't think about it in
the last delta because somehow I had managed to convince myself that
this was a nonissue. It's not.).
So- in stdone clear ST_WRITTEN for regular reads. In st_close, preserve
the state of ST_WRITTEN, and if no error and 2FM@EOD for this device and
this is a no-rewind open, backspace one filemark. This should preserve
(for this mount session) FILE - FMK - FILE - FMK - FILE ... FILE FMK FMK EOD
sequencing.
This doesn't clean up the case of EOM appends- in this case you *will* still
get (after an MTEOM operation and a write of a file) a phantom empty file,
e.g. FILE - FMK FMK - FILE - FMK FMK EOD *unless* you follow the EOM operation
with an explicit backspace. The trouble is that this makes it difficult for
seamless interchange with other systems which don't necessarily follow.
The preferrable alternative would be to eliminate the 2FM@EOD except for
1/2" Reel tapes, but that has been pretty much nixed within developers.
|
| 1.98 | 06-Aug-1998 |
drochner | fix incorrect length of LOCATE CDB (reported in PR kern/5306 by
Dave Huang <khym@bga.com>)
|
| 1.97 | 31-Jul-1998 |
mjacob | Suggestion from Matthias Drochner: If at close you decide to write
filemark 'coz you opened write only and didn't do anything else,
call st_check_eod to possibly write TWO furshlugginer filemarks.
Also- return any errors from writing filemarks out of stclose.
|
| 1.96 | 31-Jul-1998 |
mjacob | Several more changes. First of all, move CTRL_MODE to have I/O behaviour
like a no-rewind device. Secondly figure out whether the initial TUR
for a CTRL_MODE open resulted in a tape being actually found (if so,
then do a mount session).
Move the 'sun compatibility' behaviour into stdone && stclose- don't
mark a tape as having been written in stopenm, fer gosh sakes.
|
| 1.95 | 30-Jul-1998 |
mjacob | Some minor comment tweaking.
Also- to be fair and on review, kern/391 isn't really addressed by
the previous commits. In reviewing, I'm embarassed to find that this
talks about reading at EOT. I'm actually going to claim that this
is 'not a bug' or 'fixed already' in that at the end of media (at the
edge of recorded media), you may continuously open the tape (should
you choose to) issue a read, and zero bytes will transfer- this is a
sufficient EOF indicator.
|
| 1.94 | 30-Jul-1998 |
mjacob | Clarify and name some of the 'open' modes. Clarify a few comments. Remove
a now unused variable. Also, remove the restriction against at density
code being greater than the max SCSI 2 density code: 0x80..0xff are the
Vendor Unique codes and most certainly should be allowed. The check for
invalid values should be less than 0 or greater than 255.
Oh- yeah, the previous commit addressed kern/391.
|
| 1.93 | 30-Jul-1998 |
mjacob | branches: 1.93.2;
Make some changes wrt EOM behaviour. Distinguish EIO_PENDING from
EOM_PENDING. Set up a persistent EARLYWARNING behaviour flag. If
set, EOM behaviour forces a 'short read' to signal logical (as
opposed to physical) end of media. The user application may, of
course, do with this information what it will.
The EARLYWARNING behaviour may be enabled/disabled by a MTIOCTOP
operation. The default action is to not have EARLYWARNING enabled-
but this may be reversed by an option ST_ENABLE_EARLYWARN in
the kernel build.
|
| 1.92 | 19-Jul-1998 |
drochner | Remove a check which restricted raw SCSI commands to the "minor 3"
("CTLMODE") subdevice. There are legitimate uses for raw commands with
normal tape handles too.
[I'm not sure if this is a final solution. Administrators might want
to set up a more finegrained policy. However, this should not be mixed
with the "set defaults" semantics of the "CTLMODE" subdevice; another
flag should be used instead (eg execute permission or a minor number bit).]
|
| 1.91 | 15-Jul-1998 |
mjacob | part of fix for kern/3835: use of enumerated returns from target sense handlers
|
| 1.90 | 04-Jul-1998 |
mjacob | Minor enhancements:
1) Quirk entries for Storage Tek 9490 (Timberline) and D3 (Redwood)
drives.
2) Modification to st_loadtape to do a REWIND to BOT if the
action is a load and the tape doesn't support the LOAD command
(9490, SD3, and IBM 3590).
3) Cleaned up the 'undersized user record' error message to
make a little more sense.
Various bug fixes:
kern/1275: Now returns values in dsreg and erreg and sets resid
(as best as it can for a 16 but integer). See also
a recent change to mtio.h. We are declining to fix
the portion of this bug about naming a more specific
SCSI device. Since there is nothing programmatic
you can do with that information, it is not useful
to pass back at this time.
A side effect of this change is that doing MTIOCGET
also forces a mode sense (to get the current state
of WRITE PROTECT).
kern/5647: Now no longer logs to the console ILI or Filemark or (first)
EOM (on write) errors (unless SCSIDEBUG is set).
kern/5525: Substantially increased timeouts for a variety of
operations, and split them into categories of
I/O, Space, and Control operations (each have
likely different inherent times). I/O is for
reads/writes. Control is for mode sense/select.
Space is for spacing the tape.
Until EOM handling is changed, though kern/391 is still not fixed. A side
effect of EOM handling is that you now always 'lose' (to the writing
application's view) the last write since EIO is what is returned on
EOM detection during writes. Hopefully the reader applications don't
get too bent out of shape by this.
|
| 1.89 | 13-Feb-1998 |
enami | Don't include <dev/scsipi/scsi_all.h> twice.
|
| 1.88 | 07-Feb-1998 |
pk | * Add the "VIPER 150/21531" to quirk list (ST_Q_SENSE_HELP).
* Double timeout on tape position commands to 30 minutes. Some units
seem capable of taking that much time.
|
| 1.87 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.86 | 28-Dec-1997 |
is | Added ST_Q_SENSE_HELP to the Archive Viper 150S quirks so that
density 0 (autosense) actually works.
|
| 1.85 | 07-Nov-1997 |
mjacob | If we're in the middle of opening, don't let st_interpret_sense get
into the act and return EIO or whatever when we're just eating the
initial TEST UNIT READY induced check condition.
|
| 1.84 | 18-Oct-1997 |
thorpej | branches: 1.84.2;
Implement two macros, scsipi_command() and scsipi_command_direct(), and
use them to hide the structure of the function pointers we jump through
to issue a command.
|
| 1.83 | 16-Oct-1997 |
mycroft | Define FALSE and TRUE locally.
|
| 1.82 | 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.
|
| 1.81 | 12-Oct-1997 |
mjacob | Hello? If we're doing block positioning a timeout of greater than 5 seconds
is probably going to be necessary.
|
| 1.80 | 10-Oct-1997 |
explorer | Add hooks to insert timing info into the random system
|
| 1.79 | 09-Oct-1997 |
enami | Cosmetic changes;
- dereference a pointer to function explicitly.
- fold long line to fit columns < 80.
- put whitespaces around a binary operator.
- don't put a single statement into a block.
|
| 1.78 | 09-Oct-1997 |
mjacob | A) At least IBM 3590 tape drives return a NOT READY error if issued a LOAD
to BOT command- even if at BOT. Urk. B) Make READ POSITION SCSI_SILENT.
|
| 1.77 | 01-Oct-1997 |
mjacob | Many thanks to Wolfgang Rupprecht for spotting this... In order to be sure
about actual position when reading tape position, you should flush any
pending writes. Well, if the tape is write protected, some drives don't
see the zero count WRITE FILE MARK command as a no-op and complain. Dumb!
|
| 1.76 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.75 | 29-Sep-1997 |
mjacob | Add in mtop operation MTCMPRESS, which can enable or disable tape compression.
The method for doing this is to try and use the DATA COMPRESSION mode page
first, followed by the DEVICE CONFIGURATION page (this is because most newer
tape devices are now using DATA COMPRESSION instead of DEVICE CONFIGURATION
pages).
Add in the logical && hardware read position and set position ioctls. Oddly
enough, because NetBSD is limited in having the driver track file && record
numbers, the usual agony over what to do once you use logical or hardware
block positioning can be avoided. Amusing.
Make a minor change so that for SCSIVERBOSE cases that SCSI_SILENT in
the xs' flags is still observed.
|
| 1.74 | 13-Sep-1997 |
enami | Declare SCSIVERBOSE by defopt in files.scsipi. Include opt_scsiverbose.h
in scsi_base.c and st.c.
|
| 1.73 | 27-Aug-1997 |
bouyer | branches: 1.73.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.72 | 20-Aug-1997 |
mjacob | If SCSIVERBOSE is enabled, call the routine to print out detailed sense
information.
|
| 1.71 | 21-Feb-1997 |
thorpej | branches: 1.71.4;
If posting an error condition because the media has been unloaded,
make sure to set the residual count to reflect that no data was
transfered.
From Naofumi HONDA / MINOURA Makoto, PR #3007.
|
| 1.70 | 05-Dec-1996 |
cgd | branches: 1.70.6;
update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.69 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.68 | 10-Oct-1996 |
christos | printf -> kprintf, sprintf -> ksprintf
|
| 1.67 | 24-May-1996 |
thorpej | Implement ST_Q_UNIMODAL quirk, which indicates that a drive will reject
all attempts to mode select, even if the attempted mode is supported.
Add the ST_Q_UNIMODAL quirk for the HP Colorado T4000s which exhibits
this behavior. (Someone please lob a tactical nuke in that direction.)
From David Rosenthal <dshr@vitria.com> on netbsd-bugs.
|
| 1.66 | 05-May-1996 |
christos | Cleanup the rest of the SCSIDEBUG printfs. From Bernd Ernesti.
|
| 1.65 | 30-Mar-1996 |
christos | Eliminate scsi_conf.h.
|
| 1.64 | 19-Mar-1996 |
mycroft | Define a full set of [234][bl]tol() and lto[234][bl]() conversion functions,
inlined.
Use sized types in protocol structures.
Make the definition of scsi_sense_data less ugly.
|
| 1.63 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.62 | 05-Mar-1996 |
thorpej | Normalize SCSI autoconfiguration output.
From Chris Demetriou <cgd@NetBSD.ORG>. Fixes PR #1958.
|
| 1.61 | 18-Feb-1996 |
mycroft | Minor change.
|
| 1.60 | 17-Feb-1996 |
jtk | remove #ifdef notdef made unnecessary by previous changes (PR#1597)
put in missing "if (error)" that caused tape IO to always fail.
(closes PR#2086)
|
| 1.59 | 14-Feb-1996 |
christos | scsi prototypes
|
| 1.58 | 09-Feb-1996 |
briggs | PR#2048. Prototype st_erase().
|
| 1.57 | 11-Jan-1996 |
thorpej | Honor cache request and implement erase command, and add the SCSI tape
device configuration page. Fixes PRs 807, 1201, and 1705.
From John Kohl <jtk@kolvir.blrc.ma.us>.
|
| 1.56 | 05-Jan-1996 |
pk | SCSI sense `info' field is a signed value (from John Kohl; PR#1597).
|
| 1.55 | 30-Nov-1995 |
pk | Add quirk for Exabyte 8200/rev. 263H (from Paul Goyette; PR#1797).
|
| 1.54 | 13-Oct-1995 |
gwr | Add quirk entries for Wangtek SCSI tapes. Also, add entry for
Tandberg 3800 (fixes PR#1592 - from Jochen Pohl).
|
| 1.53 | 12-Aug-1995 |
mycroft | opri --> s
|
| 1.52 | 12-Aug-1995 |
mycroft | Fix oversight in previous.
|
| 1.51 | 12-Aug-1995 |
mycroft | minphys() functions really should return void.
|
| 1.50 | 24-Jul-1995 |
cgd | update SCSI minphys routines' definitions to match standard minphys()
definition and usage.
|
| 1.49 | 04-Jul-1995 |
mycroft | Make each disk and tape driver define its own read and write functions.
Deprecate rawread() and rawwrite() completely. Remove d_strategy from cdevsw to
force the abstraction barrier.
|
| 1.48 | 26-Jun-1995 |
cgd | make dump stubs consistent
|
| 1.47 | 25-Mar-1995 |
mycroft | Add a rogue entry for the ANRITSU DMT780.
|
| 1.46 | 30-Dec-1994 |
mycroft | Update the rogue entry for ancient drivers to match reality.
|
| 1.45 | 28-Dec-1994 |
mycroft | Increase the st_space() timeout.
|
| 1.44 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.43 | 22-Nov-1994 |
mycroft | Complete last change.
|
| 1.42 | 21-Nov-1994 |
mycroft | Return EIO when the drive is empty.
|
| 1.41 | 21-Nov-1994 |
mycroft | Clean up open and close routines somewhat.
|
| 1.40 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.39 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.38 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.37 | 12-Aug-1994 |
deraadt | un-aligned access is not cool on some processors
|
| 1.36 | 09-Aug-1994 |
mycroft | Change ST_Q_NEEDS_PAGE_0 to something more generic.
|
| 1.35 | 29-Jun-1994 |
cgd | branches: 1.35.2;
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.34 | 19-Jun-1994 |
mycroft | Fix a comment.
|
| 1.33 | 16-Jun-1994 |
chopps | resolve confusion over who owns the buf after calling scsi_scsi_cmd()
|
| 1.32 | 16-Jun-1994 |
mycroft | b_un.b_addr -> b_data
|
| 1.31 | 11-May-1994 |
mycroft | Add dummy *dump() routines.
|
| 1.30 | 09-May-1994 |
chopps | remove union's from sense_data struct, conditionaly define RAW_PART
|
| 1.29 | 24-Apr-1994 |
mycroft | Deal with variable-length drives that don't specify a maximum block size.
c.f. SCSI 2 protocol definition.
|
| 1.28 | 13-Apr-1994 |
mycroft | Various cleanup, but no functional differences.
|
| 1.27 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.26 | 10-Apr-1994 |
mycroft | Add another rogue...
|
| 1.25 | 05-Apr-1994 |
mycroft | This line is redundant for LD_UNLOAD (only one use, and it does this itself)
but breaks LD_RETEN, forcing a reload of the driver state on the next I/O
operation. Therefore, it goes away.
|
| 1.24 | 05-Apr-1994 |
mycroft | Reload the tape after retension.
|
| 1.23 | 05-Apr-1994 |
mycroft | Implement MTRETEN (untested).
|
| 1.22 | 05-Apr-1994 |
mycroft | Implement MTEOM, currently untested.
|
| 1.21 | 01-Apr-1994 |
mycroft | Deal with Mark Weaver's rogue tape drive.
|
| 1.20 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.19 | 06-Feb-1994 |
mycroft | Remove another use of b_actl.
|
| 1.18 | 11-Jan-1994 |
mycroft | *strategy functions return void.
|
| 1.17 | 17-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.16 | 01-Aug-1993 |
mycroft | branches: 1.16.2;
Add RCS identifiers (this time on the correct side of the branch), and
incorporate recent changes in netbsd-0-9 branch.
|
| 1.15 | 28-Jul-1993 |
cgd | incorporate changes from 0-9-base to 0-9-ALPHA
|
| 1.14 | 19-Jul-1993 |
cgd | branches: 1.14.2;
patches from allen briggs to fix a minor bug in *attach()
|
| 1.13 | 09-Jul-1993 |
cgd | fix evil interaction with new physio; don't set error bit for short reads.
|
| 1.12 | 27-Jun-1993 |
andrew | ANSIfications.
|
| 1.11 | 16-Jun-1993 |
andrew | woops - another typo.
|
| 1.10 | 16-Jun-1993 |
deraadt | whoops. typo.
|
| 1.9 | 16-Jun-1993 |
deraadt | fix to intuit the maximum number of scsi units available on a device
driver. this piece at least, should be safe from changing sizeof(dev_t)
|
| 1.8 | 20-May-1993 |
cgd | add rcsids and clean up file headers
|
| 1.7 | 04-May-1993 |
deraadt | support for making dev->id_alive be set, this is for iostat to
find disk devices. wee bit of a kludge. sub-device attach()
routines must now return 1 for successful attach(), 0 otherwise.
Other bsd's do this too..
|
| 1.6 | 22-Apr-1993 |
mycroft | Fix up error messages and return values.
|
| 1.5 | 20-Apr-1993 |
mycroft | Display more meaningful message on SCSI `unit attention'.
|
| 1.4 | 12-Apr-1993 |
deraadt | fixed various bugs like cdattach() returning garbage.
|
| 1.3 | 12-Apr-1993 |
deraadt | new scsi subsystem.
changes also in config/mkioconf.c
i386/isa/wd.c, fd.c, and all scsi drivers.
|
| 1.2 | 08-Apr-1993 |
glass | attempting to open an already opened tape drive should get you EBUSY, not
ENXIO.
|
| 1.1 | 21-Mar-1993 |
cgd | after 0.2.2 "stable" patches applied
|
| 1.14.2.2 | 31-Jul-1993 |
cgd | give names, err, wmesg's, to my "pain" -- i.e. convert sleep() to tsleep()
|
| 1.14.2.1 | 26-Jul-1993 |
briggs | When stioctl() got called with an unknown command, it was not
returning an error code consistently. This change forces it
to return EINVAL when the command is unknown.
|
| 1.16.2.12 | 16-Feb-1994 |
mycroft | More KNF-like.
|
| 1.16.2.11 | 15-Feb-1994 |
mycroft | Reset SDEV_WAITING so we actually wake up the waiting process...
|
| 1.16.2.10 | 06-Feb-1994 |
mycroft | Remove another use of b_actl.
|
| 1.16.2.9 | 01-Feb-1994 |
mycroft | Fix field name clash.
|
| 1.16.2.8 | 01-Feb-1994 |
mycroft | Change some things into arrays.
|
| 1.16.2.7 | 22-Jan-1994 |
briggs | NST -> stcd.cd_ndevs in SC_DEBUG statement.
|
| 1.16.2.6 | 29-Nov-1993 |
mycroft | Return EBUSY, not ENXIO, if alredy open.
|
| 1.16.2.5 | 25-Nov-1993 |
mycroft | Correct attach message.
|
| 1.16.2.4 | 25-Nov-1993 |
mycroft | Correct the display of some messages.
|
| 1.16.2.3 | 25-Nov-1993 |
mycroft | Compiles. Will it work?
|
| 1.16.2.2 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.16.2.1 | 24-Sep-1993 |
mycroft | cd.c, sd.c, st.c: strategy functions return void.
|
| 1.35.2.2 | 06-Oct-1994 |
mycroft | Update from trunk.
|
| 1.35.2.1 | 09-Aug-1994 |
mycroft | update from trunk
|
| 1.70.6.1 | 12-Mar-1997 |
is | Merge in changes from Trunk
|
| 1.71.4.1 | 23-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.73.2.4 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.73.2.3 | 16-Sep-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.73.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.73.2.1 | 27-Aug-1997 |
thorpej | file st.c was added on branch marc-pcmcia on 1997-08-27 23:33:41 +0000
|
| 1.84.2.2 | 30-Sep-1998 |
cgd | pull up rev 1.98 (via patch; doesn't pull up cleanly) from trunk (drochner)
|
| 1.84.2.1 | 07-Nov-1997 |
mellon | Pull rev 1.85 up from trunk (mjacob)
|
| 1.93.2.2 | 08-Aug-1998 |
eeh | Revert cdevsw mmap routines to return int.
|
| 1.93.2.1 | 30-Jul-1998 |
eeh | file st.c was added on branch eeh-paddr_t on 1998-08-08 03:06:52 +0000
|
| 1.110.4.1 | 21-Jun-1999 |
thorpej | Sync w/ -current.
|
| 1.110.2.1 | 23-Jan-2000 |
he | Pull up revision 1.118 (requested by mjacob):
Recognize (indirectly) HP4000s via density code and do the right
QIC type dance. Fixes PR#9271.
|
| 1.114.8.1 | 21-Dec-1999 |
wrstuden | Initial commit of recent changes to make DEV_BSIZE go away.
Runs on i386, needs work on other arch's. Main kernel routines should be
fine, but a number of the stand programs need help.
cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512
byte block devices. vnd, raidframe, and lfs need work.
Non 2**n block support is automatic for LKM's and conditional for kernels
on "options NON_PO2_BLOCKS".
|
| 1.114.2.8 | 29-Mar-2001 |
simonb | Make this compile for the non-SCSIVERBOSE case.
|
| 1.114.2.7 | 11-Feb-2001 |
bouyer | Sync with HEAD.
|
| 1.114.2.6 | 08-Dec-2000 |
bouyer | Sync with HEAD.
|
| 1.114.2.5 | 22-Nov-2000 |
bouyer | Sync with HEAD.
|
| 1.114.2.4 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.114.2.3 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.114.2.2 | 20-Oct-1999 |
thorpej | Remove an unnecessary comment regarding XS_CTL_NOSLEEP.
|
| 1.114.2.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.122.2.1 | 22-Jun-2000 |
minoura | Sync w/ netbsd-1-5-base.
|
| 1.124.2.1 | 03-Nov-2000 |
tv | Pullup 1.127 [pk]:
Remove useless check for SDEV_MEDIA_LOADED from st_read_block_limits().
This command should work whether or not media is present in the device.
|
| 1.134.2.10 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.134.2.9 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.134.2.8 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.134.2.7 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.134.2.6 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.134.2.5 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.134.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.134.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.134.2.2 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.134.2.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.142.2.6 | 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
|
| 1.142.2.5 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.142.2.4 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.142.2.3 | 11-Feb-2002 |
jdolecek | Sync w/ -current.
|
| 1.142.2.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.142.2.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.143.2.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.143.2.1 | 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).
|
| 1.155.2.2 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.155.2.1 | 16-May-2002 |
gehenna | Add the block/character device switches.
Replace the direct-access to devsw table with calling devsw API.
|
| 1.163.4.1 | 11-Sep-2004 |
he | Pull up revisions 1.165-1.166 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.163.2.10 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.163.2.9 | 01-Apr-2005 |
skrll | Sync with HEAD.
|
| 1.163.2.8 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.163.2.7 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.163.2.6 | 19-Oct-2004 |
skrll | Sync with HEAD
|
| 1.163.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.163.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.163.2.3 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.163.2.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.163.2.1 | 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.
|
| 1.173.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.173.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.178.2.3 | 02-Dec-2007 |
riz | Pull up following revision(s) (requested by bouyer in ticket #1880):
sys/dev/scsipi/st.c: revision 1.199-1.201 via patch
Properly set b_resid in an error case.
Clamp b_resid to b_bcount in case the info field is larger. Fix a KASSERT
in physio() when a media error occurs on my SDLT320 drive (maybe this drive
is bogus for not setting the highter bits to 0 in the INFO field in this
case).
When checking for invalid b_resid valyes, also check negative ones.
Should fix kern/36690.
Move check for b_resid value to the right place. Should definitively fix
kern/36690.
|
| 1.178.2.2 | 17-Sep-2007 |
msaitoh | Pull up following revision(s) (requested by bouyer in ticket #1837):
sys/dev/scsipi/st.c: revision 1.180
revert the only part of rev. 1.177 which made a functional
change -- it broke error reporting because st->asc(q) are
not set
(actually, I believe that SSD_RCODE_VALID is misnamed)
|
| 1.178.2.1 | 06-Apr-2005 |
tron | Pull up revision 1.179 (requested by yamt in ticket #112):
introduce a function to drain bufq and use it where appropriate.
|
| 1.181.2.4 | 27-Oct-2007 |
yamt | sync with head.
|
| 1.181.2.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.181.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.181.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.186.12.2 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.186.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.186.10.2 | 11-May-2006 |
elad | sync with head
|
| 1.186.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.186.8.4 | 03-Sep-2006 |
yamt | sync with head.
|
| 1.186.8.3 | 26-Jun-2006 |
yamt | sync with head.
|
| 1.186.8.2 | 24-May-2006 |
yamt | sync with head.
|
| 1.186.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.186.6.2 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.186.6.1 | 04-Feb-2006 |
simonb | Adapt for timecounters: mostly use get*time() and use "time_second"
instead of "time.tv_sec".
|
| 1.186.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.190.2.1 | 19-Jun-2006 |
chap | Sync with head.
|
| 1.192.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.192.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.192.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.194.8.1 | 06-Jan-2008 |
wrstuden | Catch up to netbsd-4.0 release.
|
| 1.194.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.194.2.1 | 25-Nov-2007 |
xtraeme | Pull up following revision(s) (requested by bouyer in ticket #1002):
sys/dev/scsipi/st.c: revision 1.199 (patch)
sys/dev/scsipi/st.c: revision 1.200 (patch)
sys/dev/scsipi/st.c: revision 1.201 (patch)
Properly set b_resid in an error case.
Clamp b_resid to b_bcount in case the info field is larger. Fix a KASSERT
in physio() when a media error occurs on my SDLT320 drive (maybe this drive
is bogus for not setting the highter bits to 0 in the INFO field in this
case).
When checking for invalid b_resid valyes, also check negative ones.
Should fix kern/36690.
Move check for b_resid value to the right place. Should definitively fix
kern/36690.
|
| 1.196.4.1 | 11-Jul-2007 |
mjf | Sync with head.
|
| 1.196.2.7 | 12-Oct-2007 |
ad | Sync with head.
|
| 1.196.2.6 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.196.2.5 | 09-Oct-2007 |
ad | Sync with head.
|
| 1.196.2.4 | 19-Aug-2007 |
ad | - Back out the biodone() changes.
- Eliminate B_ERROR (from HEAD).
|
| 1.196.2.3 | 15-Jul-2007 |
ad | Sync with head.
|
| 1.196.2.2 | 01-Jul-2007 |
ad | Adapt to callout API change.
|
| 1.196.2.1 | 13-May-2007 |
ad | - Pass the error number and residual count to biodone(), and let it handle
setting error indicators. Prepare to eliminate B_ERROR.
- Add a flag argument to brelse() to be set into the buf's flags, instead
of doing it directly. Typically used to set B_INVAL.
- Add a "struct cpu_info *" argument to kthread_create(), to be used to
create bound threads. Change "bool mpsafe" to "int flags".
- Allow exit of LWPs in the IDL state when (l != curlwp).
- More locking fixes & conversion to the new API.
|
| 1.197.2.1 | 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.198.10.2 | 29-Jul-2007 |
ad | It's not a good idea for device drivers to modify b_flags, as they don't
need to understand the locking around that field. Instead of setting
B_ERROR, set b_error instead. b_error is 'owned' by whoever completes
the I/O request.
|
| 1.198.10.1 | 29-Jul-2007 |
ad | file st.c was added on branch matt-mips64 on 2007-07-29 12:50:24 +0000
|
| 1.198.8.2 | 14-Oct-2007 |
yamt | sync with head.
|
| 1.198.8.1 | 06-Oct-2007 |
yamt | sync with head.
|
| 1.198.6.1 | 06-Nov-2007 |
matt | sync with HEAD
|
| 1.198.4.3 | 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.
|
| 1.198.4.2 | 07-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.198.4.1 | 02-Oct-2007 |
joerg | Sync with HEAD.
|
| 1.202.18.5 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.202.18.4 | 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.202.18.3 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.202.18.2 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.202.18.1 | 05-Apr-2008 |
mjf | - add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing
as these are always needed.
- convert many, many drivers over to the New Devfs World Order. For a
list of device drivers yet to be converted see,
http://www.netbsd.org/~mjf/devfs-todo.html.
- add a new device_unregister_all(device_t) function to remove all device
names associated with a device_t, which saves us having to construct
device names when the driver is detached.
- add a DEV_AUDIO type for devices.
|
| 1.203.4.6 | 11-Aug-2010 |
yamt | sync with head.
|
| 1.203.4.5 | 11-Mar-2010 |
yamt | sync with head
|
| 1.203.4.4 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.203.4.3 | 16-May-2009 |
yamt | sync with head
|
| 1.203.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.203.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.203.2.2 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.203.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.204.2.2 | 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.204.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.205.4.2 | 13-Dec-2008 |
haad | Update haad-dm branch to haad-dm-base2.
|
| 1.205.4.1 | 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.207.4.1 | 20-Nov-2008 |
snj | Pull up following revision(s) (requested by bouyer in ticket #78):
sys/dev/scsipi/st.c: revision 1.208
If we return an error make sure we don't return a b_resid of 0; or this
will trigger a KASSERT in physio_done(). Fix issue reported by Todd Kover
on tech-kern@
|
| 1.207.2.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.210.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.216.4.1 | 03-Jul-2010 |
rmind | sync with head
|
| 1.216.2.1 | 17-Aug-2010 |
uebayasi | Sync with HEAD.
|
| 1.217.12.5 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.217.12.4 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.217.12.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.217.12.2 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.217.12.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.217.8.3 | 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")
|
| 1.217.8.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.217.8.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.218.2.2 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.218.2.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #64):
sys/dev/scsipi/stvar.h: revision 1.24
sys/dev/scsipi/st_scsi.c: revision 1.34
sys/dev/scsipi/st_atapi.c: revision 1.28
sys/dev/scsipi/st.c: revision 1.220
Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.221.4.1 | 18-May-2014 |
rmind | sync with head
|
| 1.221.2.2 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.221.2.1 | 20-Aug-2014 |
tls | Rebase to HEAD as of a few days ago.
|
| 1.223.2.2 | 10-Aug-2014 |
tls | Rebase.
|
| 1.223.2.1 | 07-Apr-2014 |
tls | Be a little more clear and consistent about harvesting entropy from devices:
1) deprecate RND_FLAG_NO_ESTIMATE
2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE
3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE
4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME|
RND_FLAG_COLLECT_VALUE|RND_FLAG_ESTIMATE_TIME
5) Make entropy harvesting from environmental sensors a little more generic
and remove it from individual sensor drivers.
6) Remove individual open-coded delta-estimators for values from a few
places in the tree (uvm, environmental drivers).
7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers
that had stubbed out code, other minor cleanups.
|
| 1.226.10.1 | 07-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1682):
sys/dev/scsipi/st.c: revision 1.236 (patch)
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.226.6.1 | 07-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1682):
sys/dev/scsipi/st.c: revision 1.236 (patch)
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.226.4.4 | 28-Aug-2017 |
skrll | Sync with HEAD
|
| 1.226.4.3 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.226.4.2 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.226.4.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.226.2.1 | 07-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1682):
sys/dev/scsipi/st.c: revision 1.236 (patch)
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.228.2.1 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.230.8.4 | 04-Nov-2019 |
martin | Pull up following revision(s) (requested by jnemeth in ticket #1421):
sys/dev/scsipi/st.c: revision 1.238
Add simple position recovery when positioning to EOM by reading
the position with READ_POSITION.
this allows for
mt eom
mt st
to return the correct file position.
|
| 1.230.8.3 | 01-Mar-2019 |
martin | Pull up following revision(s) (requested by kardel in ticket #1197):
sys/dev/scsipi/st.c: revision 1.236
sys/dev/scsipi/st.c: revision 1.237
sys/dev/scsipi/files.scsipi: revision 1.42
Fix PR kern/53949:
Fix inconsistent/incomplete file mark handling to conform again
to mtio(4) at close(2) time. This was necessary as the PREVENT/ALLOW
bracket was reduced from a whole mount session to cover only the
open(2)/close(2) time on ~2002-03-22. The rationale was to allow
robots and humans to change the media during a mount session.
Unfortunately this lead to file marks being written to potentially other
media at the beginning on drives that used the two file marks as EOM
pattern. In order for that to happen the media had to be removed after
data and at most one file mark had been written before removal.
The mount error message has been clarified and a warning about
potential data/file mark lossage on UNIT ATTENTION
during an active mount session with unfinished file marks has been
added.
While there, fix, but disable the commented SUN compatibility to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
Additionally debug output coverage has been extended.
-
Correct printing type of b_blkno (int64_t) in st.c
Fixes build with kUBSan on NetBSD/i386.
Fix, but disable the commented SUN compatibility in st.c to write
final file marks by opening and immediately closing the device
in O_WRONLY mode. That code has not been working since around 1998.
It can now be enabled with options ST_SUNCOMPAT.
|
| 1.230.8.2 | 08-Apr-2018 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #703):
sys/dev/scsipi/st.c: 1.234
sys/dev/scsipi/stvar.h: 1.26
Use separate lock to protect internal state and release locks when
calling biodone.
|
| 1.230.8.1 | 21-Jun-2017 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #53):
sys/dev/scsipi/atapiconf.c: revision 1.91
sys/dev/scsipi/cd.c: revision 1.341
sys/dev/scsipi/scsi_base.c: revision 1.92
sys/dev/scsipi/scsiconf.c: revision 1.280
sys/dev/scsipi/scsipi_base.c: revisions 1.176, 1.177
sys/dev/scsipi/sd.c: revision 1.325
sys/dev/scsipi/ss.c: revision 1.89
sys/dev/scsipi/st.c: revision 1.231
The atapibus detach path did hold the channel mutex while calling into autoconf,
which would trigger a panic when unplugging a USB ATAPI CDROM.
Align detach code for scsibus and atapibus to fix this.
Also avoid races when detaching devices by replacing callout_stop with
callout_halt.
--
pass config_detach error to caller.
|
| 1.231.4.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.234.2.3 | 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.234.2.2 | 08-Apr-2020 |
martin | Merge changes from current as of 20200406
|
| 1.234.2.1 | 10-Jun-2019 |
christos | Sync with HEAD
|
| 1.240.12.1 | 17-Jun-2021 |
thorpej | Sync w/ HEAD.
|
| 1.31 | 14-Jul-2016 |
msaitoh | - Use aprint*() instead of printf() in xxx_attach().
- Add missing aprint_naive("\n");
- KNF
|
| 1.30 | 24-Aug-2015 |
pooka | would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.29 | 19-Apr-2012 |
bouyer | branches: 1.29.2; 1.29.16;
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.28 | 28-Feb-2012 |
mbalmer | Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.27 | 02-Feb-2012 |
tls | branches: 1.27.2;
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.
|
| 1.26 | 06-Dec-2009 |
dyoung | branches: 1.26.12; 1.26.16;
Delete do-nothing device-activation hooks.
|
| 1.25 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.24 | 19-Oct-2009 |
bouyer | Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !
|
| 1.23 | 15-Aug-2009 |
pgoyette | 1. Move the mode_select functionality into common code (in st.c) and
invoke the common routine for both scsi and atapi tapes.
2. Replace a numeric constant with some sizeof's when calculating the
size of the mode_select command buffer, clear the entire buffer, and
KASSERT to ensure the page_0_size loaded from quirk table is valid.
3. Add a quirk for my Seagate Travan-40 tape drive.
As discussed on tech-kern@
Addresses my PR kern/34832
|
| 1.22 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.21 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.20 | 16-Nov-2006 |
christos | branches: 1.20.52; 1.20.68;
__unused removal on arguments; approved by core.
|
| 1.19 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.18 | 30-Mar-2006 |
thorpej | branches: 1.18.8; 1.18.10;
Use device_private().
|
| 1.17 | 11-Dec-2005 |
christos | branches: 1.17.4; 1.17.6; 1.17.8; 1.17.10; 1.17.12;
merge ktrace-lwp.
|
| 1.16 | 29-May-2005 |
christos | branches: 1.16.2;
- Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.15 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.14 | 28-Oct-2004 |
yamt | branches: 1.14.4; 1.14.6;
move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.13 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.12 | 05-Oct-2003 |
bouyer | Remove references to University of California from my copyright notices.
|
| 1.11 | 02-Oct-2002 |
thorpej | branches: 1.11.6;
Add trailing ; to CFATTACH_DECL.
|
| 1.10 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.9 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.8 | 23-Apr-2002 |
bouyer | More copyright fixes, pointed out by Thomas. Thanks !
|
| 1.7 | 07-Dec-2001 |
yamt | add detach support for st.
|
| 1.6 | 01-Dec-2001 |
bouyer | Various quirks for the ATAPI OnStream DI-30, mostly from the FreeBSD driver.
Many thanks to Chris Pinnock for giving me remote access to his hardware.
|
| 1.5 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.4 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.3 | 18-Jun-2001 |
bouyer | branches: 1.3.2; 1.3.4;
Add my copyrigth and remove the older ones. None of the original authors of
the st driver wrote something in this file.
|
| 1.2 | 18-Jun-2001 |
bouyer | Snapshot of ATAPI tapes support. Known to be able to read/write to tape with:
st0 at atapibus0 drive 1: <Seagate STT8000A, , 5.51> type 1 sequential
removable
Major changes may still happen in order to properly support other ATAPI
tape drives.
|
| 1.1 | 04-May-2001 |
bouyer | Add an ATAPI front-end to the st driver. Completely untested for now,
but st at scsi should still work :)
|
| 1.3.4.3 | 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
|
| 1.3.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.3.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.3.2.6 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.3.2.5 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.3.2.4 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.3.2.3 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.3.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.3.2.1 | 18-Jun-2001 |
nathanw | file st_atapi.c was added on branch nathanw_sa on 2001-06-21 20:06:06 +0000
|
| 1.11.6.7 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.11.6.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.11.6.5 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.11.6.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.11.6.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.11.6.2 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.11.6.1 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.14.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.14.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.16.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.16.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.17.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.17.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.17.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.17.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.17.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.18.10.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.18.10.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.18.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.20.68.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.20.52.3 | 11-Mar-2010 |
yamt | sync with head
|
| 1.20.52.2 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.20.52.1 | 16-May-2009 |
yamt | sync with head
|
| 1.26.16.5 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.26.16.4 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.26.16.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.26.16.2 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.26.16.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.26.12.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.26.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.27.2.2 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.27.2.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #64):
sys/dev/scsipi/stvar.h: revision 1.24
sys/dev/scsipi/st_scsi.c: revision 1.34
sys/dev/scsipi/st_atapi.c: revision 1.28
sys/dev/scsipi/st.c: revision 1.220
Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.29.16.2 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.29.16.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.29.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.37 | 24-Aug-2015 |
pooka | would you like some freshly ground _KERNEL_OPT with that?
yes? excellent choice, sir/madam.
|
| 1.36 | 18-Oct-2014 |
snj | branches: 1.36.2;
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!
|
| 1.35 | 19-Apr-2012 |
bouyer | branches: 1.35.2;
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.34 | 28-Feb-2012 |
mbalmer | Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.33 | 02-Feb-2012 |
tls | branches: 1.33.2;
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.
|
| 1.32 | 06-Dec-2009 |
dyoung | branches: 1.32.12; 1.32.16;
Delete do-nothing device-activation hooks.
|
| 1.31 | 21-Oct-2009 |
rmind | Remove uarea swap-out functionality:
- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.
Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).
Discussed on <tech-kern>, reviewed by <ad>.
|
| 1.30 | 15-Aug-2009 |
pgoyette | 1. Move the mode_select functionality into common code (in st.c) and
invoke the common routine for both scsi and atapi tapes.
2. Replace a numeric constant with some sizeof's when calculating the
size of the mode_select command buffer, clear the entire buffer, and
KASSERT to ensure the page_0_size loaded from quirk table is valid.
3. Add a quirk for my Seagate Travan-40 tape drive.
As discussed on tech-kern@
Addresses my PR kern/34832
|
| 1.29 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.28 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.27 | 28-Apr-2008 |
martin | branches: 1.27.14;
Remove clause 3 and 4 from TNF licenses
|
| 1.26 | 16-Nov-2006 |
christos | branches: 1.26.48; 1.26.50; 1.26.52;
__unused removal on arguments; approved by core.
|
| 1.25 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.24 | 30-Mar-2006 |
thorpej | branches: 1.24.8; 1.24.10;
Use device_private().
|
| 1.23 | 11-Dec-2005 |
christos | branches: 1.23.4; 1.23.6; 1.23.8; 1.23.10; 1.23.12;
merge ktrace-lwp.
|
| 1.22 | 30-May-2005 |
christos | branches: 1.22.2;
- remove bogus casts
- add more const
|
| 1.21 | 27-Feb-2005 |
perry | nuke trailing whitespace
|
| 1.20 | 21-Feb-2005 |
thorpej | Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name
everything "scsi_*", since we really are talking about the SCSI command
set, ATAPI transport not withstanding. Improve the names of many structures,
and prepend "SCSI_" onto all SCSI command opcodes. Place items described
by the SCSI Primary Commands document into scsi_spc.h.
|
| 1.19 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.18 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.17 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.16 | 28-Oct-2004 |
yamt | branches: 1.16.4; 1.16.6;
move buffer queue related stuffs from buf.h to their own header, bufq.h.
|
| 1.15 | 18-Sep-2004 |
mycroft | Minor rearrangement. Whitespace and #include cleanup.
|
| 1.14 | 18-Sep-2004 |
mycroft | Standardize some variable names and the calling pattern for scsipi_command().
Use void pointer casts.
|
| 1.13 | 17-Sep-2004 |
mycroft | Remove the "xfer" argument to scsipi_command().
|
| 1.12 | 09-Sep-2004 |
bouyer | Make the xxstart() functions reentrant again, as some drivers HBA can call
scsipi_done() from their scsipi_request().
For this, add a struct scsipi_xfer * argument to scsipi_command().
If not NULL scsipi_command() will use this to enqueue this xfer, otherwise
it'll try to allocate a new one. This scsipi_xfer has to be allocated
and initialised by scsipi_make_xs() or equivalent.
In xxstart(), allocate a scsipi_xfer using scsipi_make_xs(), and if not NULL,
dequeue the buffer before calling scsipi_command(). This makes sure that
scsipi_command() will not fail, and also makes sure that xxstart() won't
be called again between the BUFQ_PEEK() and BUFQ_GET().
Fix "dequeued wrong buf" panics reported by Juergen Hannken-Illjes in
private mail and Andreas Wrede on current-users@.
Thanks to Jason Thorpe and Chuck Silver for review, and Andreas Wrede for
testing the patch.
|
| 1.11 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.10 | 02-Oct-2002 |
thorpej | branches: 1.10.6; 1.10.8;
Add trailing ; to CFATTACH_DECL.
|
| 1.9 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.8 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.7 | 07-Dec-2001 |
yamt | add detach support for st.
|
| 1.6 | 15-Nov-2001 |
lukem | don't need <sys/types.h> when including <sys/param.h>
|
| 1.5 | 13-Nov-2001 |
lukem | add RCSIDs
|
| 1.4 | 18-Jul-2001 |
thorpej | bcopy -> memcpy
|
| 1.3 | 18-Jul-2001 |
thorpej | bzero -> memset
|
| 1.2 | 18-Jun-2001 |
bouyer | branches: 1.2.2; 1.2.4;
Snapshot of ATAPI tapes support. Known to be able to read/write to tape with:
st0 at atapibus0 drive 1: <Seagate STT8000A, , 5.51> type 1 sequential
removable
Major changes may still happen in order to properly support other ATAPI
tape drives.
|
| 1.1 | 04-May-2001 |
bouyer | Add an ATAPI front-end to the st driver. Completely untested for now,
but st at scsi should still work :)
|
| 1.2.4.3 | 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
|
| 1.2.4.2 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.2.4.1 | 03-Aug-2001 |
lukem | update to -current
|
| 1.2.2.6 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.5 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.4 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.2.2.3 | 24-Aug-2001 |
nathanw | Catch up with -current.
|
| 1.2.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.2.2.1 | 18-Jun-2001 |
nathanw | file st_scsi.c was added on branch nathanw_sa on 2001-06-21 20:06:07 +0000
|
| 1.10.8.1 | 11-Sep-2004 |
he | Pull up revision 1.12 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.10.6.6 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.10.6.5 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.10.6.4 | 02-Nov-2004 |
skrll | Sync with HEAD.
|
| 1.10.6.3 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.10.6.2 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.10.6.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.16.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.16.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.22.2.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.22.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.23.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.23.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.23.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.23.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.23.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.24.10.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.24.10.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.24.8.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.26.52.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.26.52.3 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.26.52.2 | 16-May-2009 |
yamt | sync with head
|
| 1.26.52.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.26.50.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.26.48.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.27.14.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.32.16.5 | 29-Apr-2012 |
mrg | sync to latest -current.
|
| 1.32.16.4 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.32.16.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.32.16.2 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.32.16.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.32.12.2 | 23-May-2012 |
yamt | sync with head.
|
| 1.32.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.33.2.2 | 23-Apr-2012 |
riz | Pull up following revision(s) (requested by bouyer in ticket #192):
sys/dev/scsipi/cd.c: revision 1.307
sys/dev/scsipi/scsiconf.c: revision 1.266
sys/dev/scsipi/sd.c: revision 1.298
sys/dev/scsipi/st_scsi.c: revision 1.35
sys/dev/scsipi/atapiconf.c: revision 1.85
sys/dev/scsipi/scsipiconf.h: revision 1.120
sys/dev/usb/umass_scsipi.c: revision 1.44
sys/dev/scsipi/scsiconf.h: revision 1.57
sys/dev/scsipi/st_atapi.c: revision 1.29
sys/dev/scsipi/scsipi_base.c: revision 1.158
sys/dev/scsipi/st.c: revision 1.221
sys/dev/scsipi/scsipi_ioctl.c: revision 1.67
Expand struct scsipi_bustype {} in a ABI-backward-compatible way to
pass more informations about the bus:
- bustype_type has 2 different bytes, one holding the existing
SCSIPI_BUSTYPE_* (scsi, atapi, ata), and one for a per-SCSIPI_BUSTYPE_*
subtype. Introduce macros to build or extract bustype_type.
- for SCSIPI_BUSTYPE_SCSI, define subtypes for parallel SCSI, Fibre Channel,
SAS and USB, to specify the transport method. SCSIPI_BUSTYPE_SCSI_PSCSI
is 0 so that bustype_type value doesn't change for existing code
- for non-SCSIPI_BUSTYPE_SCSI busses there's no defined subtype yet,
so the bustype_type value doesn't change.
- provide scsi_fc_bustype, scsi_sas_bustype and scsi_usb_bustype
along with scsi_bustype to be used by bus driver where appropriate
- scsipi_print_xfer_mode(): more existing code under a
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_PSCSI) case, as
sync/wide parameters only make sense for parallel SCSI.
For (SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_FC) and
(SCSIPI_BUSTYPE_SCSI, SCSIPI_BUSTYPE_SCSI_SAS), only print
tagged queing status if enabled. Just be silent for other
bustypes.
This change is prompted by this problem:
right now, FC (e.g. isp(4)) and SAS (e.g. mfi(4)) don't
do anything for ADAPTER_REQ_SET_XFER_MODE, and especially never
call scsipi_async_event(ASYNC_EVENT_XFER_MODE), so sd(4) always
runs untagged. Doing a scsipi_async_event(ASYNC_EVENT_XFER_MODE) with
appropriate parameters is enough to enable tagged queuing,
but then scsipi will print:
sd0: async, 8-bit transfers, tagged queueing
which is harmless (async, 8-bit transfers doens't make sense on SAS anyway)
but will confuse users. With this change scsipi will only print:
sd0: tagged queueing
which is correct.
In the long run, knowning the underlying transport in scsipi will
allow better handling of device which are not parallel SCSI.
Another change adding an extra callback to struct scsipi_bustype {}
will come (so that scsipi_print_xfer_mode(), which is SCSI-specific,
can be moved out of scsipi_base, and split into per-subtype callback),
but this will break kernel ABI and so is not suitable for
netbsd-6, so will be commmited later. The above is enough to get
tagged queuing on FC and SAS in netbsd-6.
|
| 1.33.2.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #64):
sys/dev/scsipi/stvar.h: revision 1.24
sys/dev/scsipi/st_scsi.c: revision 1.34
sys/dev/scsipi/st_atapi.c: revision 1.28
sys/dev/scsipi/st.c: revision 1.220
Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.35.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.36.2.1 | 22-Sep-2015 |
skrll | Sync with HEAD
|
| 1.26 | 24-Mar-2018 |
mlelstv | Use separate lock to protect internal state and release locks when
calling biodone.
|
| 1.25 | 13-Apr-2015 |
riastradh | branches: 1.25.10; 1.25.16;
Convert sys/dev to use <sys/rndsource.h>.
|
| 1.24 | 28-Feb-2012 |
mbalmer | branches: 1.24.2; 1.24.16;
Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.23 | 02-Feb-2012 |
tls | branches: 1.23.2;
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.
|
| 1.22 | 19-Nov-2011 |
tls | branches: 1.22.2;
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.
|
| 1.21 | 06-Dec-2009 |
dyoung | branches: 1.21.12;
Delete do-nothing device-activation hooks.
|
| 1.20 | 15-Aug-2009 |
pgoyette | 1. Move the mode_select functionality into common code (in st.c) and
invoke the common routine for both scsi and atapi tapes.
2. Replace a numeric constant with some sizeof's when calculating the
size of the mode_select command buffer, clear the entire buffer, and
KASSERT to ensure the page_0_size loaded from quirk table is valid.
3. Add a quirk for my Seagate Travan-40 tape drive.
As discussed on tech-kern@
Addresses my PR kern/34832
|
| 1.19 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.18 | 28-Apr-2008 |
martin | branches: 1.18.14;
Remove clause 3 and 4 from TNF licenses
|
| 1.17 | 14-Apr-2006 |
blymn | branches: 1.17.58; 1.17.60; 1.17.62;
Make i/o statistics collection more generic, include tape drives and
nfs mounts in the set of devices that statistics will be reported on.
|
| 1.16 | 11-Dec-2005 |
christos | branches: 1.16.4; 1.16.6; 1.16.8; 1.16.10; 1.16.12;
merge ktrace-lwp.
|
| 1.15 | 15-Oct-2005 |
yamt | - change the way to specify a bufq strategy. (by string rather than by number)
- rather than embedding bufq_state in driver softc,
have a pointer to the former.
- move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c.
- rename method to strategy for consistency.
- move some definitions which don't need to be exposed to the rest of kernel
from sys/bufq.h to sys/bufq_impl.h.
(is it better to move it to kern/ or somewhere?)
- fix some obvious breakage in dev/qbus/ts.c. (not tested)
|
| 1.14 | 07-Aug-2005 |
blymn | Add tape statistics structure pointer.
|
| 1.13 | 29-May-2005 |
christos | branches: 1.13.2;
- Sprinkle const
- Avoid variable shadowing.
- Eliminate some caddr_t abuse.
|
| 1.12 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.11 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.10 | 31-Jan-2005 |
reinoud | As part of cleaning up sys/scsipi, replace all u_char by uint8_t and
replace all `short' with int16_t.
|
| 1.9 | 27-Aug-2004 |
bouyer | branches: 1.9.4;
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
Havard Eidnes's analysis of this problem is that the scsipi_xfer pool is
competing for resources with other pools, including the the inode and vnode
pools which can grow quite large.
*_scsipi_cmd(): don't biodone the buffer if scsipi_make_xs() fails, let the
caller deal with the problem
start function of block devices drivers: dequeue the buffer after the
scsipi_command() call. If scsipi_command() fails with ENOMEM don't dequeue
the buffer, and schedule a callout to call the start function after
some delay.
scsipi_init(): prime the scsipi_xfer_pool with one page. This ensure that
there is always some scsipi_xfer to play with. If scsipi_command() fails
because of pool_get(), we're sure there will be resources available later,
when the pending commands have completed.
Reviewed by Jason Thorpe and Havard Eidnes.
Todo: remove the "unable to allocate scsipi_xfer" and "not queued, error %d"
printfs, but I choose to keep them for now, to help make sure the code does
what it should.
|
| 1.8 | 21-Aug-2004 |
thorpej | Use ANSI function decls and make use of static.
|
| 1.7 | 21-Aug-2004 |
thorpej | - De-__P.
- Use ANSI function decls.
|
| 1.6 | 22-Jul-2002 |
hannken | branches: 1.6.6; 1.6.8;
Convert to new device buffer queue interface.
|
| 1.5 | 20-Mar-2002 |
christos | branches: 1.5.4;
Add the ability to report file number/block number. From Matt Jacob.
|
| 1.4 | 07-Dec-2001 |
yamt | add detach support for st.
|
| 1.3 | 01-Dec-2001 |
bouyer | Various quirks for the ATAPI OnStream DI-30, mostly from the FreeBSD driver.
Many thanks to Chris Pinnock for giving me remote access to his hardware.
|
| 1.2 | 18-Jun-2001 |
bouyer | branches: 1.2.2; 1.2.4;
Snapshot of ATAPI tapes support. Known to be able to read/write to tape with:
st0 at atapibus0 drive 1: <Seagate STT8000A, , 5.51> type 1 sequential
removable
Major changes may still happen in order to properly support other ATAPI
tape drives.
|
| 1.1 | 04-May-2001 |
bouyer | Add an ATAPI front-end to the st driver. Completely untested for now,
but st at scsi should still work :)
|
| 1.2.4.3 | 06-Sep-2002 |
jdolecek | sync kqueue branch with HEAD
|
| 1.2.4.2 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.2.4.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.2.2.5 | 01-Aug-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.4 | 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.2.2.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.2.2.2 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.2.2.1 | 18-Jun-2001 |
nathanw | file stvar.h was added on branch nathanw_sa on 2001-06-21 20:06:07 +0000
|
| 1.5.4.1 | 29-Aug-2002 |
gehenna | catch up with -current.
|
| 1.6.8.1 | 11-Sep-2004 |
he | Pull up revision 1.9 (via patch, requested by bouyer in ticket #837):
Improve handling of memory shortage, to fix problems like:
sd3(mpt0:0:1:0): unable to allocate scsipi_xfer
sd3: not queued, error 12
The theory is that other consumers of pool memory is causing
this memory shortage in certain somewhat hard to reproduce
situations.
This is done by giving scsipi_command an extra argument to
optionally pass a preallocated scsipi_xfer, and allocating a
scsipi_xfer before dequeueing a buffer in the various *start()
functions. If the allocation of a scsipi_xfer fails, schedule
a callout for delayed invocation of the start function. Also
reserve one page for scsipi_xfer structs, to ensure that we will
eventually have some available once pending commands complete.
Should fix PR#25670.
|
| 1.6.6.5 | 10-Nov-2005 |
skrll | Sync with HEAD. Here we go again...
|
| 1.6.6.4 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.6.6.3 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.6.6.2 | 03-Sep-2004 |
skrll | Sync with HEAD
|
| 1.6.6.1 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.9.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.13.2.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.16.12.1 | 24-May-2006 |
tron | Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
|
| 1.16.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.16.8.1 | 24-May-2006 |
yamt | sync with head.
|
| 1.16.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.16.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.17.62.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.17.62.3 | 19-Aug-2009 |
yamt | sync with head.
|
| 1.17.62.2 | 16-May-2009 |
yamt | sync with head
|
| 1.17.62.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.17.60.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.17.58.1 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.18.14.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.21.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.22.2.4 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.22.2.3 | 06-Mar-2012 |
mrg | sync to -current
|
| 1.22.2.2 | 04-Mar-2012 |
mrg | sync to latest -current.
|
| 1.22.2.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.23.2.1 | 02-Mar-2012 |
riz | Pull up following revision(s) (requested by mbalmer in ticket #64):
sys/dev/scsipi/stvar.h: revision 1.24
sys/dev/scsipi/st_scsi.c: revision 1.34
sys/dev/scsipi/st_atapi.c: revision 1.28
sys/dev/scsipi/st.c: revision 1.220
Convert st(4) to device_t, while here clean up the code and use uintXX_t
instead of u_intXX_t.
|
| 1.24.16.1 | 06-Jun-2015 |
skrll | Sync with HEAD
|
| 1.24.2.1 | 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.25.16.1 | 30-Mar-2018 |
pgoyette | Resolve conflicts between branch and HEAD
|
| 1.25.10.1 | 08-Apr-2018 |
snj | Pull up following revision(s) (requested by mlelstv in ticket #703):
sys/dev/scsipi/st.c: 1.234
sys/dev/scsipi/stvar.h: 1.26
Use separate lock to protect internal state and release locks when
calling biodone.
|
| 1.4 | 16-May-2002 |
thorpej | Garbage-collect the empty-for-years "su" device.
|
| 1.3 | 27-Aug-1997 |
bouyer | branches: 1.3.2; 1.3.32; 1.3.34; 1.3.46;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.2 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.3.46.1 | 30-May-2002 |
gehenna | Catch up with -current.
|
| 1.3.34.1 | 23-Jun-2002 |
jdolecek | catch up with -current on kqueue branch
|
| 1.3.32.1 | 20-Jun-2002 |
nathanw | Catch up to -current.
|
| 1.3.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.3.2.1 | 27-Aug-1997 |
thorpej | file su.c was added on branch marc-pcmcia on 1997-08-27 23:33:42 +0000
|
| 1.65 | 20-Nov-2016 |
mlelstv | Make scsipi framework MPSAFE.
Data structures are now protected by a per-adapter mutex at IPL_BIO
that is created by the scsibus or atapibus instance when the adapter
is configured.
The enable reference counter and the channel freeze counter which are
currently used by HBA code before the adapter is configured, are made
atomic.
The target drivers are now all tagged as D_MPSAFE.
Almost all HBA drivers still require the kernel lock to present,
so all callbacks into HBA code are still protected by kernel lock
unless the driver is tagged as SCSIPI_ADAPT_MPSAFE.
TODO: refactor sd and cd to use dksubr.
|
| 1.64 | 05-Aug-2016 |
maya | uk(4) is safe to suspend, register it as such
|
| 1.63 | 14-Jul-2016 |
msaitoh | branches: 1.63.2;
- Use aprint*() instead of printf() in xxx_attach().
- Add missing aprint_naive("\n");
- KNF
|
| 1.62 | 25-Jul-2014 |
dholland | branches: 1.62.4;
Add d_discard to all struct cdevsw instances I could find.
All have been set to "nodiscard"; some should get a real implementation.
|
| 1.61 | 16-Mar-2014 |
dholland | branches: 1.61.2;
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.
|
| 1.60 | 08-Feb-2012 |
mbalmer | branches: 1.60.6; 1.60.10;
Convert uk(4) to device_t. While here, clean up the code a bit.
releng ok (ack 3, nak 0, wait no more)
|
| 1.59 | 06-Dec-2009 |
dyoung | branches: 1.59.12; 1.59.16;
Delete do-nothing device-activation hooks.
|
| 1.58 | 12-May-2009 |
cegger | struct device * -> device_t, no functional changes intended.
|
| 1.57 | 12-May-2009 |
cegger | struct cfdata * -> cfdata_t, no functional changes intended.
|
| 1.56 | 11-Jan-2009 |
cegger | branches: 1.56.2;
make this compile
|
| 1.55 | 08-Jun-2008 |
tsutsui | branches: 1.55.6;
Use device_lookup_private() rather than using cd_devs[] directly to get softc.
XXX maybe we should change a type of cd_devs[] in struct cfdriver
from (void *) to device_t.
|
| 1.54 | 28-Apr-2008 |
martin | branches: 1.54.2;
Remove clause 3 and 4 from TNF licenses
|
| 1.53 | 05-Apr-2008 |
cegger | branches: 1.53.2; 1.53.4;
use aprint_*_dev and device_xname
|
| 1.52 | 04-Mar-2007 |
christos | branches: 1.52.36;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
|
| 1.51 | 16-Nov-2006 |
christos | branches: 1.51.4;
__unused removal on arguments; approved by core.
|
| 1.50 | 12-Oct-2006 |
christos | - sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386
|
| 1.49 | 03-Sep-2006 |
christos | branches: 1.49.2; 1.49.4;
add missing initializer
|
| 1.48 | 30-Mar-2006 |
thorpej | Use device_private().
|
| 1.47 | 28-Mar-2006 |
thorpej | Use device_unit().
|
| 1.46 | 11-Dec-2005 |
christos | branches: 1.46.4; 1.46.6; 1.46.8; 1.46.10; 1.46.12;
merge ktrace-lwp.
|
| 1.45 | 27-Feb-2005 |
perry | branches: 1.45.4;
nuke trailing whitespace
|
| 1.44 | 01-Feb-2005 |
reinoud | Backing out changes to clean up scsipi. I was pointed out there were
problems i hadn't seen. To prevent lossage i'd decided to back off all
changes and let them be reviewed on tech-kern.
|
| 1.43 | 31-Jan-2005 |
reinoud | Part of the cleanup of sys/scsipi's use of types; rename all u_int* to
uint* and change the u_long's to uint32_t's where possible. Note that the
iocl definitions/hooks have to be ulong (or u_long) or they'll bomb out.
|
| 1.42 | 21-Aug-2004 |
thorpej | branches: 1.42.4; 1.42.6;
Use ANSI function decls and make use of static.
|
| 1.41 | 08-Sep-2003 |
mycroft | We only need a single attachment for uk at scsibus and atapibus.
|
| 1.40 | 29-Jun-2003 |
fvdl | branches: 1.40.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.
|
| 1.39 | 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
|
| 1.38 | 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
|
| 1.37 | 02-Oct-2002 |
thorpej | Add trailing ; to CFATTACH_DECL.
|
| 1.36 | 30-Sep-2002 |
thorpej | Use CFATTACH_DECL().
|
| 1.35 | 27-Sep-2002 |
thorpej | Declare all cfattach structures const.
|
| 1.34 | 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.
|
| 1.33 | 10-Feb-2002 |
thorpej | branches: 1.33.8;
Remove a pretty much useless autoconfiguration message.
|
| 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 | 25-Apr-2001 |
bouyer | branches: 1.30.2; 1.30.4;
Pull up the thorpej_scsipi branch to main branch.
This is a completely rewritten scsipi_xfer execution engine, and the
associated changes to HBA drivers. Overview of changes & features:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
- support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command
will be requeued, or a REQUEST SENSE will be sent as appropriate.
Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
|
| 1.29 | 30-Mar-2000 |
augustss | branches: 1.29.6;
Get rid of register declarations.
|
| 1.28 | 29-Mar-2000 |
augustss | Add ukactivate&ukdetach so uk devices can be detached.
|
| 1.27 | 08-Dec-1998 |
thorpej | branches: 1.27.10;
When closing, wait for pending xfers to drain before deleting the reference
to the adapter.
|
| 1.26 | 20-Nov-1998 |
thorpej | Add adapter reference counting for SCSI and ATAPI devices.
|
| 1.25 | 19-Oct-1998 |
bouyer | Allow uk to attach at atapibus.
|
| 1.24 | 31-Aug-1998 |
cgd | kill the last remnants of __BROKEN_INDIRECT_CONFIG. (only the pica port
used it, and it's non-working and apparently slated for replacement.)
|
| 1.23 | 17-Aug-1998 |
mycroft | Assign my copyrights to TNF.
|
| 1.22 | 15-Aug-1998 |
mycroft | Make copyright notices with my name consistent.
|
| 1.21 | 12-Jan-1998 |
thorpej | Adjust for changes to config.
|
| 1.20 | 01-Oct-1997 |
enami | Cosmetic changes to keep coding style consistency in this directory;
- Indent with tab of width 8.
- Use four column to indent continuation line.
- Fold long line if possible.
- Use return (xx) instead of return xx.
- Compare pointer against NULL instead of testing like boolean.
- Delete whitespace at the end of line.
- Delete whitespace in front of function call operator.
- Delete whitespace after cast.
- Dereference a pointer to function explicitly.
- Add an empty line after local variable declaration.
- Use NULL instead of (char *)0.
- Dont use block for single statement.
|
| 1.19 | 27-Aug-1997 |
bouyer | branches: 1.19.2;
Merge scsipi branch in the mainline. This add support for ATAPI devices
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
|
| 1.18 | 05-Dec-1996 |
cgd | update these so they compile whether or not __BROKEN_INDIRECT_CONFIG
is defined.
|
| 1.17 | 12-Oct-1996 |
christos | revert previous kprintf change
|
| 1.16 | 10-Oct-1996 |
christos | - printf -> kprintf, sprintf -> ksprintf
- make this compile cleanly.
|
| 1.15 | 17-Mar-1996 |
thorpej | New device attachment scheme:
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
|
| 1.14 | 05-Mar-1996 |
thorpej | Normalize SCSI autoconfiguration output.
From Chris Demetriou <cgd@NetBSD.ORG>. Fixes PR #1958.
|
| 1.13 | 24-Mar-1995 |
glass | fix syntax error. Andreas Gustafsson <gson@clinet.fi>
|
| 1.12 | 28-Dec-1994 |
mycroft | Numerous changes. Many bugs fixed, better autoconfig, a few new features.
|
| 1.11 | 21-Nov-1994 |
mycroft | Clean up open and close routines somewhat.
|
| 1.10 | 21-Nov-1994 |
mycroft | Replace dev_unit with device_softc in scsi_link. Change argument to foostart()
to void*.
|
| 1.9 | 30-Oct-1994 |
cgd | be more careful with types, also pull in headers where necessary.
|
| 1.8 | 20-Oct-1994 |
mycroft | First cut at making user-level SCSI commands work. This is untested.
Partly from John Brezak.
|
| 1.7 | 29-Jun-1994 |
cgd | New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
|
| 1.6 | 11-Apr-1994 |
mycroft | Fix various types. Remove some outdated flags.
|
| 1.5 | 29-Mar-1994 |
mycroft | New SCSI system, based on Julian's more recent work.
|
| 1.4 | 16-Feb-1994 |
mycroft | More KNF-like.
|
| 1.3 | 01-Feb-1994 |
mycroft | Add scsi_stop_unit(), from Mac code. Other minor nits.
|
| 1.2 | 17-Dec-1993 |
mycroft | Canonicalize all #includes.
|
| 1.1 | 24-Nov-1993 |
mycroft | Under construction...
|
| 1.19.2.3 | 14-Oct-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.19.2.2 | 27-Aug-1997 |
thorpej | Update marc-pcmcia branch from trunk.
|
| 1.19.2.1 | 27-Aug-1997 |
thorpej | file uk.c was added on branch marc-pcmcia on 1997-08-27 23:33:43 +0000
|
| 1.27.10.3 | 20-Nov-2000 |
bouyer | Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.
Main changes to the scsipi code itself:
- add a scsipi_channel->type to allow umass to attach to both atapibus and
scsibus. Will die when IDE is converted from ata_atapi_attach to
scsipi_channel/scsipi_adapter
- Add a chan_defquirks to scsipi_channel so that adapters can pass a default
set of quirks to be set for each device attached
- add adapt_getgeom and adapt_accesschk callbacks
|
| 1.27.10.2 | 01-Nov-1999 |
thorpej | Fixup the SC_DEBUG() stuff for the new world order.
|
| 1.27.10.1 | 19-Oct-1999 |
thorpej | Completely rewritten scsipi_xfer execution engine:
- All xfers are queued in the mid-layer, rather than doing so in an
ad-hoc fashion in individual adapter drivers.
- Adapter/channel resource management in the mid-layer, avoids even trying
to start running an xfer if the adapter/channel doesn't have the resources.
- Better communication between the mid-layer and the adapters.
- Asynchronous event notification mechanism from adapter to mid-layer and
peripherals.
- Better peripheral queue management: freeze/thaw, sorted requeueing during
recovery, etc.
- Clean separation of peripherals, adapters, and adapter channels (no more
scsipi_link).
- Kernel thread for each scsipi_channel makes error recovery much easier
(no more dealing with interrupt context when recovering from an error).
- Mid-layer support for tagged queueing: commands can have the tag type
set explicitly, tag IDs are allocated in the mid-layer (thus eliminating
the need to use buggy tag ID allocation schemes in many adapter drivers).
There is a lot more work to do, but this correctly functions for the most
part on several file servers I run.
|
| 1.29.6.7 | 11-Nov-2002 |
nathanw | Catch up to -current
|
| 1.29.6.6 | 18-Oct-2002 |
nathanw | Catch up to -current.
|
| 1.29.6.5 | 17-Sep-2002 |
nathanw | Catch up to -current.
|
| 1.29.6.4 | 28-Feb-2002 |
nathanw | Catch up to -current.
|
| 1.29.6.3 | 08-Jan-2002 |
nathanw | Catch up to -current.
|
| 1.29.6.2 | 14-Nov-2001 |
nathanw | Catch up to -current.
|
| 1.29.6.1 | 21-Jun-2001 |
nathanw | Catch up to -current.
|
| 1.30.4.2 | 26-Sep-2001 |
fvdl | * add a VCLONED vnode flag that indicates a vnode representing a cloned
device.
* rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass
to VOP_REVOKE
* the revoke system call will revoke all aliases, as before, but not the
clones
* vdevgone is called when detaching a device, so make it use REVOKECLONE
to get rid of all clones as well
* clean up all uses of VOP_OPEN wrt. locking.
* add a few VOPS to spec_vnops that need to do something when it's a
clone vnode (access and getattr)
* add a copy of the vnode vattr structure of the original 'master' vnode
to the specinfo of a cloned vnode. could possibly redirect getattr to
the 'master' vnode, but this has issues with revoke
* add a vdev_reassignvp function that disassociates a vnode from its
original device, and reassociates it with the specified dev_t. to be
used by cloning devices only, in case a new minor is allocated.
* change all direct references in drivers to v_devcookie and v_rdev
to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes
when debugging race conditions that still exist wrt. locking and
revoking vnodes.
* make the locking state of a vnode consistent when passed to
d_open and d_close (unlocked). locked would be better, but has
some deadlock issues
|
| 1.30.4.1 | 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).
|
| 1.30.2.3 | 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
|
| 1.30.2.2 | 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.30.2.1 | 10-Jan-2002 |
thorpej | Sync kqueue branch with -current.
|
| 1.33.8.1 | 16-May-2002 |
gehenna | Add the character device switch.
Replace the direct-access to devsw table with calling devsw API.
|
| 1.40.2.6 | 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.40.2.5 | 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.40.2.4 | 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.40.2.3 | 25-Aug-2004 |
skrll | Sync with HEAD.
|
| 1.40.2.2 | 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.40.2.1 | 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.
|
| 1.42.6.1 | 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.42.4.1 | 29-Apr-2005 |
kent | sync with -current
|
| 1.45.4.3 | 03-Sep-2007 |
yamt | sync with head.
|
| 1.45.4.2 | 30-Dec-2006 |
yamt | sync with head.
|
| 1.45.4.1 | 21-Jun-2006 |
yamt | sync with head.
|
| 1.46.12.1 | 31-Mar-2006 |
tron | Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
|
| 1.46.10.1 | 19-Apr-2006 |
elad | sync with head.
|
| 1.46.8.2 | 14-Sep-2006 |
yamt | sync with head.
|
| 1.46.8.1 | 01-Apr-2006 |
yamt | sync with head.
|
| 1.46.6.1 | 22-Apr-2006 |
simonb | Sync with head.
|
| 1.46.4.1 | 09-Sep-2006 |
rpaulo | sync with head
|
| 1.49.4.2 | 10-Dec-2006 |
yamt | sync with head.
|
| 1.49.4.1 | 22-Oct-2006 |
yamt | sync with head
|
| 1.49.2.1 | 18-Nov-2006 |
ad | Sync with head.
|
| 1.51.4.1 | 12-Mar-2007 |
rmind | Sync with HEAD.
|
| 1.52.36.5 | 17-Jan-2009 |
mjf | Sync with HEAD.
|
| 1.52.36.4 | 29-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.52.36.3 | 02-Jun-2008 |
mjf | Sync with HEAD.
|
| 1.52.36.2 | 06-Apr-2008 |
mjf | - after some discussion with agc@ i agreed it would be a good idea to move
device_unregister_* to device_deregister_* to be more like the pmf(9)
functions, especially since a lot of the time the function calls are next
to each other.
- add device_register_name() support for dk(4).
|
| 1.52.36.1 | 05-Apr-2008 |
mjf | - add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing
as these are always needed.
- convert many, many drivers over to the New Devfs World Order. For a
list of device drivers yet to be converted see,
http://www.netbsd.org/~mjf/devfs-todo.html.
- add a new device_unregister_all(device_t) function to remove all device
names associated with a device_t, which saves us having to construct
device names when the driver is detached.
- add a DEV_AUDIO type for devices.
|
| 1.53.4.4 | 11-Mar-2010 |
yamt | sync with head
|
| 1.53.4.3 | 16-May-2009 |
yamt | sync with head
|
| 1.53.4.2 | 04-May-2009 |
yamt | sync with head.
|
| 1.53.4.1 | 16-May-2008 |
yamt | sync with head.
|
| 1.53.2.2 | 17-Jun-2008 |
yamt | sync with head.
|
| 1.53.2.1 | 18-May-2008 |
yamt | sync with head.
|
| 1.54.2.1 | 23-Jun-2008 |
wrstuden | Sync w/ -current. 34 merge conflicts to follow.
|
| 1.55.6.1 | 19-Jan-2009 |
skrll | Sync with HEAD.
|
| 1.56.2.1 | 13-May-2009 |
jym | Sync with HEAD.
Commit is split, to avoid a "too many arguments" protocol error.
|
| 1.59.16.1 | 18-Feb-2012 |
mrg | merge to -current.
|
| 1.59.12.2 | 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")
|
| 1.59.12.1 | 17-Apr-2012 |
yamt | sync with head
|
| 1.60.10.1 | 18-May-2014 |
rmind | sync with head
|
| 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.
|
| 1.61.2.1 | 10-Aug-2014 |
tls | Rebase.
|
| 1.62.4.2 | 05-Dec-2016 |
skrll | Sync with HEAD
|
| 1.62.4.1 | 05-Oct-2016 |
skrll | Sync with HEAD
|
| 1.63.2.2 | 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.63.2.1 | 06-Aug-2016 |
pgoyette | Sync with HEAD
|
