Copy stable/9 to releng/9.3 as part of the 9.3-RELEASE cycle.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

Big MFC of the physbio changes necessary for unmapped I/O. These changes
have been in production at Netflix for several months with significant
success.

MFC r246713:

Reform the busdma API so that new types may be added without modifying
every architecture's busdma_machdep.c. It is done by unifying the
bus_dmamap_load_buffer() routines so that they may be called from MI
code. The MD busdma is then given a chance to do any final processing
in the complete() callback.

MFC r249538:
Some compilers issue a warning when wider integer is casted to narrow
pointer. Supposedly shut down the warning by casting through
uintptr_t.

MFC r251479:
Simplify the checking of flags for cam_periph_mapmem(). This gets rid of
a lot of code redundancy and grossness at very minor expense.

MFC r251837:
MFC r251842:
Add infrastructure for doing compatibility shims, as has been sorely
needed for the last 10 years. Far too much of the internal API is
exposed, and every small adjustment causes applications to stop working.
To kick this off, bump the API version to 0x17 as should have been done
with r246713, but add shims to compensate. Thanks to the shims, there
should be no visible change in application behavior.

Submitted by: kib, jeffr
Approved by: kib
Obtained from: Netflix

MFC 241611:
Fix aha(4) build with i386 LINT (which includes 'device mca').

MFC 241492, 241588, 241589, 241590, 241592, 241593, 241603, 241605

Modernize and lock the aha, ahb, adv, adw, bt, and dpt drivers.

Submitted by: jhb
Obtained from: Netflix

Toss bogus mergeinfo.

MFC r235634

Fix and update battery status bits according to linux driver

Copy head to stable/9 as part of 9.0-RELEASE release cycle.

Approved by: re (implicit)

Pass bus_get_dma_tag() into the first argument of bus_dma_create_tag()
rather than NULL.

Minor typo fix, caught while skimming through the file.

Prepare for future integration between CAM and newbus. xpt_bus_register
now takes a device_t to be the parent of the bus that is being created.
Most SIMs have been updated with a reasonable argument, but a few exceptions
just pass NULL for now. This argument isn't used yet and the newbus
integration likely won't be ready until after 7.0-RELEASE.

Remove Giant from CAM. Drivers (SIMs) now register a mutex that CAM will
use to synchornize and protect all data objects that are used for that
SIM. Drivers that are not yet MPSAFE register Giant and operate as
usual. RIght now, no drivers are MPSAFE, though a few will be changed
in the coming week as this work settles down.

The driver API has changed, so all CAM drivers will need to be recompiled.
The userland API has not changed, so tools like camcontrol do not need to
be recompiled.

o break newbus api: add a new argument of type driver_filter_t to
bus_setup_intr()

o add an int return code to all fast handlers

o retire INTR_FAST/IH_FAST

For more info: http://docs.freebsd.org/cgi/getmsg.cgi?fetch=465712+0+current/freebsd-current

Approved by: re (implicit?)

o break newbus api: add a new argument of type driver_filter_t to
bus_setup_intr()

o add an int return code to all fast handlers

o retire INTR_FAST/IH_FAST

For more info: http://docs.freebsd.org/cgi/getmsg.cgi?fetch=465712+0+current/freebsd-current

Reviewed by: many
Approved by: re@

Add MODULE_DEPENDS for cam, pci, mca, eisa and isa where needed.

PR: 106543
MFC after: 3 days

Remove unused structures.

Reviewed by: cognet and imp
Approved by: cognet and imp

2nd and final commit that moves us to CAM_NEW_TRAN_CODE
as the default.

Reviewed by multitudes.

The first of 3 major steps to move the CAM layer forward to using
the CAM_NEW_TRAN_CODE that has been in the tree for some years now.

This first step consists solely of adding to or correcting
CAM_NEW_TRAN_CODE pieces in the kernel source tree such
that a both a GENERIC (at least on i386) and a LINT build
with CAM_NEW_TRAN_CODE as an option will compile correctly
and run (at least with some the h/w I have).

After a short settle time, the other pieces (making
CAM_NEW_TRAN_CODE the default and updating libcam
and camcontrol) will be brought in.

This will be an incompatible change in that the size of structures
related to XPT_PATH_INQ and XPT_{GET,SET}_TRAN_SETTINGS change
in both size and content. However, basic system operation and
basic system utilities work well enough with this change.

Reviewed by: freebsd-scsi and specific stakeholders

Add explicit dependency on cam. This is necessary when both aha.ko and
cam.ko are modules so that aha.ko's undefined symbols can be satisfied by
cam.ko.

Sumitted by: nork
Reviewed by: scottl

Remove bus_{mem,p}io.h and related code for a micro-optimization on i386
and amd64. The optimization is a trivial on recent machines.

Reviewed by: -arch (imp, marcel, dfr)

Simplify aha resource management, and fix a few bugs in unwinding
error cases.

Use the standard FreeBSD license

nits

Final attempt to make aha 1542A working. If not, oh well, I don't
have the card and no way to reproduce problems. We do this by
applying the workaround to firmware revsion 0.

PR: 14334

Start each of the license/copyright comments with /*-, minor shuffle of lines

Various whitespace nits.

Noticed by: njl

Don't use aha after calling aha_free.

#include <isa/isavar.h> instead of <i386/isa/isa_dma.h>

Convert callers to the new bus_alloc_resource_any(9) API.

Submitted by: Mark Santcroos <marks@ripe.net>
Reviewed by: imp, dfr, bde

Save the device so we can do a device_printf.
Use this in preference to aha_name.
Remove aha_name function and #define it to device_get_unitname()
Minor indentation tweaks resulting therefrom

Minor diff reduction with p4

Move 'guessing' code from the probe into the identify routine where it more
properly belongs.

Tweak the front end driver file list
Also some minor whitespace nits.

Add my copyright to aha_isa.c to make it crystal clear its license

Let bus space manage softc.

Minor tweaks to make it behave better:
o When we're resetting the board, make sure that we error out the pending
CCBs first. Otherwise the aha_cmd won't accept further commands, such
as those that are used to reset the card (AOP_INITIALIZE_MBOX). This
appears to cause a cascade failure where no more commands are possible
to the card.
o Reduce from 10s down to 1s the amount of time we're willing to tolerate
the card being awol. This helps the above case.
o Add some error checking to two commands issued in the probe.

I have a dim memory of gibbs@ trying to tell me about this problem a
few years ago, so pointy hat to imp@ for sitting on it so long.

Minor comment smithing

Make this driver a little more style(9) compliant

Use __FBSDID().
Also some minor style cleanups.

Mega busdma API commit.

Add two new arguments to bus_dma_tag_create(): lockfunc and lockfuncarg.
Lockfunc allows a driver to provide a function for managing its locking
semantics while using busdma. At the moment, this is used for the
asynchronous busdma_swi and callback mechanism. Two lockfunc implementations
are provided: busdma_lock_mutex() performs standard mutex operations on the
mutex that is specified from lockfuncarg. dftl_lock() is a panic
implementation and is defaulted to when NULL, NULL are passed to
bus_dma_tag_create(). The only time that NULL, NULL should ever be used is
when the driver ensures that bus_dmamap_load() will not be deferred.
Drivers that do not provide their own locking can pass
busdma_lock_mutex,&Giant args in order to preserve the former behaviour.

sparc64 and powerpc do not provide real busdma_swi functions, so this is
largely a noop on those platforms. The busdma_swi on is64 is not properly
locked yet, so warnings will be emitted on this platform when busdma
callback deferrals happen.

If anyone gets panics or warnings from dflt_lock() being called, please
let me know right away.

Reviewed by: tmm, gibbs

Bring back bus_dmasync_op_t. It is now a typedef to an int, though the
BUS_DMASYNC_ definitions remain as before. The does not change the ABI,
and reverts the API to be a bit more compatible and flexible. This has
survived a full 'make universe'.

Approved by: re (bmah)

I deserve a big pointy hat for having missed all those references
to bus_dmasync_op_t in my last commit.

Clean up argument comments for bus_dma_tag_create() calls.

Consolidate MIN/MAX macros into one place (param.h).

Submitted by: Hiten Pandya <hiten@unixdaemons.com>

Correct typos, mostly s/ a / an / where appropriate. Some whitespace cleanup,
especially in troff files.

Fix build on alpha. It used to case an int to a pointer in order to do
pointer arithmetic. Detour via uintptr_t, since we're generating an
offset not dereferencing them.

Change BUS_SPACE_UNRESTRICTED (~0ul) to plain ~0 when used in the
'int nsegments' argument to bus_dma_tag_create(). ~0ul does not fit in
an int on machines with 64 bit longs.

Fix new gcc-3.1 warnings. I think this gets GENERIC compiling cleanly
again.

Turn on interrupt-entropy harvesting for all/any mass storage devices
I could find. I have no doubt missed a couple.

Interrupt entropy harvesting is still conditional on the
kern.random.sys.harvest_interrupt sysctl.

Remove NAHA, NAHATOT and aha_softcs and related code. It was unused
except for setting it. Also remove count from aha and replace it with
optional.

Also add commented out pccard lines for all the old card drivers.
They have to be commented out until they are converted because it
causes problems in NEWCARD.

Remove unnecessary includes found by phk's script a long time ago.

Make a few more mallocs use M_ZERO.

Submitted by: josh@zipperup.org
Submitted by: Robert Drehmel <robd@gmx.net>
Approved by: imp

Remove unneeded #include <machine/clock.h>

Fix typo (accessable --> accessible).

PR: 18588
Submitted by: Anatoly Vorobey <mellon@pobox.com>
Reviewed by: asmodai

Back out the previous change to the queue(3) interface.
It was not discussed and should probably not happen.

Requested by: msmith and others

Change the way that the queue(3) structures are declared; don't assume that
the type argument to *_HEAD and *_ENTRY is a struct.

Suggested by: phk
Reviewed by: phk
Approved by: mdodd

Remove unneeded <sys/buf.h> includes.

Due to some interesting cpp tricks in lockmgr, the LINT kernel shrinks
by 924 bytes.

More pnpids.

Fix plug and play support:
o Cut out the probed stuff. We no longer need it since newbus implicitly
checks for this (likely bt can be changed as well in this way).
o Add preliminary support for unload. Untested because aha doesn't yet
support identify and there are some interactions with PnP that I've
not yet worked out.

With this I can boot the AHA-1542CP FW F.0. All the aha resources
appear to be picked up via pnp now.

MIN unused, so kill it

* Add struct resource_list* argument to resource_list_alloc and
resource_list_release. This removes the dependancy on the
layout of ivars.

* Move set_resource, get_resource and delete_resource from
isa_if.m to bus_if.m.

* Simplify driver code by providing wrappers to those methods:

bus_set_resource(dev, type, rid, start, count);
bus_get_resource(dev, type, rid, startp, countp);
bus_get_resource_start(dev, type, rid);
bus_get_resource_count(dev, type, rid);
bus_delete_resource(dev, type, rid);

* Delete isa_get_rsrc and use bus_get_resource_start instead.

* Fix a stupid typo in isa_alloc_resource reported by Takahashi
Yoshihiro <nyan@FreeBSD.org>.

* Print a diagnostic message if we can't assign resources to a PnP
device.

* Change device_print_prettyname() so that it doesn't print
"(no driver assigned)-1" for anonymous devices.

Kill unused includes

Use the PRVERB() macro instead of 'if (bootverbose) printf()...

Add debuging for all failure cases in aha_cmd().

Always check DATAIN_REG_READY even if we're ready to exit the
read loop.

The last fix allows the AHA-1640 MCA SCSI adapter to operate
correctly and likely fixes problems on a few other oddball
adapters. (I suspect my Dell Drive Array in 1540 mode needs this.)

Reviewed by: Warner Losh

Add support for the AHA-1640 SCSI adapter.

The changes to sys/dev/aha/aha.c that allow these
cards to function properly will be here in a day
or 2.

Use rmand_get_bus{tag,handle} rather than hard wiring things to
I386_BUS_SPACE_IO. Compiles now on the Alpha, but likely will not
work due to bus space address <-> virtual address mapping bogons that
work for i386 but not alpha.

Newbusification of aha. dfr sent me the first cut, and I made it
work. Be more verbose when one cannot allocate IRQ, et al since this
is a common configuration problem. The cards have the IRQ soft wired
into their BIOS and do not try to do collision detection. This can
cause problems when this IRQ is the same as another card/device.

The PNP hasn't been tested. My PNP board is in a deployed system.
I'll sneak in testing of it sometime later. I've been able to mount
the 3.3R cdrom that arrived today and access files off it.

Submitted by: dfr

Remove vestiges of old pnp code.

Missed this use of the old pnp includes. I don't think this is meant to
be here though.

Stub out the old pnp code till these are converted so GENERIC builds again.

$Id$ -> $FreeBSD$

Properly set the alignment argument to bus_dma_tag_create(). If we
don't care about the alignment, set it to 1, meaning single byte alignment.

The 1542 cards do not allow adapter commands to be queued while mailbox
commands are outstanding. You'd think they'd just clear the IDLE bit,
but alas, no. Delay until all pending mailbox commands have completed
in aha_cmd to work around this.

Report sync rates correctly on Fast Adaptec cards. Clones may still be
reported incorrectly since there is no documenation on how they report
extended sync values.

Clean up some unused fields in the aha softc.

Put a small delay in before sending the inquire_setup command. The
docs don't seem to shed light on why this is needed, but reports from
the field indicate this helps prevent problems in this area. Ken's
changes seem to have exposed this bug, rather than caused it, as far
as I can tell.

Thanks to Jack O'Neill for tracking this down.

Submitted by: jack@germanium.xtalwind.net

Very strong 3.2 merge candidate.

Add another kludge to the card identification probe. It appears that
some aha 1542B cards will return 0x7f for the unimplemented GEOMETRY
register.

This is a good 3.2 candidate.

PR: 11469
Submitted by: Thomas David Rivers

Use haveseen_ioport() which is now connected up to the resource manager.

ahatimeout is static..

Add a number of interrelated CAM feature enhancements and bug fixes.

NOTE: These changes will require recompilation of any userland
applications, like cdrecord, xmcd, etc., that use the CAM passthrough
interface. A make world is recommended.

camcontrol.[c8]:
- We now support two new commands, "tags" and "negotiate".

- The tags commands allows users to view the number of tagged
openings for a device as well as a number of other related
parameters, and it allows users to set tagged openings for
a device.

- The negotiate command allows users to enable and disable
disconnection and tagged queueing, set sync rates, offsets
and bus width. Note that not all of those features are
available for all controllers. Only the adv, ahc, and ncr
drivers fully support all of the features at this point.
Some cards do not allow the setting of sync rates, offsets and
the like, and some of the drivers don't have any facilities to
do so. Some drivers, like the adw driver, only support enabling
or disabling sync negotiation, but do not support setting sync
rates.

- new description in the camcontrol man page of how to format a disk
- cleanup of the camcontrol inquiry command
- add support in the 'devlist' command for skipping unconfigured devices if
-v was not specified on the command line.
- make use of the new base_transfer_speed in the path inquiry CCB.
- fix CCB bzero cases

cam_xpt.c, cam_sim.[ch], cam_ccb.h:

- new flags on many CCB function codes to designate whether they're
non-immediate, use a user-supplied CCB, and can only be passed from
userland programs via the xpt device. Use these flags in the transport
layer and pass driver to categorize CCBs.

- new flag in the transport layer device matching code for device nodes
that indicates whether a device is unconfigured

- bump the CAM version from 0x10 to 0x11

- Change the CAM ioctls to use the version as their group code, so we can
force users to recompile code even when the CCB size doesn't change.

- add + fill in a new value in the path inquiry CCB, base_transfer_speed.
Remove a corresponding field from the cam_sim structure, and add code to
every SIM to set this field to the proper value.

- Fix the set transfer settings code in the transport layer.

scsi_cd.c:

- make some variables volatile instead of just casting them in various
places
- fix a race condition in the changer code
- attach unless we get a "logical unit not supported" error. This should
fix all of the cases where people have devices that return weird errors
when they don't have media in the drive.

scsi_da.c:

- attach unless we get a "logical unit not supported" error

scsi_pass.c:

- for immediate CCBs, just malloc a CCB to send the user request in. This
gets rid of the 'held' count problem in camcontrol tags.

scsi_pass.h:

- change the CAM ioctls to use the CAM version as their group code.

adv driver:

- Allow changing the sync rate and offset separately.

adw driver

- Allow changing the sync rate and offset separately.

aha driver:

- Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs.

ahc driver:

- Allow setting offset and sync rate separately

bt driver:

- Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs.

NCR driver:

- Fix the ultra/ultra 2 negotiation bug
- allow setting both the sync rate and offset separately

Other HBA drivers:
- Put code in to set the base_transfer_speed field for
XPT_GET_TRAN_SETTINGS CCBs.

Reviewed by: gibbs, mjacob (isp), imp (aha)

Bring the 'new-bus' to the i386. This extensively changes the way the
i386 platform boots, it is no longer ISA-centric, and is fully dynamic.
Most old drivers compile and run without modification via 'compatability
shims' to enable a smoother transition. eisa, isapnp and pccard* are
not yet using the new resource manager. Once fully converted, all drivers
will be loadable, including PCI and ISA.

(Some other changes appear to have snuck in, including a port of Soren's
ATA driver to the Alpha. Soren, back this out if you need to.)

This is a checkpoint of work-in-progress, but is quite functional.

The bulk of the work was done over the last few years by Doug Rabson and
Garrett Wollman.

Approved by: core

Staticize.

Better logging when the Geometry register probe fails.

o enable plug and play support for the aha driver. Given the cumbersome
pnp system in freebsd, I'm not sure how useful this will be, but my
1542CP seems to work well in plug and play mode and does seem to
probe correctly at all the oddball addresses/irq/drqs that I tried.
[[
I was unable to get /kernel.conf or /kernel.config to read in, so
I wasn't able to verify that this method of userconfig works. that's
one thing that makes pnp so hard to use in the current scheme.
Pointers to the right new way of doing this accepted.
]]
o Add some kludges to maybe bring support for 1540A/1542A into the
driver. Since I have no 154xA cards, and the only person I know
that has them hasn't given me feedback, I'm making this commit
blind.
o Honor unit numbers that are in the config file now. This allows one
to hard wire the unit numbers (and have high unit numbers for plug
and pray devices, which can't seem to be hardwired) and have the
cards not migrate from aha1 -> aha0 should aha0 go on the fritz. I
didn't verify that hard wired scsi busses would work, but did verify
that hard wired aha addresses did work to a limited extent. Both
aha0 and aha1 must be hardwired, or when the card that was in aha0
goes away, the probe for aha0 might pick up the card that otherwise
would have been aha1.

Invalid is spelled with the v *BEFORE* the a.

Don't calculate the residual twice.

o Convert length of hccb to integer from the array. Eivind Eklund
pointed this out, but I've not seen a manifestation of this.
o Check against 0x00 as well as 0xff for geometry register, as some clone
cards don't return 0xff. Vadim Mikhailov pointed this out in PR
8743 for his Dell SCSI Array controller working in AHA-1540
emulation mode. Note that this test is likely to go away in the
future in favor of a better one Justin has recommended.

Pull in fixes similar to those recently committed to the bt driver:

- For transactions of 0 length, us a non-residual checking CCB type.

- Preserve command status if our interrupt handler completes a command
while we are polling for completion in aha_cmd.

Examine all occurrences of sprintf(), strcat(), and str[n]cpy()
for possible buffer overflow problems. Replaced most sprintf()'s
with snprintf(); for others cases, added terminating NUL bytes where
appropriate, replaced constants like "16" with sizeof(), etc.

These changes include several bug fixes, but most changes are for
maintainability's sake. Any instance where it wasn't "immediately
obvious" that a buffer overflow could not occur was made safer.

Reviewed by: Bruce Evans <bde@zeta.org.au>
Reviewed by: Matthew Dillon <dillon@apollo.backplane.com>
Reviewed by: Mike Spengler <mks@networkcs.com>

o Patch, slightly altered, from Graham Menhennitt for older aha 1542B
cards. My machine didn't tickle this problem, so I can't tell if
it fixed anything or not. However, it didn't break my already working
1542-{B,C,CF,CP} probes.
o Minor comment correction
o Minor correction in printf

Noop Commit - Change was in last revision.

If the next available mailbox is not in the free state, print a warning
and handle the condition as a temporary resource shortage. The condition
should never happen, but we shouldn't panic since recovery is possible.

Fix probes when a port address is specified.

Fix probes when a port address is specified.

Convert from BT'isms to AHA'isms

Don't fail the probe if the illegal command bit is set in the status
register. If the BusLogic MultiMaster probe preceeded us, it may well
have determined that the card we are attaching to was not a true MultiMaster
by sending us a command that fails on the 154X.

Reset the adapter before doing the inquiry. This provides extra sanity
and will also clear the illegal command status bit that my be left over
from the MultiMaster probe.

-Wunused cleanup.

Submitted by: Poul-Henning Kamp <phk@freebsd.org>

Reduce the number of ccbs to 16, per the aha documentation. Move
bogus comment to proper place.

This should fix the hangs people are seeing under very heavy load, at
least it does for me. Please let me know if you continue to have
problems.

Fix breakage introduced by last patch. bde has added CC_QUIET flag to
hasseen_isadev so this will be less noisy when conflicts do exist.
Also eliminate redundant warnings about conflicts.

Requested by: bde
Reviewed by: gibbs

Fix conficts in probe:
o For bt and aha only probe the one I/O range if a specific I/O is specified
in the config file.
o Don't even try to probe I/O ranges that have been seen already.
o If we conflict with an IRQ or DRQ, then fail the probe.

Requested by: bde, gibbs
Approved by: jkh

Hook up ahapoll so that dumps, synchronize cache commands or any other
command that comes in through xpt_polled_action works correctly.

GC unused stuff.

Misc cleanup and probe rework:

o Use the board id command to find out what kind of board
we're talking to. If we're talking to a board that is has
an ID that is shared between boards supported by the aha
driver and the bt driver, then use the bt's geometry
register to weed out the bt cards. Otherwise assume that we
support this card.
o Remove bt esetup command sending to the card. It seems to
wedge too many cards.
o Revert to doing a soft reset after an invalid command. This
change didn't fix anything, so I'm backing it out. The
whole issue of card resetting needs to be revisisted at some
point so that we can do it properly on all hardware.
o GC unused stuff in some places.

Perform a hard reset on cards when the command fails. This should help
those people that have cards that become wedged when a bogus command is
issued that are too wedged to have a soft reset help.

Several minor cleanups, inspired by bug reports and the old driver:

o Unlock mailbox interface if we have a new card. Before only newer cards
(B or newer) that had the BIOS disabled would probe. Cards with the
BIOS enabled would fail to probe in the mailbox initialization code.
o Increase the number of ccbs and sg to 17 from 16 to support 64k I/O
on a non page aligned boundary. Ideas for dynamic determination of this
value welcomed, as more of these are better.
o Took credit for this driver, even though I derived it from Justin's code.
Made sure that Justin's copyright from bt.c was preserved, along with
his name, since the error handling code is nearly identical. Add my
own, identical copyright. Point people to aha_isa.c.

Cards tested: 1542C and 1542CF. The B and CP might work now as well,
but logistical problems precluded me from testing them this evening
(if you have jumper settings for the B card, please send me private
mail).

Removed unused include of "ioconf.h" again. The CAM changes made ioconf.h
empty but regressed to including it here.

Use %p with (void *) casts to print pointers with printf.

Fix printf format bugs.

Adaptec 154X SCSI-Host Adapter driver for CAM.

Submitted by: Warner Losh <imp@village.org>