Currently, the scsi error handler will issue a START_UNIT
command if the drive indicates it needs its motor started
and the allow_restart flag is set in the scsi_device. If,
after the scsi error handler invokes a host adapter reset
due to error recovery, a device is in a unit attention
state AND also needs a START_UNIT, that device will be placed
offline. The disk array devices on an ipr RAID adapter
will do exactly this when in a dual initiator configuration.
This patch adds a single retry to the EH initiated
START_UNIT.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>

Patch modified and
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Attached is the patch I feel will address this issue. As an added
'perk' I have also added the code to detect if the controller was
previously initialized for interrupted operations by ANY operating
system should the reset_devices kernel parameter not be set and we are
dealing with a naïve kexec without the addition of this kernel
parameter. The reset handler is also improved. Related to reset
operations, but not pertinent specifically to this issue, I have also
altered the handling somewhat so that we reset the adapter if we feel
it is taking too long (three minutes) to start up.

We have not unit tested the reset_devices flag propagation to this
driver code, nor have we unit tested the check for the interrupted
operations under the conditions of a naively issued kexec. We are
submitting this modified driver to our Q/A department for integration
testing in our current programs. I would appreciate an ACK to this
patch should it resolve the issue described in this thread...
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Bump driver version.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Improve overall command performance by embedding the scatterlist
in the command block used by the adapter. This decreases
the overall number of DMAs required for a single command.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

If qc_issue fails for some reason, return a better error
to libata.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Add entry in ipr error translation table for an error
received when a device is forced into the failed state
by the user.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Add the ability to control how much error data gets logged
on a per error basis.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Fix to make sure user config accesses get re-enabled if the
PCI config write to start BIST fails.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch fixes a problem discovered on a system with some
bad SATA devices attached. If a command to a SATAPI device
times out and the device gets reset as part of error recovery,
its possible that ipr will set err_mask to indicate a device
error has occurred. If this happens, a request sense will get
issued by libata. Since scsi core thinks this scsi command is
now completed, because the device reset handler completed successfully,
scsi core will free the associated scsi command, which may
cause an oops when that request sense is completed later
by ipr. This patch ensures that any commands that get aborted
as a result of a device reset set err_mask appropriately so
that the request sense does not get sent.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

For ipr SAS adapters that support dual pathing, this
patch modifies ipr to log an error when a path fails.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The ipr driver has a sysfs attribute which can be used to
adjust the logging level of the driver for error events.
The error response data for commands can be dumped by
increasing the logging level of the ipr driver. This
currently only works for JBOD passthrough devices.
This patch enables this function for all devices, including
RAID devices, to aid in debugging problems.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Adds support for some new PCI-E ipr adapters.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Remove the auto RAID 0 array creation module parameter, since
support for this function has been removed from the firmware.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Some ipr adapters may take longer than others to come operational.
This patch makes this timeout different for different adapters,
while still preserving the module parameter which can be used
to globally override the default.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Fixed the kernel-doc comment for ibmvscsi_slave_configure.  Thanks to
Randy Dunlap for pointing this out.

Adding a slave_configure function for the driver. Now the disks can be
restarted by the scsi mid-layer when the are disconnected and reconnected.
Signed-off-by: N"Robert Jennings" <rcj@linux.vnet.ibm.com>
Signed-off-by: N"Santiago Leon" <santil@us.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The request limit calculations used previously on the client failed to
mirror the state of the server.  Additionally, when a value < 3 was provided
there could be problems setting can_queue and handling abort and reset
commands.
Signed-off-by: N"Robert Jennings" <rcj@linux.vnet.ibm.com>
Signed-off-by: NSantiago Leon <santil@us.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The Adapter build date that is to be printed on instantiation was not
displayed as a result of the supplemental adapter information structure
not being in sync with the Firmware; the driver took an early test cycle
version that had a miss-sized padded region at the head and the
structure was not re-checked at the end of qualification. The Build Date
was not a priority and is merely a cosmetic enhancement, and the wrong
location for the start of the structure member would not induce any
side-effect problems. We updated the structure to match the actual
format, and added the TSID (Tech Support Identification) value print,
should it be present, to the adapter instantiation announcements during
driver load.

This later enhancement should improve the relationship between Service
folk & Tech Support if the printed value of the TSID found it's way into
the circular file labeled G...

Neither of these values show in sysfs (yet).
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Just sweeping the floor clean in one spot. Some of these constants have
never been used in the driver or in the firmware (and thus are
meaningless). Triggered this patch because I discovered one of the
unused constants was actually incorrect and figured it was better to
clean them out than correct and update. There are no side effects at all
regarding this patch, it is purely cosmetic.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Correctly initialize the timer for FSF requests with jiffies + timeout.

Cc: Swen Schillig <swen@vnet.ibm.com>
Acked-by: NHeiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: NChristof Schmitt <christof.schmitt@de.ibm.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

On the path qla2x00_probe_one() -> probe_failed -> qla2x00_free_device(),
pci_disable_device() is executed twice, once in qla2x00_free_device()
and once in qla2x00_probe_one().

This patch removes the unnecessary call.
Signed-off-by: NBernhard Walle <bwalle@suse.de>
Acked-by: NAndrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Unsigned long is not always the same size as a pointer, namely on 32 bit
systems with 64 bit address space. Ptrdiff_t is the same size as a
pointer in all configurations. By using ptrdiff_t we can mitigate the
warning messages on these configurations. There should be no side
effects of this patch.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Add SMC and SUN products to aacraid documentation
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Acked-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

- proper prototypes for global code in aacraid.h
- aac_rx_start_adapter() can now become static
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Acked-by: N"Salyzyn, Mark" <mark_salyzyn@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

The Coverity checker spotted this dead code introduced by
commit a07f3537.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Acked-by: N"Salyzyn, Mark" <mark_salyzyn@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Acked-by: NEric Moore <Eric.Moore@lsi.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Remove the unused SCSI-related kernel config variables

	SCSI_NCR53C8XX_PROFILE_SUPPORT
	SCSI_NCR53C8XX_PROFILE
	53C700_IO_MAPPED
	AIC79XX_ENABLE_RD_STRM
	AIC7XXX_PROBE_EISA_VL
Signed-off-by: NRobert P. J. Day <rpjday@mindspring.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Remove some trailing whitespaces and some replace whitespaces with tabs.
Signed-off-by: NHenrik Kretzschmar <henne@nachtwindheim.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Acked-by: N"Salyzyn, Mark" <mark_salyzyn@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Remove the Kconfig requirement that the PCMCIA SCSI drivers be built
only as modules, and allow them to be built into the kernel.
Signed-off-by: NRobert P. J. Day <rpjday@mindspring.com>
Acked-by: NDominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Add some likely() and unlikely() compiler hints in some of the aacraid
hardware interface layers. There should be no operational side effects
resulting from this patch and the changes should be mostly benign on x86
platforms.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This patch fixes byte swap issue in qla2xxx driver
to fix corrupted fabric_name passed to
/sys/class/fc_host/host*/fabric_name.
Signed-off-by: NRichard Lary <rlary@us.ibm.com>
Acked-by: NSeokmann Ju <seokmann.ju@qlogic.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

We have seen two cases where VPD on an emulex card has been incorrect
and we end up walking off the end of memory. It looks like someone made
an update (increased the length of a string) without increasing the
Length field. Then we do:

	Length -= (3+i);

And since Length is unsigned it becomes very large and we loop forever
in the encapsulating:

	while (Length > 0) {

If we make Length signed then we fall out of the loop and proceed on.

Its important to note we have only seen this in the lab and it may be
the only two cases of this in existence, but since the rest of the code
has been written to be resilient against bad VPD we may as well fix this
too.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Acked-by: NJames Smart <James.Smart@Emulex.Com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

There is some residual cleanup of the last series of patches and the
need to bump the revision number to draw the line in the sand.

The cmd->SCp.phase is set in the aac_valid_context routine, then set
again to the same value following it's return. The cmd->scsi_done is set
twice in the aac_queuecommand routine. Free up the scsidev FILO in
aac_probe_container as it is not needed further down the function in any
case. Improve the efficiency of the abort handler kernel print
parameters. Bump revision number of driver to approximate the equivalent
in the Adaptec supplied version.
Signed-off-by: NMark Salyzyn <aacraid@adaptec.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

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>

The #ifdef MODULE around the export of scsi_complete_async_scans()
which is the API the scsi_wait_scan module uses is incorrect and
causes the symbol to be undefined in certain circumstances leading to
a build failure.  Remove the defines.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

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

Implement SBC START/STOP management.  sdev->mange_start_stop is added.
When it's set to one, sd STOPs the device on suspend and shutdown and
STARTs it on resume.  sdev->manage_start_stop defaults is in sdev
instead of scsi_disk cdev to allow ->slave_config() override the
default configuration but is exported under scsi_disk sysfs node as
sdev->allow_restart is.

When manage_start_stop is zero (the default value), this patch doesn't
introduce any behavior change.
Signed-off-by: NTejun Heo <htejun@gmail.com>

Rejections fixed and
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

sd_sync_cache() should return -errno on error, fix it.
Signed-off-by: NTejun Heo <htejun@gmail.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

This address the issue of VMWare guest OS being remounted as read-only
becuase the underlying device was held busy too long and at the
same time address Engenio MPP driver concerns over infinite retries.
This patch removes the code that snoops the SAM STATUS on busy, which
would be returning DID_BUS_BUSY, instead we return the status as is.
Retry hanlding seems to be properly handled in scsi_softirq_done,
where a busy sam status would only occurr for the time specified by
(cmd->allowed +1) * cmd->timeout_per_command.
Signed-off-by: NEric Moore <Eric.Moore@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

Honour the return value of pci_enable_device(), which
seems to be a desirable thing to do:

  2.6.20-rc4
  gcc (GCC) 4.1.2 20061115 (prerelease) (Debian 4.1.1-21)

    CC [M]  drivers/message/fusion/mptbase.o
    drivers/message/fusion/mptbase.c: In function `mpt_resume':
    drivers/message/fusion/mptbase.c:1541: warning: ignoring return value
    of `pci_enable_device', declared with attribute warn_unused_result

It also in turn has mptscsih_resume() honour the return value of
mpt_resume()

I'm not sure about the handling of the other potential error cases
in mpt_resume(), of which there appear to be many. But this does
seem to be a good start.
Signed-off-by: NSimon Horman <horms@verge.net.au>
Acked-by: N"Moore, Eric" <Eric.Moore@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>