For crying out loud, they have devices which do not like port resets.
So, do what usb-storage does and try both bulk and port resets.
We start with a port reset (which usb-storage does at the end of transport),
then do a Bulk reset, then a port reset again. This seems to work for me.

The code is getting dirtier and dirtier here, but I swear that I'll
do something about it (see those two new XXX). Honest.
Signed-off-by: NPete Zaitcev <zaitcev@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If SCSI commands are submitted while other commands are still processed,
the dispatch loop turns, and we stop the work_timer. Then, if URB fails
to complete, ub hangs until the device is unplugged.

This does not happen often, becase we only allow one SCSI command per
block device, but does happen (on multi-LUN devices, for example).

The fix is to stop timer only when we actually going to change the state.

The nicest code would be to have the timer stopped in URB callback, but
this is impossible, because it can be called from inside a timer, through
the urb_unlink. Then we get BUG in timer.c:cascade(). So, we do it a
little dirtier.
Signed-off-by: NPete Zaitcev <zaitcev@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The blk_cleanup_queue does not necesserily destroy the queue. When we
destroy the corresponding ub_dev, it may leave the queue spinlock pointer
dangling.

This patch moves spinlocks from ub_dev to static memory. The locking
scheme is not changed. These spinlocks are still separate from the ub_lock.
Signed-off-by: NPete Zaitcev <zaitcev@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Cc: David Brownell <david-b@pacbell.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This moves the previously widely-used ehci-pci.c BIOS handoff
code into the pci-quirks.c file, replacing the less widely used
"early handoff" version that seems to cause problems lately.

One notable change:  the "early handoff" version always enabled
an SMI IRQ ... and did so even if the pre-Linux code said it was
not using EHCI (and not expecting EHCI SMIs).  Looks like a goof
in a workaround for some unknown BIOS version.

This merged version only forcibly enables those IRQs when pre-Linux
code says it's using EHCI.  And now it always forces them off "just
in case".
Signed-off-by: NDavid Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I get storms of warnings from local_bh_enable().  Better-tested patches,
please.

Cc: Ingo Molnar <mingo@elte.hu>
Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Modern versions of gcc do not like case statements at the end of a block
statement: you need at least an empty statement.  Using just a "break;"
is preferred for visual style.
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

A critical thing the ServeRAID driver MUST do is hide the physical DASDI
devices from the OS. It does this by intercepting the INQUIRY commands.

In recent 2.6.15 testing, I discovered this to be failing.

The cause was the driver assuming that the INQUIRY response data was in a
simple single buffer, when it was actually a 1 element scatter gather list.

This patch makes ips always look at the correct data when examining an
INQUIRY response.
Signed-off-by: NJack Hammer <jack_hammer@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This adds a sanity check in the interrupt routine
insures incoming message frames are a valid
message frames.

The code for setting 0xdeadbeaf in the freed message
frames, apparently was already submitted by Christoph
in previous patch submission.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch inhibits sending spi negotiation parameters
for non-configured devices from the slave_destroy function.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The ioc->alt_ioc->alt_ioc pointer is not getting cleared
during driver unload time.   This dangling pointer
can result in panic in certain circumstances, such
as error recovery, or firmware download in flashless
environments. This only impacts dual functions controllers,
such as 1030. Please apply.

This patch also includes a small cosmetic name change
for mpt_spi_log_info.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

When people use the userspace scanning facilities on SAS hardware the
LLDD gets bogus slave_alloc calls.  Just fail those gracefully instead
of printing a warning in mptsas and another one in the midlayer.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Adding verbose message returned from firmware
when a task mangment request fails.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

On Mon, Jan 16, 2006 at 06:53:24PM -0700, Moore, Eric wrote:
> Adding MSI support, and command line for enabling
> it.  By default, the command line option has MSI disabled.

mpt_msi_enable is initialized to 0 implicitly, no need to do that.  Also
replace if (mpt_msi_enable == 1) tests with just if (mpt_msi_enable).

Updated patch below:
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

A customer request to send raid asyn actions
from firmware to the event syslog.  This shows
when raid volumes go degraded, or complete resync,
or volumes created/deleted, etc.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Increasing the reply frame size by 16 bytes, to
be in sync with the other fusion drivers.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

On Mon, Jan 16, 2006 at 06:53:13PM -0700, Moore, Eric wrote:
> The task managment request timeout in the eh threads was set
> for U320 timing, which is between 2-5 seconds.
> This is too small for FC and SAS.
> According to the firmware engineers, Fibre needs to be 40 seconds
> and SAS needs to be 10 seconds.

The timeout selection should probably be done in a little helper instead
of duplicated in a few places.  Updated patch below.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Increase the port enable timeout only for SAS from 30 to 300 seconds.
A customer request for the handling large topologies.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch is for spi.  This issues bus reset when driver
loads. Handling cases when initator has negotiated for packetized,
and target negotiated for non-packetized; effectly this bus reset
is getting both target and initiator on the same sheet of music.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Fix the timer handling in aic79xx to use the SCSI-ML provided handling
instead of implementing our own.
It also fixes a deadlock in the command recovery code.
Signed-off-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch updates the documentation for aic7xxx and aic79xx with fixes
from the adaptec driver.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch introduces the SLOWCRC handling for certain buggy chipsets.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch updates the aic79xx sequencer with latest fixes from adaptec.
The sequencer code now corresponds with adaptec version 2.0.15.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patchset updates aicasm code with the latest fixes from adaptec.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The prior fusion patches moved an invocation of a function,
mptscsih_TMHandler(), static to mptscsih.c into mptsas.c

Make the function unstatic, move the header to mptscsih.h and export it.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This fix's problems with recent fc submission regarding
i/o being redirected to the wrong target.
Signed-off-by: NMichael Reed <mdr@sgi.com>
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This moves code intented for SAS from
the generic mptscsih module over to the
mptsas module.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The issuing of the target reset
used in device hot removal case so the
firmware queue is flushed out off outstanding
commands.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

RAID event support.

This will hot add and remove raid volumes
when managment application creates and
deletes the volumes.  The driver is basically
responding to firmware asyn events, and reporting
the changes to the above layers.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch displays the port identifier on
the folder attribute; located in the middle digit.

/sys/class/sas_rphy/rphy-%x:%x:%x

The port identifier is basically the unique identifier
for each sas domain.
Signed-off-by: NEric Moore <Eric.Moore@lsil.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>