ieee1394 import from FreeBSD.

branches: 1.7.4;
nuke trailing whitespace

branches: 1.6.16; 1.6.24; 1.6.26;
Change some wording on fw's to make it more logical sounding.

heirarchy -> hierarchy

branches: 1.4.2;
Add locators support

Point out where the documentation is wrong.

Rework implementation notes/ideas with a more fleshed out example.

branches: 1.1.4; 1.1.6; 1.1.8;
random thoughts on how i'm implementing this.

merge ktrace-lwp.

ieee1394 import from FreeBSD.

branches: 1.3.6; 1.3.22;
New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

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.

branches: 1.1.4; 1.1.6; 1.1.8;
Install ieee1394reg.h since user application (someday) will need it.

ieee1394 import from FreeBSD.

branches: 1.2.4;
nuke trailing whitespace

branches: 1.1.2; 1.1.8; 1.1.16; 1.1.18;
First pass (checkpoint checkin) for firewire scsi code.

There are a number of issues here for anyone trying to use this today:

1. On my test drive the command engine on the drive seems to stall after the
inquire is done. So the mode sense times out for a long time before
aborting. This obviously needs to be tracked down and fixed.

However it does do a proper inquire:

scsibus0 at sbpscsi0: 1 target, 1 lun per target
sd0 at scsibus0 target 0 lun 0: <Maxtor, 1394 storage, 60> disk fixed

2. This code is quite ugly in places as debug code was added to test things.
Definitly needs cleanup/documention in places where it's using command
structures. The structure for alloc'ing orbs, running them through the
command engine and getting state back is mostly set but implementation needs
an overhaul in places.

3. For testing I use the following config options:

fwohci* at cardbus? dev ? function ? # IEEE1394 Controller
fw* at fwbus?
options FW_DEBUG
options FWNODE_DEBUG
options P1212_DEBUG
options SBP2_DEBUG
options SBPSCSI_DEBUG
fwnode* at fwbus?
sbpscsi* at fwnode?
scsibus* at sbpscsi?

ieee1394 import from FreeBSD.

branches: 1.11.4;
nuke trailing whitespace

branches: 1.10.2; 1.10.10; 1.10.12;
Updates

First pass (checkpoint checkin) for firewire scsi code.

There are a number of issues here for anyone trying to use this today:

1. On my test drive the command engine on the drive seems to stall after the
inquire is done. So the mode sense times out for a long time before
aborting. This obviously needs to be tracked down and fixed.

However it does do a proper inquire:

scsibus0 at sbpscsi0: 1 target, 1 lun per target
sd0 at scsibus0 target 0 lun 0: <Maxtor, 1394 storage, 60> disk fixed

2. This code is quite ugly in places as debug code was added to test things.
Definitly needs cleanup/documention in places where it's using command
structures. The structure for alloc'ing orbs, running them through the
command engine and getting state back is mostly set but implementation needs
an overhaul in places.

3. For testing I use the following config options:

fwohci* at cardbus? dev ? function ? # IEEE1394 Controller
fw* at fwbus?
options FW_DEBUG
options FWNODE_DEBUG
options P1212_DEBUG
options SBP2_DEBUG
options SBPSCSI_DEBUG
fwnode* at fwbus?
sbpscsi* at fwnode?
scsibus* at sbpscsi?

Note done sections related to fwnode/ieee1212 stuff

done - Make fwohci detach correctly (for cardbus/etc type interfaces)

branches: 1.6.2; 1.6.4;
Add a few nits. Hubs seem problematic on resets sometimes.
Also, make sure fwohci can detach cleanly (for cardbus).

Clean up comment about unregistering callbacks

Add locators support

Add some notes

Add a few more items

First pass at the overall TODO list. Some of these may come off before I get the full code checked in even

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?).

branches: 1.10.168;
Indented.

branches: 1.9.4; 1.9.6; 1.9.8; 1.9.10; 1.9.12;
merge ktrace-lwp.

ieee1394 import from FreeBSD.

branches: 1.7.6; 1.7.22;
First pass (checkpoint checkin) for firewire scsi code.

There are a number of issues here for anyone trying to use this today:

1. On my test drive the command engine on the drive seems to stall after the
inquire is done. So the mode sense times out for a long time before
aborting. This obviously needs to be tracked down and fixed.

However it does do a proper inquire:

scsibus0 at sbpscsi0: 1 target, 1 lun per target
sd0 at scsibus0 target 0 lun 0: <Maxtor, 1394 storage, 60> disk fixed

2. This code is quite ugly in places as debug code was added to test things.
Definitly needs cleanup/documention in places where it's using command
structures. The structure for alloc'ing orbs, running them through the
command engine and getting state back is mostly set but implementation needs
an overhaul in places.

3. For testing I use the following config options:

fwohci* at cardbus? dev ? function ? # IEEE1394 Controller
fw* at fwbus?
options FW_DEBUG
options FWNODE_DEBUG
options P1212_DEBUG
options SBP2_DEBUG
options SBPSCSI_DEBUG
fwnode* at fwbus?
sbpscsi* at fwnode?
scsibus* at sbpscsi?

Rototill fwnode.c to rely on the p1212 routines to parse and validate the ROM

branches: 1.5.2;
Add locators support

Put in the structure for fwnode/fwscsi

branches: 1.3.2;
whitespace police

branches: 1.2.2;
Add a fwlynx device. Move the file for fwohci to files.ieee1394

First Prototype implementation of network interface part for IEEE1394 (if_fw).

Current status:
Only OHCI chip is supported (fwohci).
ping (IPv4) works with Sony's implementation (SmartConnect) on Win98.
sometimes works but not stable.
Not implemented yet:
IRM (Isochronous Resource Manager) functionality.
Link layer fragmentation.
Topology map.
More to do:
clean ups
MCAP
charactor device part
dhcp

There is no entry in GENERIC config file yet.
Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.

fix various small typos, mainly in comments.

ieee1394: Need kernel lock around config_found.

follow fcambus steps by fixing typos for "unknown" in few more error messages.

Merge thorpej-cfargs2.

branches: 1.51.8;
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?).

branches: 1.50.10;
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.

Use unsigned to avoid undefined behavior. Found by kUBSan.

branches: 1.48.4;
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!)

branches: 1.47.14; 1.47.16;
Kill some more tsleep.

And no, surrounding tsleep by mutex_exit/mutex_enter does not fix
problems even if the LOCKDEBUG panic symptom goes away...

Defer initialization of isodma channels until we know how many.

Should fix a bug I introduced four years ago in:
https://mail-index.netbsd.org/source-changes/2012/08/04/msg036211.html

branches: 1.45.2; 1.45.4;
src is too big these days to tolerate superfluous apostrophes. It's
"its", people!

Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.

Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.

Pass the device name in, so we can debug what deferred drivers did not work.

branches: 1.42.2; 1.42.4;
...and be sure to config_pending_decr if kthread_create fails.

This error branch looks suspect, though. Shouldn't we bail at this
point rather than blithely try to proceed?

Restore config pending stuff, adjusted to avoid the race.

Leave a note about what this code probably should look like for
anyone who comes rummaging around with the intent to really fix it.
I would do that myself if I had time and resources to test this at
the moment, and more comfort with our firewire stack, but I don't.

Fix error branches and config pending races in firewire init.

This way, if anything fails, it just fails; you don't panic. This can
happen if suspending and resuming of firewire is broken (e.g., as I
encountered in PR kern/44581).

Change to consistently use M_FW for all malloc/free.
It probably doesn't matter any more, but the code doesn't appear to
have matched its mallocs and frees - so the stats would have been awol.

branches: 1.38.8; 1.38.12;
do not assume all fw devices speak sbp.
teach the fw attach code to deal with different fw device classes.
this allows other fw drivers than sbp to attach

fix crash when detaching/re-attaching a cable where three firewire devices are on the bus.

Fix some code paths where pointers are dereferenced after checking that
they are NULL (oops?)

XXX pull-ups for NetBSD-4 and NetBSD-5.

Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

don't allocate rom scratch memory from an interrupt context:

fwohci0: BUS reset
panic: kernel diagnostic assertion "!cpu_intr_p()" failed: file "../../../../ker
n/subr_kmem.c", line 195
fatal breakpoint trap in supervisor mode
trap type 1 code 0 rip ffffffff8022db1d cs 8 rflags 246 cr2 0 cpl 6 rsp fffffff
f80fafb08
breakpoint() at netbsd:breakpoint+0x5
panic() at netbsd:panic+0x2ba
kern_assert() at netbsd:kern_assert+0x2d
kmem_alloc() at netbsd:kmem_alloc+0x18a
kmem_zalloc() at netbsd:kmem_zalloc+0xf
fw_busreset() at netbsd:fw_busreset+0x23b
fwohci_intr() at netbsd:fwohci_intr+0xa56
...

Add mutex.h.

May be easy to understand 'pay_len > 0' more than 'paylaod != NULL'.

Use kmem(9) instead of malloc(9).

KASSERT(FALSE) to enter the debugger on this check. When DDB is not defined
but KGDB is, this would result in cpu_Debugger() dont being defined! Please be
kind to developers that use KGDB instead of DDB! :)

Remove more my debugging code.

Remove debugging process.

Bye-bye fw_port.h.

branches: 1.26.2;
various aprint_* fixes.

branches: 1.25.2;
struct cfdata * -> cfdata_t, no functional changes intended.

bzero -> memset

bcmp -> memcmp

branches: 1.22.4;
Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

branches: 1.21.4; 1.21.6;
Use aprint

branches: 1.20.2; 1.20.4; 1.20.6;
Split device_t/softc.
And cosmetic change.

branches: 1.19.6;
__FUNCTION__ -> __func__

use __KERNEL_RCSID()

branches: 1.17.2;
Merge jmcneill-pm branch.

branches: 1.16.4; 1.16.6;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.14.6; 1.14.8; 1.14.12;
Merge some of the less invasive changes from the vmlocking branch:

- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements

firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.12.2; 1.12.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.11.4;
__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

Fix device name in {fwohci,firewire}_print(). valiable aux is not string
pointer.

branches: 1.8.4; 1.8.8; 1.8.10;
Don't allocate > 1K on the stack

branches: 1.7.2;
Coverity CID 1118: Protect NULL deref.

fw_bus_probe_thread: change wmesg from "-" to "fwprobe".

branches: 1.5.4; 1.5.6; 1.5.8; 1.5.10; 1.5.12;
merge ktrace-lwp.

branches: 1.4.6;
kill some more simple submatch() functions, use config_stdsubmatch()

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

minor cleanup: some "const", global namespace sanity and fix a maclo name

ieee1394 import from FreeBSD.

convert union member from [0] array to [3] array, so that attempts
to access these members later will work.

does not affect size of the structure or layout in any other way.

branches: 1.6.76;
Include dependencies directly.

Bye-bye fw_port.h.

branches: 1.4.20; 1.4.40; 1.4.42;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

branches: 1.3.22; 1.3.24; 1.3.28;
make the struct fw_crom_buf's len uint32_t, so that it can never be negative
the real fix to
http://archives.neohapsis.com/archives/fulldisclosure/2006-11/0261.html

branches: 1.2.18; 1.2.22; 1.2.24;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

Bye-bye fw_port.h.

branches: 1.2.18; 1.2.76; 1.2.98; 1.2.100;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

Kill some more tsleep.

And no, surrounding tsleep by mutex_exit/mutex_enter does not fix
problems even if the LOCKDEBUG panic symptom goes away...

Defer initialization of isodma channels until we know how many.

Should fix a bug I introduced four years ago in:
https://mail-index.netbsd.org/source-changes/2012/08/04/msg036211.html

branches: 1.18.2; 1.18.16; 1.18.20;
Fix error branches and config pending races in firewire init.

This way, if anything fails, it just fails; you don't panic. This can
happen if suspending and resuming of firewire is broken (e.g., as I
encountered in PR kern/44581).

Change to consistently use M_FW for all malloc/free.
It probably doesn't matter any more, but the code doesn't appear to
have matched its mallocs and frees - so the stats would have been awol.

branches: 1.16.4; 1.16.8;
Inclusion for malloc decls (missed in previous commit).

branches: 1.15.2;
Include dependencies directly.

do not assume all fw devices speak sbp.
teach the fw attach code to deal with different fw device classes.
this allows other fw drivers than sbp to attach

Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

don't allocate rom scratch memory from an interrupt context:

fwohci0: BUS reset
panic: kernel diagnostic assertion "!cpu_intr_p()" failed: file "../../../../ker
n/subr_kmem.c", line 195
fatal breakpoint trap in supervisor mode
trap type 1 code 0 rip ffffffff8022db1d cs 8 rflags 246 cr2 0 cpl 6 rsp fffffff
f80fafb08
breakpoint() at netbsd:breakpoint+0x5
panic() at netbsd:panic+0x2ba
kern_assert() at netbsd:kern_assert+0x2d
kmem_alloc() at netbsd:kmem_alloc+0x18a
kmem_zalloc() at netbsd:kmem_zalloc+0xf
fw_busreset() at netbsd:fw_busreset+0x23b
fwohci_intr() at netbsd:fwohci_intr+0xa56
...

Bye-bye fw_port.h.

branches: 1.10.2; 1.10.4;
unifdef -U__DragonFlyBSD__ -U__FreeBSD__ -D__NetBSD__.

XXX Compiles, but untested.

buildfix: don't redefine MIN and MAX. They are defined in <sys/param.h>.

branches: 1.8.4; 1.8.18;
Split device_t/softc.
And cosmetic change.

branches: 1.7.16;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

branches: 1.6.4; 1.6.6; 1.6.10; 1.6.12;
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.

branches: 1.5.2;
firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.4.2; 1.4.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.3.18; 1.3.28;
merge ktrace-lwp.

branches: 1.2.6;
minor cleanup: some "const", global namespace sanity and fix a maclo name

ieee1394 import from FreeBSD.

Bye-bye fw_port.h.

branches: 1.33.2; 1.33.4;
unifdef -U__FreeBSD__ -D__NetBSD__, et cetera.

Add queue macros SLIST_FOREACH_SAFE, STAILQ_FOREACH_SAFE, STAILQ_LAST,
SIMPLEQ_CONCAT, and SIMPLEQ_LAST.

Remove seperate definitions outside of queue.h.

bzero -> memset

branches: 1.30.4; 1.30.10;
Digraphs as lvalues must return the same type.

branches: 1.29.4; 1.29.6; 1.29.8;
use aprint_*_dev and device_xname

Split device_t/softc.
And cosmetic change.

branches: 1.27.6;
Convert many of the uses of __attribute__ to equivalent
__packed, __unused and __dead macros from cdefs.h

branches: 1.26.2; 1.26.4; 1.26.8;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.24.2; 1.24.6; 1.24.8;
Improve per-CPU support for the workqueue(9):
- Make structures CPU-cache friendly, as suggested and explained
by Andrew Doran. CACHE_LINE_SIZE definition is invented.
- Use current CPU if NULL is passed to the workqueue_enqueue().
- Implemented MI CPU index, which could be used as an index of array.
Removed linked-lists usage for work queues.

The roundup2() function avoids division, but works only with power of 2.

Reviewed by: <ad>, <yamt>, <tech-kern>

branches: 1.23.2; 1.23.6;
Merge some of the less invasive changes from the vmlocking branch:

- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements

firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.21.2; 1.21.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.20.4;
__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

we have rounddown in <sys/param.h> now

Add "name" parameter to powerhook_establish, to aid debugging. No objections
on tech-kern@

branches: 1.16.2; 1.16.4;
fix order of decl

branches: 1.15.6;
Use An appropriate flag in macro fw_bus_dma_*.

Coordination of macro fw_bus_dma_tag_create().

Remove `;'.

No parentheses are needed around the return value.

Coverity CID 1120: Avoid NULL deref.

Coverity CID 2696: Don't use SLIST_FOREACH when freeing elements.

Use device_unit().

branches: 1.8.2; 1.8.4; 1.8.6;
Change "inline" back to "__inline" in .h files -- C99 is still too
new, and some apps compile things in C89 mode. C89 keywords stay.

As per core@.

branches: 1.7.2; 1.7.4; 1.7.6;
No longer need our own device_t typedef.

merge ktrace-lwp.

Merge the 3 copies of m_getcl() so that fast ipsec compiles again together
with net80211. XXX: We don't really have an m_getcl(), we just emulate it.

branches: 1.4.6;
kill some more simple submatch() functions, use config_stdsubmatch()

Since fwip_unicast_input_recycle() is called from interrupt context, one
can't sleep there.

minor cleanup: some "const", global namespace sanity and fix a maclo name

ieee1394 import from FreeBSD.

No reason not to make the string const.

ran sed too agressively.

fix cut-n-paste, use arrays instead of separate variables to avoid code
duplication.

branches: 1.14.4;
Deal with snprintf more gracefully.

branches: 1.13.8; 1.13.18; 1.13.22;
- print expected crc when it does not match
- add and print secondary text leaf. This prints the model name of my webcam.
- add csr keys that are printed as 'unknown' otherwise

ANSIfy, KNF

Bye-bye fw_port.h.

branches: 1.10.2; 1.10.4;
Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch

bcopy -> memcpy

bzero -> memset

branches: 1.7.8; 1.7.14;
Constify, so that usr.sbin/fwctl can be built with WARNS=4.

branches: 1.6.8; 1.6.10; 1.6.12;
use __KERNEL_RCSID()

branches: 1.5.8; 1.5.16; 1.5.18; 1.5.20;
firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.4.2; 1.4.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.3.6; 1.3.16;
Fix a bunch of cast lvalues.

branches: 1.2.4; 1.2.6; 1.2.8; 1.2.10; 1.2.12;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

s/avaliable/available/

memory leak, found by Mootja; it seems that we should check the return
value of 'fw_bindadd' in several other places, but whatever

Kill some more tsleep.

And no, surrounding tsleep by mutex_exit/mutex_enter does not fix
problems even if the LOCKDEBUG panic symptom goes away...

branches: 1.30.4; 1.30.8;
Add d_discard to all struct cdevsw instances I could find.

All have been set to "nodiscard"; some should get a real implementation.

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.

branches: 1.28.2;
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.

branches: 1.27.2; 1.27.4;
Change to consistently use M_FW for all malloc/free.
It probably doesn't matter any more, but the code doesn't appear to
have matched its mallocs and frees - so the stats would have been awol.

branches: 1.26.2; 1.26.6;
Put back #include "ioconf.h" per popular demand.

Declare cfdrivers using extern rather than including ioconf.h.

branches: 1.24.2;
don't call tsleep() with any mutexes hold

do not hold mutexes when calling tsleep(9).
Fixes LOCKDEBUG panics

Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

Use kmem(9) instead of malloc(9).

Use kmem(9) instead of malloc(9).

Remove unnecessary cast.

Bye-bye fw_port.h.

branches: 1.17.2; 1.17.4;
Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch

bcopy -> memcpy

bzero -> memset

branches: 1.14.12; 1.14.20; 1.14.26;
use __KERNEL_RCSID()

branches: 1.13.4; 1.13.6; 1.13.8;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.11.4; 1.11.6; 1.11.10; 1.11.12;
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.

branches: 1.10.2;
firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.9.2; 1.9.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.8.4;
undo previous change - crom_buf->len has been changed to uint32_t
(this is backward ABI compatible) so the check for negative value is
no longer necessary

Avoid integer overflow, as reported by Rodrigo Rubira Branco in a
message to full-disclosure:

http://archives.neohapsis.com/archives/fulldisclosure/2006-11/0261.html

Also see comment from bouyer@:

http://mail-index.netbsd.org/tech-security/2006/11/16/0001.html

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.4.2; 1.4.4;
- fix missing initializations
- remove stray semicolon, that screwed up error checking

branches: 1.3.4; 1.3.8; 1.3.18;
merge ktrace-lwp.

branches: 1.2.6;
minor cleanup: some "const", global namespace sanity and fix a maclo name

ieee1394 import from FreeBSD.

Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.

branches: 1.17.16;
Fix fwdma_free() for when bus_dmamap_unload() clobbers the dmamap.

Prevents KASSERTs on detaches of sbp(4) and fwohci(4) on amd64.

We should pass the dmamem segs around seperately, not in loaded dmamap...

branches: 1.16.58; 1.16.60;
Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

Use kmem(9) instead of malloc(9).

Call malloc(M_ZERO) instead of malloc() and memset().

Bye-bye fw_port.h.

branches: 1.12.10; 1.12.30; 1.12.32;
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.

use __KERNEL_RCSID()

branches: 1.10.2; 1.10.4; 1.10.6; 1.10.8;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.8.12; 1.8.14; 1.8.18;
Remove #include <machine/vmparam.h>.

branches: 1.7.4; 1.7.8;
__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.5.6; 1.5.10; 1.5.12;
Fix memory leak.

Use An appropriate flag in macro fw_bus_dma_*.

Coordination of macro fw_bus_dma_tag_create().

branches: 1.2.4; 1.2.6; 1.2.8; 1.2.10; 1.2.12;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

Bye-bye fw_port.h.

branches: 1.5.40; 1.5.62; 1.5.64;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.4.12; 1.4.22;
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@.

branches: 1.3.2; 1.3.4; 1.3.6;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

ieee1394 import from FreeBSD.

branches: 1.5.4;
nuke trailing whitespace

branches: 1.4.4; 1.4.6;
return an error on invalid mode.

add missing __KERNEL_RCSID()

branches: 1.2.2;
Add some missing splx()'s noted by Krister Walfridsson

branches: 1.1.2;
New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

ieee1394 import from FreeBSD.

branches: 1.1.2; 1.1.8; 1.1.24;
New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

ieee1394 import from FreeBSD.

branches: 1.1.2; 1.1.8; 1.1.24;
New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

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!)

branches: 1.18.28; 1.18.30;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.

Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.

branches: 1.17.2; 1.17.4;
Change to consistently use M_FW for all malloc/free.
It probably doesn't matter any more, but the code doesn't appear to
have matched its mallocs and frees - so the stats would have been awol.

branches: 1.16.2; 1.16.6;
Put back #include "ioconf.h" per popular demand.

Declare cfdrivers using extern rather than including ioconf.h.

branches: 1.14.2;
Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

Use kmem(9) instead of malloc(9).

Bye-bye fw_port.h.

branches: 1.11.2; 1.11.4;
Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch

bcopy -> memcpy

branches: 1.9.4;
Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

branches: 1.8.8; 1.8.12; 1.8.18; 1.8.20;
use __KERNEL_RCSID()

branches: 1.7.4; 1.7.6; 1.7.8;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.5.4; 1.5.6; 1.5.10; 1.5.12;
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.

branches: 1.4.2;
firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.3.2; 1.3.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.2.18; 1.2.28;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

Bye-bye fw_port.h.

branches: 1.3.40; 1.3.62; 1.3.64;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.2.18; 1.2.28;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

ieee1394 import from FreeBSD.

branches: 1.20.4;
nuke trailing whitespace

branches: 1.19.2; 1.19.10; 1.19.12;
Use aprint_normal() in cfprint routines.

Change structure/usage of interface provided callbacks. (read, write, reset)
Push these up to the ieee1394 softc level and make the naming scheme
consistant.

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

remove trailing \n in panic(). approved perry.

Only dump the rom up to ab_retlen in error cases. Also add in
P1212_ALLOW_VENDOR_DIRECTORY_TYPE to the flags passed to the p1212 parser

Update copyright date

Make this file compiles w/o FWNODE_DEBUG.

Rototill fwnode.c to rely on the p1212 routines to parse and validate the ROM

s/ab_csr/ab_addr/ so it compiles again.

Call malloc(9) with M_ZERO flag instead of memset() after malloc().

don't need <sys/types.h> when including <sys/param.h>

add RCSID

branches: 1.5.2; 1.5.4;
Make sure and zero out the malloc'd abufs before using them.

Change ab_node's to ab_req.
Use read/write instead of input/output.
Fix a bug in attach where the abuf wasn't getting zero'd out.

Retabify to correct style. Also, change to using DPRINTF interface for debug output.

Wrap some variable definitions with #ifdef FW_DEBUG that only apply in debug mode.

Initial checkin for fwnode code. Needs extensive cleanup

ieee1394 import from FreeBSD.

branches: 1.1.2; 1.1.24; 1.1.40;
Initial checkin for fwnode support

ieee1394 import from FreeBSD.

branches: 1.6.4;
nuke trailing whitespace

branches: 1.5.6; 1.5.14; 1.5.16;
Change structure/usage of interface provided callbacks. (read, write, reset)
Push these up to the ieee1394 softc level and make the naming scheme
consistant.

Rototill fwnode.c to rely on the p1212 routines to parse and validate the ROM

branches: 1.3.2; 1.3.4;
Rename input/output to read/write to match ieee1394var.h

Style police. Got the tab'ing correct

Initial checkin for fwnode support

don't assign struct pointers to smaller then structure regions of memory.

in all cases here, the later parts of the structure are not actually
accessed, so there are no existing bugs here beyond general UB. for the
ufs ones, this also removes some casts.

found by GCC 12.

s/bounaries/boundaries/ in comments.

fix various typos in comments.

s/transcat/transact/ in comment.

fix some more typos in comments/log messages, improve wording as well.

s/helt/held+s/eroneously/erroneously/+s/splitted/split/+s/recommented/recommended/

Merge thorpej-cfargs2.

branches: 1.145.8;
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?).

branches: 1.144.10;
Fix operand priority of bit op vs ternary operator.

Use unsigned to avoid undefined behavior. Found by kUBSan.

branches: 1.142.2;
s/recieve/receive/

branches: 1.141.16;
Let's try initializing the pointers *before* dereferencing them!

Oops.

KASSERT(mutex_owner(...)) ---> KASSERT(mutex_owned(...))

If this is not correct, then there's something bogus in this code
anyway, so better to fail early.

Final part of PR kern/47114.

Kill some more tsleep.

And no, surrounding tsleep by mutex_exit/mutex_enter does not fix
problems even if the LOCKDEBUG panic symptom goes away...

Defer initialization of isodma channels until we know how many.

Should fix a bug I introduced four years ago in:
https://mail-index.netbsd.org/source-changes/2012/08/04/msg036211.html

branches: 1.137.6; 1.137.10;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.

Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.

add missing ifdefs

Make it compilable with gcc 4.8.1

branches: 1.134.4;
Add '\n' to tail of printing string.

branches: 1.133.2;
Fix error branches and config pending races in firewire init.

This way, if anything fails, it just fails; you don't panic. This can
happen if suspending and resuming of firewire is broken (e.g., as I
encountered in PR kern/44581).

branches: 1.132.2;
Put back #include "ioconf.h" per popular demand.

Declare cfdrivers using extern rather than including ioconf.h.

branches: 1.130.2;
convert tsleep to kpause

fix grammar in error message

Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

Don't allocate sid scratch memory from an interrupt context:

fwohci0: BUS reset
fwohci0: node_id=0xc800ffc0, gen=1, CYCLEMASTER mode
panic: kernel diagnostic assertion "!cpu_intr_p()" failed: file "../../../../ker
n/subr_kmem.c", line 195
fatal breakpoint trap in supervisor mode
trap type 1 code 0 rip ffffffff8022db1d cs 8 rflags 246 cr2 0 cpl 6 rsp fffffff
f80fafb68
breakpoint() at netbsd:breakpoint+0x5
panic() at netbsd:panic+0x2ba
kern_assert() at netbsd:kern_assert+0x2d
kmem_alloc() at netbsd:kmem_alloc+0x18a
fwohci_intr() at netbsd:fwohci_intr+0xbe2
...

I will send-pr for the next one... Looks like someone did not use DIAGNOSTIC
when made the changes.

Use kmem(9) instead of malloc(9).

+ Fix typo.
+ Call aprint_error_dev() instead of aprint_error().

Fix a bug.
STAILQ is not NULL termination.

Remove more my debugging code.

Remove debugging process.

Bye-bye fw_port.h.

branches: 1.120.2; 1.120.4;
Remove extra whitespace added by a stupid tool.
XXX: more in src/sys/arch

bcopy -> memcpy

bzero -> memset

buildfix: undo previous

Ansify function definitions w/o arguments. Generated with sed.

branches: 1.115.4;
Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

branches: 1.114.4; 1.114.6;
Use aprint

branches: 1.113.10; 1.113.12; 1.113.14; 1.113.16;
Merge jmcneill-pm branch.

branches: 1.112.2; 1.112.4;
The argument of *_intr_establish() is changed to fwohci_flit.
Our ieee1394 coordinates from FreeBSD implementation.
Add fwohci_pci_shutdown() and fwohci_pci_power().
Move __KERNEL_RCSID to ahead.

on NetBSD, cardbus_intr_establish's fifth arg's is a symbol which returns
an int, not a void; change fwohci_intr accordingly.

branches: 1.110.2;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.108.2;
Use selinit() / seldestroy().

branches: 1.107.6; 1.107.8; 1.107.10;
firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.106.2; 1.106.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.105.2;
Include ioconf.h it for `struct cfdriver fwohci_cd'.

Fix memory allocation, in case isochronous transfer fail this.

branches: 1.103.2;
__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

Fix device name in {fwohci,firewire}_print(). valiable aux is not string
pointer.

branches: 1.100.6; 1.100.8;
Clean up bogus whitespace

branches: 1.99.2;
Back out for 1.94 from 1.93.

Fix memory leak.

Use An appropriate flag in macro fw_bus_dma_*.

use fw_bus_dmamap_create().

Indented.

Remove sync-ope for descritor of DMA, because its descriptor set
BUS_DMA_COHERENT flag.

Coordination of macro fw_bus_dma_tag_create().

Coverity CID 727: Comment out dead code.

branches: 1.91.4; 1.91.6; 1.91.8; 1.91.10; 1.91.12;
merge ktrace-lwp.

minor cleanup: some "const", global namespace sanity and fix a maclo name

ieee1394 import from FreeBSD.

branches: 1.88.2;
Change the name of a shadowed variable in FW_DEBUG code to make this
compile again.

fix handling of volatile and remove bogus casts.

No 0x in front of %p...

move an assignment up.

branches: 1.84.2; 1.84.4; 1.84.6;
nuke trailing whitespace

branches: 1.83.4; 1.83.6;
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"

netbsd.org -> NetBSD.org

This was the last commit of this kind to src/sys, which is now totally
"NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.

Fix unitialized var warning, add a comment to indicate that it's use
is suspect.

Fix unitialized variable warnings.

Typos in comments. From OpenBSD.

branches: 1.78.2;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

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

Bugfix: fwohci_ir_read(): loop continues when the buffer size not only
greater than but also same as packet size. This bug is pointed out by
FUKAUMI Naoki.

DMA, not dma nor Dma.

PR/20982: Eric Fair: fix debugging printf format errors.

Use aprint_*().

synchronous, not syncronous.

Add some missing splx()'s noted by Krister Walfridsson

Remove arbitrary bufcnt limit in at_output.

Use aprint_normal() in cfprint routines.

Add additional debugging for fwohci_arrq_input: no listener conditions

Improve performance with disks almost 60x by redoing the callback handler to
accept ranges as well as single addresses. Still need to go through any key
areas and remove the malloc's and replace these with some sort of pooling
instead.

Restructure how the at_output dma map gets setup. Ask for a dma map that can
support up to the max ohci descriptor segments. Then attempt to fill it in
via a load. Use the number of segments filled in as the basis for figuring
out how many descriptors to supply to the ohci DMA engine.

XXX: Still needs to account for requests which because of splits may overflow
the 6 dma descriptors available today. For now this fixes cases where a
single 512 byte write was getting split into 2 dma segments from 1 incoming
iov request

Fix fwohci_write's sanity check against large packets. Max rec is only used
when writing directly to the device (ala IEEE1394_TCODE_WRITE_REQ_BLOCK).
Otherwise the largest size is determined strictly by speed

Fill in sc1394_node_id after a self id cycle

Cast appropriately so that this compiles on sparc, sparc64, macppc etc.

New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

Correct calculating of offsets when memcpy returned data.

Print out some more details from incoming packets (tlabel's).

Also, when receiving a write packet, respond with a proper ack if something
was available to process it. (Fixes most of the sbp2 random hangs)

Add ULL to 64-bit integer constants.

Change structure/usage of interface provided callbacks. (read, write, reset)
Push these up to the ieee1394 softc level and make the naming scheme
consistant.

remove trailing \n in panic(). approved perry.

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.

remove obsolete use of M_LINK0. it was planned to be used to identify
isosynchronous stream from others, but end up not being used.
From: Hayakawa Koichi <haya@ilink.sony.co.jp>

branches: 1.54.2;
Clean up some comments

Change/pretty up some debugging info.

Also, add unit location fields to the ROM image. The p1212 spec requires
unit locations if you have 2 or more unit directories which is the case for
ipv4 and ipv6 compiled into a kernel.

Print the tcode in hex when debugging like all other debug statements

Add an unreg function.

Add a uio struct to the abuf so direct uio supporting calls can be setup.

Change all refs of csr in the abuf to addr as thats it's real function.

Fix illegal pointer manipulations that cause alignment problems.

Call malloc(9) with M_ZERO flag instead of memset() after malloc().

cosmetics: print a "\n" terminated string if no "fw at fwohci" is
configured

implement fwohci detach routine.

don't need <sys/types.h> when including <sys/param.h>

add RCSID

Bugfix: re-order double buffer only when second buffer exists.

Fix typo in printf format.

Bugfix: re-order buffers when they are released.

branches: 1.41.2;
make it compilable:
fw_verbose, fw_dump -> fwdebug (DPRINTFN)
add sc1394_ifsetiso into ieee1394_softc

New Feature: Add isochronous stream input routine. This feature has
been tested on i386. It does not have any interface for useland to
get isochoronous stream. The isochoronous acquisition interface
should be determined.

branches: 1.39.2;
performance improvement in fwohci_pkt_get() for asynchronous read.

set/reset low level receive handler according to IFF_UP flag of network
interface.

FW_DEBUG only: move descriptor dump from DPRINTFN(1) to DPRINTFN(2).
fix indent in packet dump.

Fixes to make if_fw work:
- fwohci_arrq_input: Do not return if next packet is in the buffer,
or the next packet cannot be received until the next receive interrupt.
- fwohci_uid_collect: XXX change M_WAITOK to M_NOWAIT for now, since there is
no lock but only one instance must be allowed here.
split fwohci_uid_req() and retry to get uid.
XXX need timer to wait some moment between retries...
- Add DELAY(10) for wait loop.
- split fwohci_buf_stop() into _stop_tx() and _stop_rx() because receiver
should not be stopped at bus-reset.
- split fwohci_buf_input() into _input() and _input_ppb() to improve
performance slightly.

simplify and localize to access sc_intmask in fwohci_event_thread().

Simplify to update sc_intmask in fwohci_intr().

FW_DEBUG: split debug printf in fwohci_intr() and fwohci_phy_input() into
fwohci_show_intr() and fwohci_show_phypkt() respectively.

fix printf to DPRINTFN inside FW_DEBUG

Various small changes.
Clean up one bug in a DPRINTF in arrs_input which could panic on some packets.
Gut the ack/response functionality and clean it up so all packets get checked
correctly and the abuf struct isn't used once the ab_cb has happened (there
still could be ack packets waiting to be processed at that time).
Finally, add some documentation explaining read/write/inreg and their
purpose/argument calling.

Add locators support

Add first pass generic ack callback code (to eventually replace mbuf specific code).
Change ab_node's to ab_req.
Fix a bug where setting BIBValid on OHCI 1.0 chips (this is a reserved bit
on 1.0)

Change to DPRINTF/DPRINTFN for debug output.

Avoid null pointer dereference.

Fix whitespace usage (fow now, previously existing code only just due to
my time shortage):
- Use 8 column for basic indent.
- Use 4 space for 2nd level indent.
- Use tab instead of 8 spaces.
- Don't put space before function call operator. That's unary operator.
- Wrap lines so that it fits in 80 columns.

Remove some one-off debug printf's

Extensive changes from fwnode integration
Move interrupt routine to simply updating state flags and ack'ing interrupts
Move main processing into kthread
Change IPL level to IPL_BIO and try not to hold splbio very much if at all.
Add baseline support for attaching/detaching/updating fwnode's.
Start adding higher level API which isn't tied to if_fw/mbuf's

Remove the use of splimp() from the NetBSD kernel. splnet()
and only splnet() is allowed for the protection of data structures
used by network devices.

Make fwohci works on big endian cpu (e.g., powermac g4).

make compilable again without FW_DEBUG.

Fix to validate selfid buffer according to the OHCI spec.
1. read 'gen' from first quadlet of the buffer.
2. raed selfid packets in the buffer.
3. confirm 'gen' field of the SELFID register matches 1.

Reinitialize context match register at fwohci_hw_init()
to receive asynchronous stream data after suspend/resume.

use OHCI_BITVAL(val, name), instead of
((val) & (name##_MASK)) >> name##_BITPOS

change fw_verbose inside #ifdef FW_DEBUG
0) no verbose messages.
1) print non-fatal errors, nodeid, unique ID for other nodes.
2) extra printfs.

branches: 1.16.2;
Do not require GUID ROM, which is optional according to the OHCI spec.
Use Global UID register as is, it should be initialized by firmware.
If it is not initialized (i.e. 0), try to read GUID ROM if exists.
On my VAIO PCG-N505AS, the version register says it implement GUID ROM,
but reading GUID ROM returns all zero bytes. I'm not sure where the bug is.

fix length calculation not to overrun in receiving packets around MHLEN.

Add RCS Id.

Pass the right size of object (in this case, a trailer) to fwohci_buf_pktget.
Since the fp_trail is a pointer to trailer, we have to pass sizeof(*fp_trail)
instead of sizeof(fp_trail).

Make this file compiles on alpha even if FW_DEBUG is defined.

Pass the right max segment size to bus_dmamap_create.

Add new powerhook argument values, PWR_SOFTSUSPEND, PWR_SOFTSTANDBY and
PWR_SOFTRESUME. Apm calls powerhook with the values in normal interrupt
priority level while others are protected with splhigh().

Use DMA from mbuf instead of copy in transmit.
Still use memcpy in receiving because we must use buffer fill mode
and many packets may share single receive buffer.

XXX: Workaround(?) for CXD3222: it fails to DMA for selfid packet according
to code placement. I'm not sure about the reason (cache? timing? bug?).

Fixed the bug: transmitter sometimes stop and OACTIVE bit of if_fw
never be cleared.
Fixed the bug: freeing free buffer.

Enable ieee1394_drain and ieee1394_watchdog for loss of fragment.

branches: 1.8.2;
Add support for link fragmentation and reassemble for IEEE-1394.
XXX: drain is still not yet implemented, thus memory leak will occur
in case of any of fragment lost.

Add powerhook and shutdownhook to support Notebook.
Use PAGE_SIZE instead of OHCI_PAGE_SIZE for each DMA mem.
OHCI_PAGE_SIZE(2kB) will be round up PAGE_SIZE, anyway.
XXX: need more sophisticated way to manage buffer.
Clear IntEvent first not to lose SelfIDCompleted in multiple bus reset.
Be careful not to clear BusReset int event in SelfID duration to follow
the OHCI spec.
Initiate BusReset if selfid is not completed within 3sec after busreset.

make it compilable with FW_DEBUG

Merge in my some of my stuff.

make compilable without FW_DEBUG

First Prototype implementation of network interface part for IEEE1394 (if_fw).

Current status:
Only OHCI chip is supported (fwohci).
ping (IPv4) works with Sony's implementation (SmartConnect) on Win98.
sometimes works but not stable.
Not implemented yet:
IRM (Isochronous Resource Manager) functionality.
Link layer fragmentation.
Topology map.
More to do:
clean ups
MCAP
charactor device part
dhcp

There is no entry in GENERIC config file yet.
Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.

branches: 1.2.4;
Make this file compile. (I know it's just a placeholder, but I couldn't
resist trying it.)

Beginning of a IEEE 1394 framework. An attachment for PCI OHCI controllers
and bus-independent module that just begins to print things out. No real
code behind it. THIS IS A WORK IN PROGRESS. The *reg.h are woefully
incomplete.

s/vender/vendor/ in comments, log message and fw ohci register property.

branches: 1.20.32;
Use unsigned to avoid undefined behavior. Found by kUBSan.

branches: 1.19.2; 1.19.8;
Listed UniNorth device is in fact the UniNorth 2 interface, rename.
https://pci-ids.ucw.cz/read/PC/106b/0031

branches: 1.18.58;
Bye-bye fw_port.h.

branches: 1.17.20; 1.17.40; 1.17.42;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

branches: 1.16.14; 1.16.16; 1.16.20;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.15.26;
merge ktrace-lwp.

ieee1394 import from FreeBSD.

branches: 1.13.16;
Typos in comments. From OpenBSD.

branches: 1.12.6;
New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

OCHI -> OHCI

branches: 1.10.2;
fix typo: CLear -> Clear

Add mask for getting the ACK bits out of the status area of a DMA context descriptor

- Wrap long line.
- Consistently use #define<TAB> instead of #define<SPACE>.

use OHCI_BITVAL(val, name), instead of
((val) & (name##_MASK)) >> name##_BITPOS

branches: 1.6.2;
Do not require GUID ROM, which is optional according to the OHCI spec.
Use Global UID register as is, it should be initialized by firmware.
If it is not initialized (i.e. 0), try to read GUID ROM if exists.
On my VAIO PCG-N505AS, the version register says it implement GUID ROM,
but reading GUID ROM returns all zero bytes. I'm not sure where the bug is.

Add RCS Id.

Use DMA from mbuf instead of copy in transmit.
Still use memcpy in receiving because we must use buffer fill mode
and many packets may share single receive buffer.

XXX: Workaround(?) for CXD3222: it fails to DMA for selfid packet according
to code placement. I'm not sure about the reason (cache? timing? bug?).

Fixed the bug: transmitter sometimes stop and OACTIVE bit of if_fw
never be cleared.
Fixed the bug: freeing free buffer.

Enable ieee1394_drain and ieee1394_watchdog for loss of fragment.

branches: 1.3.2;
Merge in my some of my stuff.

First Prototype implementation of network interface part for IEEE1394 (if_fw).

Current status:
Only OHCI chip is supported (fwohci).
ping (IPv4) works with Sony's implementation (SmartConnect) on Win98.
sometimes works but not stable.
Not implemented yet:
IRM (Isochronous Resource Manager) functionality.
Link layer fragmentation.
Topology map.
More to do:
clean ups
MCAP
charactor device part
dhcp

There is no entry in GENERIC config file yet.
Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.

branches: 1.1.4;
Beginning of a IEEE 1394 framework. An attachment for PCI OHCI controllers
and bus-independent module that just begins to print things out. No real
code behind it. THIS IS A WORK IN PROGRESS. The *reg.h are woefully
incomplete.

Fix error branches and config pending races in firewire init.

This way, if anything fails, it just fails; you don't panic. This can
happen if suspending and resuming of firewire is broken (e.g., as I
encountered in PR kern/44581).

Change to consistently use M_FW for all malloc/free.
It probably doesn't matter any more, but the code doesn't appear to
have matched its mallocs and frees - so the stats would have been awol.

branches: 1.32.8; 1.32.12;
Revert all previous kmem_ commits. This needs to be done in a different way
because we cannot call kmem_ from an interrupt context. I opened PR/43341 for
it.

Don't allocate sid scratch memory from an interrupt context:

fwohci0: BUS reset
fwohci0: node_id=0xc800ffc0, gen=1, CYCLEMASTER mode
panic: kernel diagnostic assertion "!cpu_intr_p()" failed: file "../../../../ker
n/subr_kmem.c", line 195
fatal breakpoint trap in supervisor mode
trap type 1 code 0 rip ffffffff8022db1d cs 8 rflags 246 cr2 0 cpl 6 rsp fffffff
f80fafb68
breakpoint() at netbsd:breakpoint+0x5
panic() at netbsd:panic+0x2ba
kern_assert() at netbsd:kern_assert+0x2d
kmem_alloc() at netbsd:kmem_alloc+0x18a
fwohci_intr() at netbsd:fwohci_intr+0xbe2
...

I will send-pr for the next one... Looks like someone did not use DIAGNOSTIC
when made the changes.

Bye-bye fw_port.h.

branches: 1.29.18; 1.29.38; 1.29.40;
The argument of *_intr_establish() is changed to fwohci_flit.
Our ieee1394 coordinates from FreeBSD implementation.
Add fwohci_pci_shutdown() and fwohci_pci_power().
Move __KERNEL_RCSID to ahead.

on NetBSD, cardbus_intr_establish's fifth arg's is a symbol which returns
an int, not a void; change fwohci_intr accordingly.

branches: 1.27.2;
+ Sync to FreeBSD.
firewire.c Rev.1.101
firewire.h Rev.1.21
firewirereg.h Rev.1.50
fwdev.c Rev.1.52
fwdma.c Rev.1.9
fwmem.c Rev.1.34
fwohci.c Rev.1.93
fwohcireg.h Rev.1.23
fwohcivar.h Rev.1.16
if_fwip.c Rev.1.16
if_fwipvar.h Rev.1.5
sbp.c Rev.1.92
+ Cleanup macros in fw_port.h.
+ Fix the occurrence of the error at the resume. Don't set the buffer again.

branches: 1.26.6; 1.26.8; 1.26.12;
firewire.c sync to Rev.1.86 for FreeBSD.
firewirereg.h sync to Rev.1.41 for FreeBSD.
fwcrom.c sync to Rev.1.14 for FreeBSD.
fwdev.c sync to Rev.1.49 for FreeBSD.
fwmem.c sync to Rev.1.32 for FreeBSD.
fwohci.c sync to Rev.1.86 for FreeBSD.
fwohcivar.h sync to Rev.1.15 for FreeBSD.
if_fwip.c sync to Rev.1.14 for FreeBSD.
if_fwipvar.h sync to Rev.1.4 for FreeBSD.
sbp.c sync to Rev.1.89 for FreeBSD.

branches: 1.25.26; 1.25.30; 1.25.32;
merge ktrace-lwp.

ieee1394 import from FreeBSD.

branches: 1.23.2;
fix handling of volatile and remove bogus casts.

nuke trailing whitespace

branches: 1.21.8; 1.21.10;
function prototype must not have variable name

branches: 1.20.6;
Improve performance with disks almost 60x by redoing the callback handler to
accept ranges as well as single addresses. Still need to go through any key
areas and remove the malloc's and replace these with some sort of pooling
instead.

New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

don't pull in sys/mbuf.h just for prototype decl

branches: 1.17.8;
Fix illegal pointer manipulations that cause alignment problems.

implement fwohci detach routine.

New Feature: Add isochronous stream input routine. This feature has
been tested on i386. It does not have any interface for useland to
get isochoronous stream. The isochoronous acquisition interface
should be determined.

branches: 1.14.2;
Various small changes.
Clean up one bug in a DPRINTF in arrs_input which could panic on some packets.
Gut the ack/response functionality and clean it up so all packets get checked
correctly and the abuf struct isn't used once the ab_cb has happened (there
still could be ack packets waiting to be processed at that time).
Finally, add some documentation explaining read/write/inreg and their
purpose/argument calling.

Add locators support

Add first pass of generalized hooks for packet completion (i.e. ack) codes. This will eventually replace the mbuf code when it moves back to if_fw.c where it belongs.

Style police. Got the tab'ing correct

Fix whitespace usage (fow now, previously existing code only just due to
my time shortage):
- Use 8 column for basic indent.
- Use 4 space for 2nd level indent.
- Use tab instead of 8 spaces.
- Don't put space before function call operator. That's unary operator.
- Wrap lines so that it fits in 80 columns.

Add proc struct for kthread, interrupt handler flags, and nodelist

Make fwohci works on big endian cpu (e.g., powermac g4).

branches: 1.7.2;
Add RCS Id.

Use DMA from mbuf instead of copy in transmit.
Still use memcpy in receiving because we must use buffer fill mode
and many packets may share single receive buffer.

XXX: Workaround(?) for CXD3222: it fails to DMA for selfid packet according
to code placement. I'm not sure about the reason (cache? timing? bug?).

Fixed the bug: transmitter sometimes stop and OACTIVE bit of if_fw
never be cleared.
Fixed the bug: freeing free buffer.

Enable ieee1394_drain and ieee1394_watchdog for loss of fragment.

branches: 1.5.2;
Add support for link fragmentation and reassemble for IEEE-1394.
XXX: drain is still not yet implemented, thus memory leak will occur
in case of any of fragment lost.

Add powerhook and shutdownhook to support Notebook.
Use PAGE_SIZE instead of OHCI_PAGE_SIZE for each DMA mem.
OHCI_PAGE_SIZE(2kB) will be round up PAGE_SIZE, anyway.
XXX: need more sophisticated way to manage buffer.
Clear IntEvent first not to lose SelfIDCompleted in multiple bus reset.
Be careful not to clear BusReset int event in SelfID duration to follow
the OHCI spec.
Initiate BusReset if selfid is not completed within 3sec after busreset.

Merge in my some of my stuff.

First Prototype implementation of network interface part for IEEE1394 (if_fw).

Current status:
Only OHCI chip is supported (fwohci).
ping (IPv4) works with Sony's implementation (SmartConnect) on Win98.
sometimes works but not stable.
Not implemented yet:
IRM (Isochronous Resource Manager) functionality.
Link layer fragmentation.
Topology map.
More to do:
clean ups
MCAP
charactor device part
dhcp

There is no entry in GENERIC config file yet.
Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.

branches: 1.1.4;
Beginning of a IEEE 1394 framework. An attachment for PCI OHCI controllers
and bus-independent module that just begins to print things out. No real
code behind it. THIS IS A WORK IN PROGRESS. The *reg.h are woefully
incomplete.

Bye-bye fw_port.h.

branches: 1.2.18; 1.2.76; 1.2.98; 1.2.100;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

First pass (checkpoint checkin) for firewire scsi code.

There are a number of issues here for anyone trying to use this today:

1. On my test drive the command engine on the drive seems to stall after the
inquire is done. So the mode sense times out for a long time before
aborting. This obviously needs to be tracked down and fixed.

However it does do a proper inquire:

scsibus0 at sbpscsi0: 1 target, 1 lun per target
sd0 at scsibus0 target 0 lun 0: <Maxtor, 1394 storage, 60> disk fixed

2. This code is quite ugly in places as debug code was added to test things.
Definitly needs cleanup/documention in places where it's using command
structures. The structure for alloc'ing orbs, running them through the
command engine and getting state back is mostly set but implementation needs
an overhaul in places.

3. For testing I use the following config options:

fwohci* at cardbus? dev ? function ? # IEEE1394 Controller
fw* at fwbus?
options FW_DEBUG
options FWNODE_DEBUG
options P1212_DEBUG
options SBP2_DEBUG
options SBPSCSI_DEBUG
fwnode* at fwbus?
sbpscsi* at fwnode?
scsibus* at sbpscsi?

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

remove trailing \n in panic(). approved perry.

Basic prototype/beginning of firewire disk support using sbp2/scsi. Commiting
to have a baseline.

don't need <sys/types.h> when including <sys/param.h>

add RCSID

branches: 1.2.2; 1.2.4;
Style police. Got the tab'ing correct

Place holder until real fwscsi code is fleshed out.

Fix build.

branches: 1.5.8; 1.5.18; 1.5.22;
- print expected crc when it does not match
- add and print secondary text leaf. This prints the model name of my webcam.
- add csr keys that are printed as 'unknown' otherwise

Bye-bye fw_port.h.

branches: 1.3.20; 1.3.22;
Constify, so that usr.sbin/fwctl can be built with WARNS=4.

branches: 1.2.18; 1.2.72; 1.2.74; 1.2.76;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

Bye-bye fw_port.h.

branches: 1.3.14; 1.3.16;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.2.18; 1.2.72; 1.2.76; 1.2.78; 1.2.82;
merge ktrace-lwp.

branches: 1.1.6;
ieee1394 import from FreeBSD.

ieee1394 import from FreeBSD.

branches: 1.18.4;
nuke trailing whitespace

branches: 1.17.8; 1.17.10;
sprinkle some LL suffixes on >32 bit number literals

branches: 1.16.2;
Move SBP data area's below the 4G mark as some broken devices seem to clip data
address's to 32 bits.
XXX - Need to move the if_fw fifo addr as well but this needs some testing

Provide a macro for turning a speed number into a max rec value

New Feature: add pseudo device for IEEE 1394 isochronous stream and
isochronous reception routine for IEEE 1394 OHCI (fwohci). The
transmission part is under construction.

The minimum configuration options for this feature are:

# IEEE 1394 (i.LINK)
fwohci* at pci? dev ? function ?
pseudo-device fwiso 1

First pass (checkpoint checkin) for firewire scsi code.

There are a number of issues here for anyone trying to use this today:

1. On my test drive the command engine on the drive seems to stall after the
inquire is done. So the mode sense times out for a long time before
aborting. This obviously needs to be tracked down and fixed.

However it does do a proper inquire:

scsibus0 at sbpscsi0: 1 target, 1 lun per target
sd0 at scsibus0 target 0 lun 0: <Maxtor, 1394 storage, 60> disk fixed

2. This code is quite ugly in places as debug code was added to test things.
Definitly needs cleanup/documention in places where it's using command
structures. The structure for alloc'ing orbs, running them through the
command engine and getting state back is mostly set but implementation needs
an overhaul in places.

3. For testing I use the following config options:

fwohci* at cardbus? dev ? function ? # IEEE1394 Controller
fw* at fwbus?
options FW_DEBUG
options FWNODE_DEBUG
options P1212_DEBUG
options SBP2_DEBUG
options SBPSCSI_DEBUG
fwnode* at fwbus?
sbpscsi* at fwnode?
scsibus* at sbpscsi?

Move definition of if_fw fifo location to ieee1394reg.h as it's needed for
filling in unit location in the rom.

XXX: This all is really too welded together. The if_fw code should be
decoupled from this and simply be handed a method by which it can provide
the config rom back to the main code.

Add 3 more macros for extracting info from bus info block

Add IEEE1394_MAX_REC macro