Captured a panic on an older kernel where an application issuing
commands via sg was sending requests that lacked a request_buffer, thus
the buffer pointer used in aac_internal_transer was NULL. The
application was fixed closing the issue, but felt it was advised to
immunize the driver against the eventuality.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

This set of fixes improve error handling stability of the driver. A popular
manifestation of the problems is an NULL pointer reference in the interrupt
handler when referencing portions of the scsi command context, or in the
scsi_done handling when an offlined device is referenced.

The aacraid driver currently does not get notification of orphaned command
completions due to devices going offline. The driver also fails to handle the
commands that are finished by the error handler, and thus can complete again
later at the hands of the adapter causing situations of completion of an
invalid scsi command context. Test Unit Ready calls abort assuming that the
abort was successful, but are not, and thus when the interrupt from the adapter
occurs, they reference invalid command contexts. We add in a TIMED_OUT flag to
inform the aacraid FIB context that the interrupt service should merely release
the driver resources and not complete the command up. We take advantage of this
with the abort handler as well for select abortable commands. And we detect and
react if a command that can not be aborted is currently still outstanding to
the controller when reissued by the retry mechanism.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

The raw srb ioctl is supposed to be able to take packets with 32 and 64 bit
virtual address SG elements, it did not handle the frames with 64 bit SG
elements well when communicating with 64 bit DMA capable adapters, and it did
not handle the 32 bit limited DMA adapters at all.  The enclosed patch now
handles all four quadrants (32 bit / 64 bit SG elements in SRB requests + 32
bit or 64 bit DMA capable adapters)

This fix is required before Java based management applications in a 64 bit user
space can submit raw srb requests to the array physical components via the
ioctl mechanism, the allocated user memory pool on 64 bit machines under this
environment forced the management software's hands to submit 64 bit user space
virtual address SG elements in via the ioctl.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

This little patch removes the ',cid)' container identification argument
from some of the functions. The argument is used in some cases as merely
a debug helper and thus not used, and in others, the value can be
quickly acquired from the scsi command in their single solitary use in
the procedure rather than wasting resources on passing the argument in
from above.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

Outstanding ioctl calls still have some problems with aborting cleanly
in the face of a reset iop recovery action should the adapter ever enter
into a Firmware Assert (BlinkLED) condition. The enclosed patch resolves
some uncovered flawed handling.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

The aac_probe_container call blocks. This is an issue because it is called on
occasion in the context of the queuecommand handler. Once in a blue moon this
has resulted in a kernel panic sleeping during interrupt; or problems with some
embedded system versions of the kernel that depend on queuecommand to not
block. This ugly patch rewrites the aac_probe_container call into a new routine
_aac_probe_container that is an asynchronous state machine to complete the
series of operations. The legacy blocking aac_probe_container call used in
other areas of the driver (during initialization scanning for all targets and
in the separate hot-add/remove [aacraid] thread) merely issues
_aac_probe_container and then simple spins calling schedule() waiting for
completion.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

This patch is to resolve a namespace issue that will result from a patch
expected in the future that adds a new interface; rationalized as
correcting a long term issue where hw_fib, instead of hw_fib_va, refers
to the virtual address space and hw_fib_pa refers to the physical
address space. A small fragment of this patch also cleans up an unused
variable that was close to the patch fragments.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

This patch updates the adapter restart function to deal with some
adapters that have specific IOP reset needs. Since the code for
restarting the adapter was in two places, changed over to utilizing a
platform function in one place.
Signed-off-by: NMark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

After Al Viro (finally) succeeded in removing the sched.h #include in module.h
recently, it makes sense again to remove other superfluous sched.h includes.
There are quite a lot of files which include it but don't actually need
anything defined in there.  Presumably these includes were once needed for
macros that used to live in sched.h, but moved to other header files in the
course of cleaning it up.

To ease the pain, this time I did not fiddle with any header files and only
removed #includes from .c-files, which tend to cause less trouble.

Compile tested against 2.6.20-rc2 and 2.6.20-rc2-mm2 (with offsets) on alpha,
arm, i386, ia64, mips, powerpc, and x86_64 with allnoconfig, defconfig,
allmodconfig, and allyesconfig as well as a few randconfigs on x86_64 and all
configs in arch/arm/configs on arm.  I also checked that no new warnings were
introduced by the patch (actually, some warnings are removed that were emitted
by unnecessarily included header files).
Signed-off-by: NTim Schmielau <tim@physik3.uni-rostock.de>
Acked-by: NRussell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Many struct file_operations in the kernel can be "const".  Marking them const
moves these to the .rodata section, which avoids false sharing with potential
dirty data.  In addition it'll catch accidental writes at compile time to
these shared resources.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Received from Mark Salyzyn,

Take the expose_physicals flag and allow the user to select default (physicals
available via /dev/sg), exposed (physicals available via /dev/sd for
experimental reasons) and hidden (physicals blocked from all access). This
expands the functionality of the previous expose_physicals insmod parameter
which was added to support some experimental configurations.

Signed-off-by Mark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

Replace all if/else packet formations with platform function calls. This is in
recognition of the proliferation of read and write packet types, and in the
need to migrate to up-and-coming packets for new products.

Signed-off-by Mark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

Add in the NEMER/ARK physical register mapping, represented in up and coming
products currently under test at Adaptec.

Signed-off-by Mark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn,

Replace all if/else communication transports with a platform function call.
This is in recognition of the need to migrate to up-and-coming transports.
Currently the Linux driver does not support two available communication
transports provided by our products, these will be added in future patches, and
will expand the platform function set.

Signed-off-by Mark Haverkamp <markh@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Update drivers/scsi/aacraid/linit.c and Documentation/scsi/aacraid.txt
file with the current list of
adapters supported by the aacraid driver. Deprecated a few adapters that
never shipped, corrected a
few and added new adapters that matched the family code support. No
functional changes to the driver.
No side effects.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Run this:

	#!/bin/sh
	for f in $(grep -Erl "\([^\)]*\) *k[cmz]alloc" *) ; do
	  echo "De-casting $f..."
	  perl -pi -e "s/ ?= ?\([^\)]*\) *(k[cmz]alloc) *\(/ = \1\(/" $f
	done

And then go through and reinstate those cases where code is casting pointers
to non-pointers.

And then drop a few hunks which conflicted with outstanding work.

Cc: Russell King <rmk@arm.linux.org.uk>, Ian Molton <spyro@f2s.com>
Cc: Mikael Starvik <starvik@axis.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Roman Zippel <zippel@linux-m68k.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Greg KH <greg@kroah.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Karsten Keil <kkeil@suse.de>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: James Bottomley <James.Bottomley@steeleye.com>
Cc: Ian Kent <raven@themaw.net>
Cc: Steven French <sfrench@us.ibm.com>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Neil Brown <neilb@cse.unsw.edu.au>
Cc: Jaroslav Kysela <perex@suse.cz>
Cc: Takashi Iwai <tiwai@suse.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Received from Mark Salyzyn:

Version patch, update to reflect a rough estimate of the Adaptec build
(2423) that coincides with the sources on kernel.org.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn:

Add code to abort outstanding management ioctl fibs when the blinkLED recovery
is performed. This code is 'clunky' and does not have any real feedback in that
the reset could progress before the user application has gotten it's
notification of command completion. We put a schedule() call to delay just the
right amount for most cases, because we tried a spin and still managed to find
cases where we would spin forever waiting for the management application to
acknowledge the impending doom surrounding the cause of the BlinkLED. Will
cause an oops in the context of the management application if we proceed too
quickly. I view this as the lesser of many evils since currently if there are
outstanding management ioctls during a need to reset/recover the adapter, the
management application just locks up and waits forever. The best practices fix
for this problem not going to be simple or easy (at least the fixes I imagine
today); and we found a balance between the needs of the driver to proceed, and
the applications that locked or confused that would hold back the driver. I
just do not like the idea of a kernel oops in an application to deal with low
priority, sluggish or misbehaving applications.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn:

Blinkled at startup is useful for catching Adapters in a lot of pain, in a
BlinkLED assert, quickly; rather than waiting several minutes for commands to
timeout.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Maintain a per-CPU global "struct pt_regs *" variable which can be used instead
of passing regs around manually through all ~1800 interrupt handlers in the
Linux kernel.

The regs pointer is used in few places, but it potentially costs both stack
space and code to pass it around.  On the FRV arch, removing the regs parameter
from all the genirq function results in a 20% speed up of the IRQ exit path
(ie: from leaving timer_interrupt() to leaving do_IRQ()).

Where appropriate, an arch may override the generic storage facility and do
something different with the variable.  On FRV, for instance, the address is
maintained in GR28 at all times inside the kernel as part of general exception
handling.

Having looked over the code, it appears that the parameter may be handed down
through up to twenty or so layers of functions.  Consider a USB character
device attached to a USB hub, attached to a USB controller that posts its
interrupts through a cascaded auxiliary interrupt controller.  A character
device driver may want to pass regs to the sysrq handler through the input
layer which adds another few layers of parameter passing.

I've build this code with allyesconfig for x86_64 and i386.  I've runtested the
main part of the code on FRV and i386, though I can't test most of the drivers.
I've also done partial conversion for powerpc and MIPS - these at least compile
with minimal configurations.

This will affect all archs.  Mostly the changes should be relatively easy.
Take do_IRQ(), store the regs pointer at the beginning, saving the old one:

	struct pt_regs *old_regs = set_irq_regs(regs);

And put the old one back at the end:

	set_irq_regs(old_regs);

Don't pass regs through to generic_handle_irq() or __do_IRQ().

In timer_interrupt(), this sort of change will be necessary:

	-	update_process_times(user_mode(regs));
	-	profile_tick(CPU_PROFILING, regs);
	+	update_process_times(user_mode(get_irq_regs()));
	+	profile_tick(CPU_PROFILING);

I'd like to move update_process_times()'s use of get_irq_regs() into itself,
except that i386, alone of the archs, uses something other than user_mode().

Some notes on the interrupt handling in the drivers:

 (*) input_dev() is now gone entirely.  The regs pointer is no longer stored in
     the input_dev struct.

 (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking.  It does
     something different depending on whether it's been supplied with a regs
     pointer or not.

 (*) Various IRQ handler function pointers have been moved to type
     irq_handler_t.
Signed-Off-By: NDavid Howells <dhowells@redhat.com>
(cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Received from Mark Salyzyn:

Until the system is stabilized, I am suggesting the enclosed
modification to prevent the driver from tickling the panic. Once sysfs
and friends are stabilized, the patch may be backed out. We have yet to
evaluate if we really want to relinquish existing Scsi Devices in any
case, holding on to them as configuration of arrays comes and goes makes
some sense as well. As a result, we have opted to pull the lines rather
than comment them in legacy.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn:

The only real difference between the rkt and rx platform modules is the
offset of the message registers. This patch recognizes this similarity
and simplifies the driver to reduce it's code footprint and to improve
maintainability by reducing the code duplication.

Visibly, the 'rkt.c' portion of this patch looks more complicated than
it really is. View it as retaining the rkt-only specifics of the
interface.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn:

I am placing this functionality into an insmod parameter. Normally the physical
components are exported to sg, and are blocked from showing up in sd.

Note that the pass-through I/O path via the driver through the Firmware to the
physical disks is not an optimized path, the card is designed for Hardware
RAID, elevator sorting and caching. This should not be used as a means for
utilizing the aacraid based controllers as a generic scsi/SATA/SAS controller,
performance should suck by a few percentage points, any RAID meta-data on the
drives will confuse the controller about who owns the drives and there is a
high risk of destroying content in both directions. Unreliable and for
experimentation or strange controlled circumstances only.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn:

Basically cleanup, nothing here will have an affect. Adjusting some
error codes, removing superfluous definitions and code fragments.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn

If the adapter is in blinkled (Firmware Assert) when error recovery
timeout actions have been triggered, perform an adapter warm reset and
restart the initialization.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn

The enclosed patch cleans up some code fragments, adds some paranoia
(unproven causes of potential driver failures).
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn

If the adapter should be in a blinkled (Firmware Assert) state when the
driver loads, we will perform a warm restart of the Adapter Firmware to
see if we can rescue the adapter. Possible causes of a blinkled can
occur on some early release motherboard BIOSes, transitory PCI bus
problems on embedded systems or non-x86 based architectures, transitory
startup failures of early release drives or transitory hardware
failures; some of which can bite the adapter later at runtime. Future
enhancements will include recovery during runtime.

Fixed extra whitespace space issue.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn

This patch allows the FSACTL_SEND_LARGE_FIB, FSACTL_SENDFIB and
FSACTL_SEND_RAW_SRB ioctl calls into the aacraid driver to be
interruptible. Only necessary if the adapter and/or the management
software has gone into some sort of misbehavior and the system is being
rebooted, thus permitting the user management software applications to
be killed relatively cleanly. The FIB queue resource is held out of the
free queue until the adapter finally, if ever, completes the command.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: James Bottomley <James.Bottomley@steeleye.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This may seem like a DILLIGAF, but after chatting with the F/W folks,
there is no harm in dropping the page calculation as denoted in the
enclosed patch for these older adapters in this new age of 4GB+ memory
sticks. Any resource optimization within the old-old-old adapters for
systems with less than 4G of memory is of little consequence. The
existing AAC_QUIRK_31BIT flag in linit.c should look after the rest of
the legacy hardware DMA limitations.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

 - Rename the GART_IOMMU option to IOMMU to make clear it's not
   just for AMD
 - Rewrite the help text to better emphatise this fact
 - Make it an embedded option because too many people get it wrong.

To my astonishment I discovered the aacraid driver tests this
symbol directly. This looks quite broken to me - it's an internal
implementation detail of the PCI DMA API. Can the maintainer
please clarify what this test was intended to do?

Cc: linux-scsi@vger.kernel.org
Cc: alan@redhat.com
Cc: markh@osdl.org
Signed-off-by: NAndi Kleen <ak@suse.de>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NEric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Cc: Bartlomiej Zolnierkiewicz <B.Zolnierkiewicz@elka.pw.edu.pl>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: James Bottomley <James.Bottomley@steeleye.com>
Acked-by: N"Salyzyn, Mark" <mark_salyzyn@adaptec.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Received from Mark Salyzyn

Spelling correction, orphaned comment removal & update branch name.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received From Mark Salyzyn

Some of the cards product names changed.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received from Mark Salyzyn

clear_user return is 0 for success, the code fragment is written to
assume that it is the count of the number of bytes zero'd.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Use ARRAY_SIZE macro instead of sizeof(x)/sizeof(x[0]) and remove
duplicates of the macro.
Signed-off-by: NTobias Klauser <tklauser@nuerscht.ch>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Various scsi drivers use scsi_cmnd.buffer and scsi_cmnd.bufflen in their
queuecommand functions.  Those fields are internal storage for the
midlayer only and are used to restore the original payload after
request_buffer and request_bufflen have been overwritten for EH.  Using
the buffer and bufflen fields means they do very broken things in error
handling.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received From Mark Salyzyn

The queue tracking is just not being used, not even for debugging. Information
about outstanding commands can be acquired from the scsi structures.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Received From Mark Salyzyn

A race condition existed that could result in a lost completion of a
command to the ppc based cards.
Signed-off-by: NMark Haverkamp <markh@osdl.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>