branches: 1.4.8;
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.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.2.2; 1.2.4;
Install only headers that are useful to userspace (ioctl interface is not
finished, but will be here soon).

Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

This isn't relevant any more.

branches: 1.4.2;
- Allocate any private space that's requested.
- Fix a few bogons with the last update.
- Add some more LAN defs.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.2.2;
- Only configure un-claimed devices. Implement re-configuration.
- Adhere to the spec better in some places. Also, work around some quirks
noted in the Linux I2O code.
- Register event handlers for the executive and RBS devices.
- Fix a number of buglets and tidy a little.
- Implement a message pass-through interface and some other useful ioctls.

branches: 1.1.2;
TODO

Make the I2O management cdevs MPSAFE.

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: 1.49.2; 1.49.4;
Kill some more extern struct cfdriver declarations.

Down with externs in .c!

Add d_discard to all struct cdevsw instances I could find.

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

branches: 1.47.2;
make this compile again.
XXX: 32 bit pointer issues.

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: 1.45.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.44.2; 1.44.12;
Replace some wakeup_one(9) uses with mutex(9) or plain wakeup(9).

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.

branches: 1.42.4;
struct device * -> device_t, no functional changes intended.

struct cfdata * -> cfdata_t, no functional changes intended.

branches: 1.40.8;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.39.4;
Replace device_lookup() with device_lookup_private() on getting softc
for future device_t/softc spilt.

Include machine/cputypes.h explicitly.

branches: 1.37.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.36.2; 1.36.4;
use aprint_*_dev and device_xname

branches: 1.35.16;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.34.6; 1.34.8; 1.34.12;
- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.33.2; 1.33.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.32.2;
Replace some uses of lockmgr() / simplelocks.

Change kauth(9) KPI for kauth_authorize_device_passthru() to add another
argument, u_long, serving as a bit-mask of generic requests for the
passthru request.

Discussed on tech-security@ and tech-kern@. Okay tls@.

branches: 1.30.2;
__unused removal on arguments; approved by core.

Replace securelevel checks with kauth(9) calls.

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

fix empty if

branches: 1.26.2; 1.26.4;
fix incomplete initializers.

Acting as Elad's keyboard:

Place the securelevel checks in their logical locations.
This will be clearer in the future when code changes to use kauth(9) calls.

input and okay ad@

Use device_private().

Use device_unit().

Use device_parent().

branches: 1.21.4; 1.21.6; 1.21.8; 1.21.10; 1.21.12;
merge ktrace-lwp.

Initialize "mapped" before any of the "goto"'s, so we're not testing an
uninitialized variable at "bad:".

branches: 1.19.4; 1.19.10;
nuke trailing whitespace

branches: 1.18.4; 1.18.6;
sprintf -> snprintf

Grab the configuration lock when handling I2ORESCANCMD.

Remove some assigned-to but otherwise unused variables.

Reorder, eliminating a bogus initializer.

Fix uninitialized variable warnings

branches: 1.13.2;
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

__'ify MACHINE_ARCH names.

- Handle DPT ioctls from native NetBSD binaries.
- SCSI targets could disapper mid-way through autoconfig if the board
notifed us of a changed LCT at the wrong time. Fix that.

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

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

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.

Use __mips__ instead of mips.

branches: 1.3.2; 1.3.4; 1.3.12; 1.3.14;
add RCSID

branches: 1.2.2;
Make firmware updates work with dpti.

branches: 1.1.2;
Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

Fix various typos in comments:
s/defininitions/definitions/
s/ininitialise/initialise/
s/collasped/collapsed/
s/optionaly/optionally/

branches: 1.10.10;
fix typos, mainly in words minimum and maximum, but also few others.

split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.8.2; 1.8.12;
Replace some wakeup_one(9) uses with mutex(9) or plain wakeup(9).

Remove clause 3 and 4 from TNF licenses

branches: 1.6.36; 1.6.38; 1.6.40;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.5.26;
merge ktrace-lwp.

branches: 1.4.4;
nuke trailing whitespace

branches: 1.3.6; 1.3.14; 1.3.16;
Implement the DPT EATA ioctl() interface.

branches: 1.2.2; 1.2.6; 1.2.8; 1.2.18;
Make firmware updates work with dpti.

branches: 1.1.2;
Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

The I2O LAN stuff was never tested and is just more code to maintain -
remove it.

I2OVERBOSE/opt_i2o.h were removed

defopt I2OVERBOSE

branches: 1.7.6; 1.7.8; 1.7.12; 1.7.14;
- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.6.32; 1.6.74; 1.6.76;
- 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

branches: 1.5.4;
Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

branches: 1.4.2;
Add iopl.

branches: 1.3.2; 1.3.4;
whitespace police

lsu -> ld, by popular request.

branches: 1.1.2;
Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

Remove unused I2O LAN class defs (the LAN driver was deleted years ago).

s/identifing/identifying/ and s/multipler/multiplier/ in comments.

Add some #includes to make a few more header files compile.

branches: 1.15.38;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.14.2; 1.14.6;
Remove clause 3 and 4 from TNF licenses

branches: 1.13.70; 1.13.72; 1.13.74;
merge ktrace-lwp.

nuke trailing whitespace

branches: 1.11.4; 1.11.6;
Spell removable with only two es. Inspired by jmc@openbsd.

Fix some typos. From Tom Cosgrove via jmc@openbsd.

branches: 1.9.4; 1.9.20;
Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

branches: 1.8.2;
Parameter group changes:

- Do FIELD_SET correctly for scalar types.
- Add some basic table ops.
- Push error reporting back into iop.c.
- Add some field index and (yet more) LAN defs to i2o.h.

Add more LAN defs.

Some changes from OpenBSD, and some of my own:

- More SCSI port defs.
- Nuke vtophys().
- Release resources in iop_init() upon failure.
- Don't use a message wrapper when initalising the outbound FIFO.
- A couple of field size/endian fixes.
- Just use iop_post() when we don't need special handling.
- IM_DISCARD is now pointless, since we don't queue at the driver level.
- Map data transfers from/to userspace directly.
- A few comment and stylistic changes.

branches: 1.5.2;
receive, not recieve

- Allocate any private space that's requested.
- Fix a few bogons with the last update.
- Add some more LAN defs.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.2.2;
- Only configure un-claimed devices. Implement re-configuration.
- Adhere to the spec better in some places. Also, work around some quirks
noted in the Linux I2O code.
- Register event handlers for the executive and RBS devices.
- Fix a number of buglets and tidy a little.
- Implement a message pass-through interface and some other useful ioctls.

branches: 1.1.2;
Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.4.2; 1.4.6;
Remove clause 3 and 4 from TNF licenses

branches: 1.3.70; 1.3.72; 1.3.74;
merge ktrace-lwp.

nuke trailing whitespace

branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10; 1.1.26; 1.1.34; 1.1.36;
Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

Make the I2O management cdevs MPSAFE.

Merge thorpej-cfargs2.

branches: 1.91.8;
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: 1.90.10;
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.

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: 1.88.2; 1.88.4;
Kill some more extern struct cfdriver declarations.

Down with externs in .c!

Fix a buffer pointer check in iop_msg_unmap(). The number of im_xfer
is IOP_MAX_MSG_XFERS and it's three. iop_systab_set() makes three ximit
buffers. In this case, iop_msg_unmap() overrruns by "if((++ix)->ix_size == 0)".
Check overrun first.

This change fixes a bug that iop(4) panics while attaching iopsp(4).

branches: 1.86.4;
Add d_discard to all struct cdevsw instances I could find.

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

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.

remove unused variables and __USE a debugging variable

#if 0 iop_status, but keep it for documentation purposes.

branches: 1.82.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.81.8; 1.81.18;
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.

branches: 1.80.2;
Fix confusion about PRE/POST in previous; and another place that has the
same issue. Also add a BUS_DMASYNC_POSTWRITE to the POLL operation.
Problem pointed out by tsutsui@.
Still not perfect (a BUS_DMASYNC_PREREAD would be needed after the
testing the condition in the POLL() macro), but closer.

Fix bus_dmamap_sync(): it should be a BUS_DMASYNC_POSTWRITE after
writing to DMA memory.
Found by code inspection.

Don't mix BUS_DMASYNC_PREREAD and BUS_DMASYNC_POSTWRITE.
Suggested by <bouyer>, fixes PR kern/42662

branches: 1.77.2;
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>.

struct device * -> device_t, no functional changes intended.

branches: 1.75.8;
Use the physical address from the bus_dmamap_load() map, not the one from
the bus_dmamem_alloc(). Fixes iop(4) on alpha, and possibly sparc64 as
described in the thread
http://mail-index.netbsd.org/port-sparc64/2008/06/04/msg000413.html.

Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.73.4;
Replace device_lookup() with device_lookup_private() on getting softc
for future device_t/softc spilt.

Fix up bus_dmamap_sync calls.

branches: 1.71.2;
- Convert hashinit() to use kmem_alloc(). The hash tables can be large
and it's better to not have them in kmem_map.
- Convert a couple of minor items along the way to kmem_alloc().
- Fix some memory leaks.

Remove clause 3 and 4 from TNF licenses

branches: 1.69.2; 1.69.4;
use aprint_*_dev and device_xname

branches: 1.68.12;
Match the docs: MUTEX_DRIVER/SPIN are now only for porting code written
for Solaris.

branches: 1.67.2; 1.67.4;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.66.6; 1.66.8; 1.66.12;
Merge some of the less invasive changes from the vmlocking branch:

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

- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.64.2; 1.64.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.63.2;
Replace some uses of lockmgr() / simplelocks.

Change kauth(9) KPI for kauth_authorize_device_passthru() to add another
argument, u_long, serving as a bit-mask of generic requests for the
passthru request.

Discussed on tech-security@ and tech-kern@. Okay tls@.

branches: 1.61.2;
__unused removal on arguments; approved by core.

Replace securelevel checks with kauth(9) calls.

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

fix empty if.

branches: 1.57.2; 1.57.4;
- fix initializers
- comment out dead code.

Acting as Elad's keyboard:

Place the securelevel checks in their logical locations.
This will be clearer in the future when code changes to use kauth(9) calls.

input and okay ad@

Coverity CID 1081: Avoid NULL deref.

Use device_private().

Use device_unit().

A device that has just been detached can't be reconfigured. Coverity ID 2674.
Note: this was not a problem right now because there's no drivers with both
a detach and a ii_reconfig callback.

branches: 1.51.4; 1.51.6; 1.51.8; 1.51.10; 1.51.12;
merge ktrace-lwp.

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.

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

branches: 1.48.2;
avoid variable shadowing.

nuke trailing whitespace

branches: 1.46.4; 1.46.6;
oops, another typo, should fix report by taya

actually pass the locators,
should fix panic reported by Shin'ichiro TAYA

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"

sprintf -> snprintf

Match Promise controllers.

Remove some assigned-to but otherwise unused variables.

Fix the previous differently.

Fix uninitialized variable warnings

branches: 1.38.2;
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

DMA, not dma nor Dma.

Remove variable that is only assigned too but not referenced.

Merge the nathanw_sa branch.

Use aprint_normal() in cfprint routines.

Fix typo.

- Handle DPT ioctls from native NetBSD binaries.
- SCSI targets could disapper mid-way through autoconfig if the board
notifed us of a changed LCT at the wrong time. Fix that.

Cosmetic changes.

Add support for Adaptec 200[05]S
reviewed by Andrew Doran

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

PR 18760: don't issue I2O_EXEC_IOP_CLEAR to AMI boards, since some
firmware revisions can't handle it.

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.

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.

branches: 1.24.2; 1.24.4;
Implement mstohz() as discussed on tech-kern, and use it in SCSI drivers
to convert xs->timeout to callout() parameter.

PR 15552: don't explode if the board fails to respond during initalization.

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

Set sc_framesize to a reasonable value early on so that DEBUG code
doesn't choke before we discern the real value.

add RCSID

branches: 1.19.4;
Make firmware updates work with dpti.

Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

Give initiali[sz]e all the "i"s it deserves.

branches: 1.16.2;
Parameter group changes:

- Do FIELD_SET correctly for scalar types.
- Add some basic table ops.
- Push error reporting back into iop.c.
- Add some field index and (yet more) LAN defs to i2o.h.

Some changes from OpenBSD, and some of my own:

- More SCSI port defs.
- Nuke vtophys().
- Release resources in iop_init() upon failure.
- Don't use a message wrapper when initalising the outbound FIFO.
- A couple of field size/endian fixes.
- Just use iop_post() when we don't need special handling.
- IM_DISCARD is now pointless, since we don't queue at the driver level.
- Map data transfers from/to userspace directly.
- A few comment and stylistic changes.

branches: 1.14.2;
receive, not recieve

- Allocate any private space that's requested.
- Fix a few bogons with the last update.
- Add some more LAN defs.

Decrease the timeouts on QUEIESCE and CLEAR lest the adapter doesn't reply.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.10.2;
Fix thinko in previous.

- Fix a number of bugs with the configuration stuff.
- Bash the passthrough interface into working order.
- Add an ioctl to retrieve the `tidmap'.

Don't try to configure `uninteresting' devices.

Nit.

Fix a comment.

- Only configure un-claimed devices. Implement re-configuration.
- Adhere to the spec better in some places. Also, work around some quirks
noted in the Linux I2O code.
- Register event handlers for the executive and RBS devices.
- Fix a number of buglets and tidy a little.
- Implement a message pass-through interface and some other useful ioctls.

branches: 1.4.2;
NBPG -> PAGE_SIZE

unused variable

- A few minor changes (typos and the like).
- Use SPLs in iop_msg_send() and not the interrupt mask - it accesses
shared resources lower down.

Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

fix various typos in comments.

Add some #includes to make a few more header files compile.

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/).

branches: 1.6.44; 1.6.64;
Remove clause 3 and 4 from TNF licenses

branches: 1.5.6; 1.5.8; 1.5.10;
Don't get under the device_t abstraction by measuring the length
of dv_xname, especially not to set the length of a buffer that
(apparently) belongs to the kernel ABI. Instead, set the buffer
length to 16, which is the current length of dv_xname.

branches: 1.4.28; 1.4.34;
undo previous. Andy says the struct is used in pkgsrc.

Sprinkle _KERNEL.

branches: 1.2.4; 1.2.36; 1.2.50;
Make firmware updates work with dpti.

branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8;
Split limits and ioctl defs out of iopvar.h.

The I2O LAN stuff was never tested and is just more code to maintain -
remove it.

I2OVERBOSE/opt_i2o.h were removed

defopt I2OVERBOSE

branches: 1.29.2;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.28.4;
Once again, use ether_ioctl() instead of (cmd == SIOCADDMULTI) ?
ether_addmulti() : ether_delmulti().

branches: 1.27.2;
Constify.

branches: 1.26.2; 1.26.6;
Merge some of the less invasive changes from the vmlocking branch:

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

- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.24.2; 1.24.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.23.2;
Fix uninitialised warnings. All are false positives, except one which was
indeed real.

branches: 1.22.2;
remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.

Use device_private().

Use device_parent().

branches: 1.19.4; 1.19.6; 1.19.8; 1.19.10; 1.19.12;
merge ktrace-lwp.

branches: 1.18.2;
split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.

nuke trailing whitespace

Remove an erroneous semicolon.

Eliminate use of M_HASFCS.

branches: 1.14.4; 1.14.6;
When adding/deleting multicast addresses, only whack the address
filter if the interface is marked RUNNING.

Fixes kern/27678.

s/the the/the/ (only in sources that aren't regularly imported from
elsewhere).

sprintf -> snprintf

branches: 1.11.6; 1.11.10;
Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

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

Bump IOPL_MAX_SEGS to 16 in line with other Ethernet drivers, and in one
spot replace it with the capped value learned from the board.

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

add RCSID

Give initiali[sz]e all the "i"s it deserves.

branches: 1.3.2; 1.3.4; 1.3.6;
Implement/fix FDDI, receive checksums, media selection and a ton of other
stuff.

Thinko.

Rough I2O LAN interface driver. Needs a bit of work.

The I2O LAN stuff was never tested and is just more code to maintain -
remove it.

branches: 1.5.44; 1.5.46; 1.5.52;
merge ktrace-lwp.

branches: 1.4.4;
nuke trailing whitespace

branches: 1.3.10; 1.3.18; 1.3.20;
Bump IOPL_MAX_SEGS to 16 in line with other Ethernet drivers, and in one
spot replace it with the capped value learned from the board.

branches: 1.2.2; 1.2.4;
Implement/fix FDDI, receive checksums, media selection and a ton of other
stuff.

Rough I2O LAN interface driver. Needs a bit of work.

Remove clause 3 and 4 from TNF licenses

branches: 1.1.2; 1.1.12; 1.1.126; 1.1.128; 1.1.130;
Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

fix various typos in comments and log messages.

Merge thorpej-cfargs2.

branches: 1.39.8;
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: 1.38.10;
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.

branches: 1.37.18;
Fix device_t/softc split bug.

branches: 1.36.14;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.35.12; 1.35.22;
struct device * -> device_t, no functional changes intended.

struct cfdata * -> cfdata_t, no functional changes intended.

branches: 1.33.8;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.32.2; 1.32.6;
Remove clause 3 and 4 from TNF licenses

branches: 1.31.2; 1.31.4;
use aprint_*_dev and device_xname

branches: 1.30.14;
Noisy printfs.

branches: 1.29.2;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.28.6; 1.28.8; 1.28.12;
- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.27.2; 1.27.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.26.2;
Replace some uses of lockmgr() / simplelocks.

__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: 1.23.8; 1.23.10;
Use device_private().

Use device_parent().

branches: 1.21.2; 1.21.4; 1.21.6; 1.21.8; 1.21.10;
Make almost everything #include <sys/bswap.h> instead of <machine/bswap.h>
The bswap.h and endian.h files are all rather incestuous, but I want to
get the constant folding stuff into one place - sys/bswap.h

branches: 1.20.2;
merge ktrace-lwp.

branches: 1.19.4;
nuke trailing whitespace

branches: 1.18.2; 1.18.10; 1.18.12;
Fix the way the scsipi adapter openings are adjusted when the array
controller adjusts our queue length.

- Handle DPT ioctls from native NetBSD binaries.
- SCSI targets could disapper mid-way through autoconfig if the board
notifed us of a changed LCT at the wrong time. Fix that.

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

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

branches: 1.12.10;
Call malloc(9) with M_ZERO flag instead of memset() after malloc().

add RCSID

branches: 1.10.6;
Parameter group changes:

- Do FIELD_SET correctly for scalar types.
- Add some basic table ops.
- Push error reporting back into iop.c.
- Add some field index and (yet more) LAN defs to i2o.h.

Some changes from OpenBSD, and some of my own:

- More SCSI port defs.
- Nuke vtophys().
- Release resources in iop_init() upon failure.
- Don't use a message wrapper when initalising the outbound FIFO.
- A couple of field size/endian fixes.
- Just use iop_post() when we don't need special handling.
- IM_DISCARD is now pointless, since we don't queue at the driver level.
- Map data transfers from/to userspace directly.
- A few comment and stylistic changes.

branches: 1.8.2;
Make it compile with SCSIPI_DEBUG

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

- Allocate any private space that's requested.
- Fix a few bogons with the last update.
- Add some more LAN defs.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.4.2;
FC peripherals on DPT/Adaptec boards only respond to private class messages.

- Only configure un-claimed devices. Implement re-configuration.
- Adhere to the spec better in some places. Also, work around some quirks
noted in the Linux I2O code.
- Register event handlers for the executive and RBS devices.
- Fix a number of buglets and tidy a little.
- Implement a message pass-through interface and some other useful ioctls.

branches: 1.2.2;
- A few minor changes (typos and the like).
- Use SPLs in iop_msg_send() and not the interrupt mask - it accesses
shared resources lower down.

Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.8.34; 1.8.44;
Remove clause 3 and 4 from TNF licenses

branches: 1.7.28; 1.7.30; 1.7.32;
- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.6.18; 1.6.60; 1.6.62;
Fix the way the scsipi adapter openings are adjusted when the array
controller adjusts our queue length.

branches: 1.5.6;
Some changes from OpenBSD, and some of my own:

- More SCSI port defs.
- Nuke vtophys().
- Release resources in iop_init() upon failure.
- Don't use a message wrapper when initalising the outbound FIFO.
- A couple of field size/endian fixes.
- Just use iop_post() when we don't need special handling.
- IM_DISCARD is now pointless, since we don't queue at the driver level.
- Map data transfers from/to userspace directly.
- A few comment and stylistic changes.

branches: 1.4.2;
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

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.2.2; 1.2.4;
- A few minor changes (typos and the like).
- Use SPLs in iop_msg_send() and not the interrupt mask - it accesses
shared resources lower down.

Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.23.12; 1.23.22;
struct device * -> device_t, no functional changes intended.

branches: 1.22.8;
Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

branches: 1.21.2; 1.21.6;
Remove clause 3 and 4 from TNF licenses

branches: 1.20.28; 1.20.30; 1.20.32;
Merge some of the less invasive changes from the vmlocking branch:

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

- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.18.2; 1.18.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.2;
Replace some uses of lockmgr() / simplelocks.

merge ktrace-lwp.

branches: 1.15.4;
nuke trailing whitespace

branches: 1.14.4; 1.14.6;
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"

Match Promise controllers.

branches: 1.12.6;
Cosmetic changes.

Add support for Adaptec 200[05]S
reviewed by Andrew Doran

branches: 1.10.4; 1.10.14;
Make firmware updates work with dpti.

Add `dpti', which is an I2O vendor extension implementing the DPT/Adaptec
control interface. This makes it possible to run the Linux versions of
dptmgr, raidutil, dptelog and other tools under NetBSD.

branches: 1.8.2;
Parameter group changes:

- Do FIELD_SET correctly for scalar types.
- Add some basic table ops.
- Push error reporting back into iop.c.
- Add some field index and (yet more) LAN defs to i2o.h.

Some changes from OpenBSD, and some of my own:

- More SCSI port defs.
- Nuke vtophys().
- Release resources in iop_init() upon failure.
- Don't use a message wrapper when initalising the outbound FIFO.
- A couple of field size/endian fixes.
- Just use iop_post() when we don't need special handling.
- IM_DISCARD is now pointless, since we don't queue at the driver level.
- Map data transfers from/to userspace directly.
- A few comment and stylistic changes.

branches: 1.6.2;
- Allocate any private space that's requested.
- Fix a few bogons with the last update.
- Add some more LAN defs.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.4.2;
- Fix a number of bugs with the configuration stuff.
- Bash the passthrough interface into working order.
- Add an ioctl to retrieve the `tidmap'.

- Only configure un-claimed devices. Implement re-configuration.
- Adhere to the spec better in some places. Also, work around some quirks
noted in the Linux I2O code.
- Register event handlers for the executive and RBS devices.
- Fix a number of buglets and tidy a little.
- Implement a message pass-through interface and some other useful ioctls.

branches: 1.2.2;
- A few minor changes (typos and the like).
- Use SPLs in iop_msg_send() and not the interrupt mask - it accesses
shared resources lower down.

Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.

ld(4): Convert blkno argument for sc_dump() to daddr_t

PR kern/59153

(1) For backends that accept 64-bit block address, i.e.,
nvme(4), virtio(4), aac(4), iop(4), and mainbus(usermode/4),
this should enable to dump beyond 2Gi blocks.

(2) sdmmc(4) backend allows to dump up to the last block.

(3) For other backends, block address is handled as `int`.
Some of them may support blocks up to 4Gi, but I do not have
enough time to examine datasheets. So, continue to reject >2Gi
blocks as before.

XXX
This is KABI change, and cannot be pulled up into netbsd-{10,9}.

XXX
Compile-test only (for amd64/ALL) due to lack of large SSDs ;)

Thanks mlelstv@ for discussion and careful review!!

fix a/an grammar in obvious cases

branches: 1.40.6;
Fix typo.

While ld(4) is MP safe, many backends are not.

Add a flag for backends that are MP safe. Take KERNEL_LOCK when calling
into a backend that doesn't have the flag set. Do the same for the
discard routine.

Fixes PR 52462.

Don't wait for allocation in polling mode.

branches: 1.37.6;
refactor the ld(4) DIOCCACHESYNC hook into general ioctl hook, so that attachments
would be able to implement arbitrary other ioctls

branches: 1.36.2;
modify ldattach() to have default strategy as a parameter

branches: 1.35.2;
Convert sys/dev to use <sys/rndsource.h>.

branches: 1.34.6; 1.34.24;
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.

branches: 1.33.14; 1.33.18;
More device_t/softc split fallout, this time in iop(4). Fix from
Juan Romero Pardines. Tested on alpha by me.

branches: 1.32.2; 1.32.4;
Complete device_t/softc split for the ld(4) attachments. This should
prevent crashes while attaching a drive.

Patch supplied by Juan RP in PR kern/39468.

Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h
preprocessor macros.

Add a flags argument to the ld(4) flush handlers, and call these with a
"poll" flag when called from ld_shutdown().

This is the infrastructure part of kern/38655 - in itself it doesn't
fix the panic referenced in that PR.

XXX: At least ld_twa.c and perhaps ld_iop.c and ld_icp.c need to
check for this new poll flag and do something useful.

branches: 1.29.2; 1.29.4;
Fix a tyop in a comment.

branches: 1.28.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.27.2; 1.27.4;
use aprint_*_dev and device_xname

branches: 1.26.16;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.25.4; 1.25.6; 1.25.10; 1.25.12;
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: 1.24.2;
- Merge changes from the vmlocking branch.
- Eliminate opt_i2o.h, since I2OVERBOSE is gone.

branches: 1.23.8; 1.23.10;
__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: 1.21.8; 1.21.10;
Coverity CID 1532: Off-by-one array overrun.

Use device_private().

Use device_parent().

branches: 1.18.4; 1.18.6; 1.18.8; 1.18.10; 1.18.12;
merge ktrace-lwp.

branches: 1.17.2;
add const

nuke trailing whitespace

branches: 1.15.4; 1.15.6;
move buffer queue related stuffs from buf.h to their own header, bufq.h.

Spell removable with only two es. Inspired by jmc@openbsd.

branches: 1.13.6;
Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

add RCSID

branches: 1.9.6;
Parameter group changes:

- Do FIELD_SET correctly for scalar types.
- Add some basic table ops.
- Push error reporting back into iop.c.
- Add some field index and (yet more) LAN defs to i2o.h.

Some changes from OpenBSD, and some of my own:

- More SCSI port defs.
- Nuke vtophys().
- Release resources in iop_init() upon failure.
- Don't use a message wrapper when initalising the outbound FIFO.
- A couple of field size/endian fixes.
- Just use iop_post() when we don't need special handling.
- IM_DISCARD is now pointless, since we don't queue at the driver level.
- Map data transfers from/to userspace directly.
- A few comment and stylistic changes.

branches: 1.7.2;
Centralise the geometry fudge.

Major update; basically, improve performance somewhat and play nicer with
Intel and AMI IOPs. There still exists one nasty problem with Intel adapters
when under load, which I'm working on.

branches: 1.5.2;
Sync with ld.c.

Make dumps work, and put the first part of detach into ld.c.

branches: 1.3.2;
Oops.

- Only configure un-claimed devices. Implement re-configuration.
- Adhere to the spec better in some places. Also, work around some quirks
noted in the Linux I2O code.
- Register event handlers for the executive and RBS devices.
- Fix a number of buglets and tidy a little.
- Implement a message pass-through interface and some other useful ioctls.

lsu -> ld, by popular request.

lsu -> ld

branches: 1.1.2;
Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs
more testing with different configurations, and work in a number of areas
(which I'm not able to do for a little while), but is at least functional
and stable on i386 with DPT adapters.