When READ_16 command is issued, the setting of SILI Bit in CDB is confirmed
and if SILI bit is off, the processing of relavent Errata is executed.

Added code for checking SILI bit for READ_16 as described in "SSC-4".
Signed-off-by: NNagalakshmi Nandigama <nagalakshmi.nandigama@lsi.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

[Resend patch as per Bernd Schubert comment ]

Issue:

Device goes offline while doing aggressive HBA reset
along with IO using some utility.

Root cause:

FW goes into bad state due to aggressive reset. Softreset does not
help to recover FW. And also aggressive reset open up the window for
Error handling thread to kicked off at the same time HBA will be in
constant RESET loop as part of aggressive reset test case can lead
Device to goes offline.

Changes:

1. Added extra check as below inside eh_timed_out call back as below.
   if(ioc->ioc_reset_in_progress) Rc = EH_TIMER_RESET

2. Removed " DOORBELL_ACTIVE" check for SAS controller from task
   management context.  Since SAS controller uses high priority queue
   for task management. This check is not required for SAS controller.

3. Moved SoftReset call to HardReset from Task Mgmt context.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Find Non-Operation IOC and remove it from OS: Detecting
dead(non-functional) ioc will be done reading doorbell register value
from fault reset thread, which has been called from work thread
context after each specific interval. If doorbell value is 0xFFFFFFFF,
it will be considered as IOC is non-operational and marked as dead
ioc.

Once Dead IOC has been detected, it will be removed at pci layer using
"pci_remove_bus_device" API.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

When READ_6 command is issued, the setting of SILI Bit in CDB is confirmed and
if SILI bit is off, the processing of relavent Errata is executed.

Earlier we did not have check for SILI bit in READ_6 CDB.
As described in "ssc-r22", Now Driver is checking SILI bit for READ_6.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Remove debug print from mptscsih_qcmd function call.
This debug print cause flood of prints and difficult to debug other issues.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

The SCSI midlayer stops all command processing when in error handling, which
means there is no chance for command reuse when the abort handler is called.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

There's a branch at the end of this function that
is supposed to normalize the return value with what
the mid-layer expects. In this one case, we get it wrong.

Also increase the verbosity of the INFO level printk
at the end of mptscsih_abort to include the actual return value
and the scmd->serial_number. The reason being success
or failure is actually determined by the state of
the internal tag list when a TMF is issued, and not the
return value of the TMF cmd. The serial_number is also
used in this decision, thus it's useful to know for debugging
purposes.

Cc: stable@kernel.org
Reported-by: NPeter M. Petrakis <peter.petrakis@canonical.com>
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Issue description:
In multipath topology, when device deletion is in transient state,
multipath driver can call blk_flush_queue() as part of path failure.
Before device get deleted from OS, Device may go OFFLINE as part of error
handling kicked off triggered from multipathing driver. Above condition hits
more frequently if device missing delay timer (which is LSI specific firmware
parameter) is non zero value.

root cause of this issue is Error handling thread is getting kicked off for
device which is not really present(in transient state of deleting).

This patch has solution for this issue. driver is now using eh_timed_out
callback. See below.

mptsas_transport_template->eh_timed_out = mptsas_eh_timed_out

Using mptsas_eh_timed_out function, driver can decide weather vdevice is
under Device missing delay or deleting state.

for either of those cases, there is BLK_EH_RESET_TIMER return to scsi mid
and error handling thread will not be kicked off for that particular scsi
command.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Cc: Stable Tree <stable@kernel.org>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Printing Doorbell register in a case of hard reset and timeout
should be useful for figuring out the state of the system.
Signed-off-by: NKei Tokunaga <tokunaga.keiich@jp.fujitsu.com>
Acked-by: N"Desai, Kashyap" <Kashyap.Desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Issue:
target reset will be queued to driver's internal queue to get schedule
later. When driver add target into internal target_reset queue we will block IOs
on those target using scsi midlayer API. Now due to some cause driver is not
executing those target_reset list and it is always in block state.

Changes:
now we are clearing target_reset queue from all other Callback context
instead of only DeviceReset context.Now wherever driver is clearing
taskmgmt_in_progress flag it is considering target_reset queue cleanup
also.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Changed the return value for Nexus Loss IOs to be DID_TRANSPORT_DISRUPTED.
What this will allow is the multi-path driver to delay the fail over
process. They would like the path to keep up as long as the nexus loss
Loginfo is return from firmware. With DID_BUS_BUSY the path fails over
immediately.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

I may have an explanation for the LSI 1068 HBA hangs provoked by ATA
pass-through commands, in particular by smartctl.

First, my version of the symptoms.  On an LSI SAS1068E B3 HBA running
01.29.00.00 firmware, with SATA disks, and with smartd running, I'm seeing
occasional task, bus, and host resets, some of which lead to hard faults of
the HBA requiring a reboot.  Abusively looping the smartctl command,

    # while true; do smartctl -a /dev/sdb > /dev/null; done

dramatically increases the frequency of these failures to nearly one per
minute.  A high IO load through the HBA while looping smartctl seems to
improve the chance of a full scsi host reset or a non-recoverable hang.

I reduced what smartctl was doing down to a simple test case which
causes the hang with a single IO when pointed at the sd interface.  See
the code at the bottom of this e-mail.  It uses an SG_IO ioctl to issue
a single pass-through ATA identify device command.  If the buffer
userspace gives for the read data has certain alignments, the task is
issued to the HBA but the HBA fails to respond.  If run against the sg
interface, neither the test code nor smartctl causes a hang.

sd and sg handle the SG_IO ioctl slightly differently.  Unless you
specifically set a flag to do direct IO, sg passes a buffer of its own,
which is page-aligned, to the block layer and later copies the result
into the userspace buffer regardless of its alignment.  sd, on the other
hand, always does direct IO unless the userspace buffer fails an
alignment test at block/blk-map.c line 57, in which case a page-aligned
buffer is created and used for the transfer.

The alignment test currently checks for word-alignment, the default
setup by scsi_lib.c; therefore, userspace buffers of almost any
alignment are given directly to the HBA as DMA targets.  The LSI 1068
hardware doesn't seem to like at least a couple of the alignments which
cross a page boundary (see the test code below).  Curiously, many
page-boundary-crossing alignments do work just fine.

So, either the hardware has an bug handling certain alignments or the
hardware has a stricter alignment requirement than the driver is
advertising.  If stricter alignment is required, then in no case should
misaligned buffers from userspace be allowed through without being
bounced or at least causing an error to be returned.

It seems the mptsas driver could use blk_queue_dma_alignment() to advertise
a stricter alignment requirement.  If it does, sd does the right thing and
bounces misaligned buffers (see block/blk-map.c line 57).  The following
patch to 2.6.34-rc5 makes my symptoms go away.  I'm sure this is the wrong
place for this code, but it gets my idea across.
Acked-by: N"Desai, Kashyap" <Kashyap.Desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

The function name must be followed by a space, hypen, space, and a
short description.
Signed-off-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NRandy Dunlap <randy.dunlap@oracle.com>
Cc: Eric Moore <Eric.Moore@lsi.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

1) corrected return value as SUCCESS instead of 0.
2) Added check in mptscsih_abort.
mptfusion do not support task abort for Volumes.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Added sanity checks before accessing vdevice and added vdevice->deleted
setting for mptfc.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Current design of mptsas is as follow.
MPTSAS will do probe() if pci id matches for available card in
system, irrespective of mode of controller. If controller is I/T mode
or I mode, things are fine. If controller is only in T mode, mptsas is
not doing complete process of mptsas_probe(). It will only make
sure IOC structure is created and IOC reference is available for
mptstm driver. Now While removing module we should take care
case of Target mode only mptsas. If we are removing IOC which is
only in Target mode, We should only detach IOC instead of
following rest of the cleanup process which is only required for T
mode controller. Now For T mode controller, only part clean up is
done instead of complete cleanup. mpt_detach will call early in case
of Target mode only controller.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Message Unit Reset - instructs the IOC to reset the Reply Post and
Free FIFO's. All the Message Frames on Reply Free FIFO are
discarded. All posted buffers are freed, and event notification is
turned off.  IOC doesnt reply to any outstanding request. This will
transfer IOC to READY state.  Message unit ready is less expensive
operations than Hard Reset.  soft reset will not force Firmware to
reload again, it only do clean up of Message units.

mpt_Soft_Hard_ResetHandler will first try for Soft Reset,if
it fails then go for big hammer reset which is Hard Reset.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h

percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files.  percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.

percpu.h -> slab.h dependency is about to be removed.  Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability.  As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.

  http://userweb.kernel.org/~tj/misc/slabh-sweep.py

The script does the followings.

* Scan files for gfp and slab usages and update includes such that
  only the necessary includes are there.  ie. if only gfp is used,
  gfp.h, if slab is used, slab.h.

* When the script inserts a new include, it looks at the include
  blocks and try to put the new include such that its order conforms
  to its surrounding.  It's put in the include block which contains
  core kernel includes, in the same order that the rest are ordered -
  alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
  doesn't seem to be any matching order.

* If the script can't find a place to put a new include (mostly
  because the file doesn't have fitting include block), it prints out
  an error message indicating which .h file needs to be added to the
  file.

The conversion was done in the following steps.

1. The initial automatic conversion of all .c files updated slightly
   over 4000 files, deleting around 700 includes and adding ~480 gfp.h
   and ~3000 slab.h inclusions.  The script emitted errors for ~400
   files.

2. Each error was manually checked.  Some didn't need the inclusion,
   some needed manual addition while adding it to implementation .h or
   embedding .c file was more appropriate for others.  This step added
   inclusions to around 150 files.

3. The script was run again and the output was compared to the edits
   from #2 to make sure no file was left behind.

4. Several build tests were done and a couple of problems were fixed.
   e.g. lib/decompress_*.c used malloc/free() wrappers around slab
   APIs requiring slab.h to be added manually.

5. The script was run on all .h files but without automatically
   editing them as sprinkling gfp.h and slab.h inclusions around .h
   files could easily lead to inclusion dependency hell.  Most gfp.h
   inclusion directives were ignored as stuff from gfp.h was usually
   wildly available and often used in preprocessor macros.  Each
   slab.h inclusion directive was examined and added manually as
   necessary.

6. percpu.h was updated not to include slab.h.

7. Build test were done on the following configurations and failures
   were fixed.  CONFIG_GCOV_KERNEL was turned off for all tests (as my
   distributed build env didn't work with gcov compiles) and a few
   more options had to be turned off depending on archs to make things
   build (like ipr on powerpc/64 which failed due to missing writeq).

   * x86 and x86_64 UP and SMP allmodconfig and a custom test config.
   * powerpc and powerpc64 SMP allmodconfig
   * sparc and sparc64 SMP allmodconfig
   * ia64 SMP allmodconfig
   * s390 SMP allmodconfig
   * alpha SMP allmodconfig
   * um on x86_64 SMP allmodconfig

8. percpu.h modifications were reverted so that it could be applied as
   a separate patch and serve as bisection point.

Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: NTejun Heo <tj@kernel.org>
Guess-its-ok-by: NChristoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>

retval should be SUCCESS/FAILED which is defined at scsi.h
retval = 0 is directing wrong return value. It must be retval = SUCCESS.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

There is a 'ioprio' field in the BIO and the Request structure.
check this priority field and set MPI_SCSIIO_CONTROL_HEADOFQ
to pass down I/O priority.
An enhancement to the LSI Disk Array Controller firmware is being
developed to look at the Head Of Queue bit to allow I/Os with the HOQ bit
set to be processed before I/Os which do not have the HOQ bit set.
In order to set the HOQ bit, the mpt fusion driver  needs to look at the
'ioprio' field in the request structure associated with the scsi command.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

This patch modifies scsi_host_template->change_queue_depth so that
it takes an argument indicating why it is being called. This will be
used so that if a LLD needs to do some extra processing when
handling queue fulls or later ramp ups, it can do so.

This is a simple port of the drivers setting a change_queue_depth
callback. In the patch I just have these LLDs adjust the queue depth
if the user was requesting it.
Signed-off-by: NMike Christie <michaelc@cs.wisc.edu>

[Vasu.Dev: v2
	Also converted pmcraid_change_queue_depth and then verified
all modules compile  using "make allmodconfig" for any new build
warnings on X86_64.

	Updated original description after combing two original
patches from Mike to make this patch git bisectable.]
Signed-off-by: NVasu Dev <vasu.dev@intel.com>
[jejb: fixed up 53c700]
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Errata:
Certain conditions on the scsi bus may casue the 53C1030 to incorrectly signal
a SCSI_DATA_UNDERRUN to the host.

Workaround 1:
For an Errata on LSI53C1030 When the length of request data
and transfer data are different with result of command (READ or VERIFY),
DID_SOFT_ERROR is set.

Workaround 2:
For potential trouble on LSI53C1030. It is checked whether the length of
request data is equal to the length of transfer and residual.
MEDIUM_ERROR is set by incorrect data.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Driver is modified to return DID_NO_CONNECT for all pending I/O
requests for bus type SAS, if it founds the target is removed at
the firmware level.
Signed-off-by: NKashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Do not set max_id value received from FW. Once SAS transport layer is
introduced max_id value is missleading to SCSI mid layer. Use max_id to
infinite value.

logic of can queue of scsi host is changed.
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Handle insufficient resources status as similar to busy status.
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Removed mptscsih_timer_expired. This timer is no more use.
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Fix recent fusion driver kernel-doc fatal error and warnings.
Signed-off-by: NRandy Dunlap <randy.dunlap@oracle.com>
Cc: Eric.Moore@lsi.com
Cc: support@lsi.com
Cc: DL-MPTFusionLinux@lsi.com
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Several of the doc book in the previous patches had incorrect multi-line short
function descriptors.  Fixed it all to be the correct single line descriptor.
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

Firmware is able to handle Broadcast primitives, but upstream driver does not
have support for broadcast primitive handling. Now this patch is mainly to
support broadcast primitives.
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

1. Handle integrated Raid device(Add/Delete) and error condition and check
   related to Raid device. is_logical_volume will represent logical volume
   device.
2. Raid device dual port support is added. Main functions to support this
   feature are mpt_raid_phys_disk_get_num_paths and mpt_raid_phys_disk_pg1.
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

Resending patch considering Grants G's code review.

Main goal to submit this patch is code cleaup.
1. Better driver debug prints and code indentation.
2. fault_reset_work_lock is not used anywhere. driver is using taskmgmt_lock
instead of fault_reset_work_lock.
3. setting pci_set_drvdata properly.
4. Ingore config request when IOC is in reset state.( ioc_reset_in_progress
is set).
5. Init/clear managment frame proprely.(INITIALIZE_MGMT_STATUS and
CLEAR_MGMT_STATUS)
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

1) rewrite of ioctl_cmds internal generated function that issue commands to
firmware, porting them to be single threaded using the generic MPT_MGMT
struct. All wait Queues are replace by completion Queue.
2) added seperate callback handler for ioctl task managment
(mptctl_taskmgmt_reply), to handle command that timeout
3) rewrite mptctl_bus_reset
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

1.) 	Added taskmgmt_quiesce_io flag in IOC and removed resetPending from
	_MPT_SCSI_HOST struct.
2.)	Reset from Scsi mid layer and internal Reset are seperate context.
	Adding DeviceResetCtx for internal Device reset frame.
	mptsas_taskmgmt_complete is optimized as part of implementation.
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

1.)	 rewrite taskmanagement request and completion routines, making them
single threaded and using the generic MPT_MGMT struct, deleting
mptscsih_TMHandler, replacing with single request TM handler
mptscsih_IssueTaskMgmt, and killing the watchdog timer functions.
2.) cleanup ioc_reset callback handlers, introducing wrappers for
synchronizing error recovery (mpt_set_taskmgmt_in_progress_flag,
mpt_clear_taskmgmt_in_progress_flag), as the fusion firmware only handles
one task management request at a time
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

Rewrite of all internal generated functions that issue commands to firmware,
porting them to be single threaded using the generic MPT_MGMT
struct. Implemented using completion Queue.
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

The reason for this change is there is a data corruption when four different
physical memory regions in the 36GB to 37GB region are
accessed. This is only affecting 1078.

The solution is we need to use different addressing when filling in
the scatter gather table for the effected memory regions.  So instead
of snooping on all four different memory holes, we treat any physical
addresses in the 36GB address with the same algorithm.

The fix is explained below
1) Ensure that the message frames are NOT located in the trouble
region. There is no remapping available for message frames, they must
be allocated outside the problem region.
2) Ensure that Sense buffers are NOT in the trouble region. There is
no remapping available.
3) Walk through the SGE entries and if any are inside the trouble region
   then they need to be remapped as discussed below.
	1) Set the Local Address bit in the SGE Flags field.
  	MPI_SGE_FLAGS_LOCAL_ADDRESS
  	2) Ensure we are using 64-bit SGEs
  	3) Set MSb (Bit 63) of the 64-bit address, this will indicate buffer
	location is Host Memory.
Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

Signed-off-by: NKashyap Desai <kadesai@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

A bug in the fusion driver was exposed by the switch to block timeout.
Basically, drivers are supposed to terminate commands once error
handling begins on them.  The fusion apparently wasn't doing this.
Under the old timeout regime, completions on terminated commands would
by and large get ignored because of the way command timeouts used to
work. The new block timers are very intolerant to this, though,
becuase the request gets cleaned and freed.

Fixes: http://bugzilla.kernel.org/show_bug.cgi?id=12195Reported-by: NAlex Shi <alex.shi@intel.com>
Tested-by: NMing Lin <ming.m.lin@intel.com>
Cc: Eric Moore <Eric.Moore@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

Make SPI timeout 10s the same as SAS
Signed-off-by: NBernd Schubert <bs@q-leap.de>
Acked-by: N"Prakash, Sathya" <Sathya.Prakash@lsi.com>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>

[jejb: fixed up a ton of missed conversions.

 All of you are on notice this has happened, driver trees will now
 need to be rebased]
Signed-off-by: NHarvey Harrison <harvey.harrison@gmail.com>
Cc: SCSI List <linux-scsi@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJames Bottomley <James.Bottomley@HansenPartnership.com>