There's a zero day mistake in the megaraid driver in that the code that
obtains the version number does a >> 8 on a char quantity.  This >>8 causes a
sparse warning because it always produces zero.  Al Viro suggested these
shifts should be >> 4 thus treating the firmware version as a BCD quantity.
However, in the interests of safety we've elected to replace the >> 8
quantities with an explicit zero, thus quieting the sparse warning while
preserving the same (albeit incorrect) version number as had previously been
seen.
Signed-off-by: NAdam Radford <aradford@gmail.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Instead of open coding this function use kstrtoul_from_user() directly.
Signed-off-by: NStephen Boyd <sboyd@codeaurora.org>
Acked-by: NDouglas Gilbert <dgilbert@interlog.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

The sdev is deleted from starved list and then try to dispatch from this
device. It's quite possible the sdev can't eventually dispatch a request,
then the sdev will be in starved list tail. This isn't fair.
There are two cases here:
1. unplug path. scsi_request_fn() calls to scsi_target_queue_ready(), then
the dev is removed from starved list, but quite possible host queue isn't
ready, the dev is moved to starved list without dispatching any request.
2. scsi_run_queue path. It deletes the dev from starved list first (both
global and local starved lists), then handles the dev. Then we could have
the same process like case 1.

This patch fixes the first case. Case 2 isn't fixed, because there is a
rare case scsi_run_queue finds host isn't busy but scsi_request_fn finds
host is busy (other CPU is faster to get host queue depth). Not deleting
the dev from starved list in scsi_run_queue will keep scsi_run_queue
looping (though this is very rare case, because host will become busy).
Fortunately fixing case 1 already gives big improvement for starvation in
my test. In a 12 disk JBOD setup, running file creation under EXT4, this
gives 12% more throughput.
Signed-off-by: NShaohua Li <shaohua.li@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

When expander connected in x2 or x4 mode and with IO runnning, if
a cable from wideport is plugged out from the phy, IO's start failing
on all the targets.

Observed that when cable is pulled with IO running, cominit is
happening on all the links and IO's start dropping to 0 and eventually
the whole IO fails. Second observation, target is trying to open and
SCU is responding with "Open reject no destination".

A cause of the problem is when the port went from the "ready
configuring substate" back to "ready configuring substate" as a result
of phy being pulled off, scic suspended the port task scheduler
register. As a result no IO was allowed and in the "substate
configuring enter" routine the IO never goes back to 0. As a result
the port never comes out of "ready substate configuring".

The patch adds a mechanism of activate and deactivate phy when a port
link up, which fixes the problem.
Signed-off-by: NBartek Nowakowski <bartek.nowakowski@intel.com>
Signed-off-by: NMaciej Trela <maciej.trela@intel.com>
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Split scu_link_layer_start_oob function to reset and enable and
add flush after all steps.
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

When the first phy of a wide port comes up, don't report the port ready
yet, always wait for 250 miliseconds then config the port with all phys
added to the port. So that we can avoid reporting wide port device too
early to kernel, which caused the first IOs (report luns, inquirys)
failed due to not all the phys are configured into its port. Changes
also made that the phys in a wide port don't need to go through half
second wait time for consuming power.
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

When hot insert the wide port device through the mini-sas port,
the first IOs (Report Luns or Inquiry) may fail due to the device
trying to open to a SCU phy that is still in suspended state. This
IO failure causes the wide port device stuck in UPDATING_PORT_WIDTH
state.
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Arrange for task_contexts prepared for the wide targets to account for
all the attached phys in the port.
Signed-off-by: NBartek Nowakowski <bartek.nowakowski@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Failure seen pulling a cable from a x4 port configured in manual port
configuration (MPC) mode (MPC mode is set by the the OEM paramaters
provided by the platform or isci_firmware.bin).  While IO running to
devices behind and expander, plugging out the cable from phy is causing
IO failures and IO drops on disks and never recover.

It happens because during link up/down the phy were being taken out of
the port.

Fix: during link down the phy is kept in the same logical port.
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Bump the version now that the driver has atapi support and the initial
round of hotplug fixes.  The EXPERIMENTAL tag should have been removed a
while back.  While we're here also kill the "select SCSI_SAS_HOST_SMP"
as the build error was separately fixed by commit d962480e "[SCSI]
libsas: fix try_test_sas_gpio_gp_bit() build error".
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

As the field was never set, isci_print_tmf() using 'isci_tmf->device'
sometimes causes a kernel crash if the dev_dbg() statement is enabled.
Remove the unused field both from isci_tmf struct definition and from
isci_print_tmf()
Signed-off-by: NMaciej Trela <maciej.trela@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Gen-3 operation is marginal, default to gen-2 for now.
Signed-off-by: NJeff Skirvin <jeffrey.d.skirvin@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

"No task timeout timer reduced from 20 to 2 This timer controls how
long the SCU hardware will hold open the TX side of the connection
before sending a DONE.  The timer allows the hardware to attempt to
optimize the DONE/CLOSE behavior to allow for new COMMAND IU to be
posted.  In practice closing the connection quicker is better."
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

v1.3 allows the attenuation of the attached cables to be specified to
the driver in terms of 'short', 'medium', and 'long' (see probe_roms.h).
These settings (per phy) are retrieved from the platform oem-parameters
(BIOS rom) or via a module parameter override.
Reviewed-by: NJiangbi Liu <jiangbi.liu@intel.com>
Signed-off-by: NJeff Skirvin <jeffrey.d.skirvin@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

v1.1 allows finer grained tuning of the SSC (spread-spectrum-clocking)
settings for SAS and SATA.  See notes in probe_roms.h
Signed-off-by: NDave Jiang <dave.jiang@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

C1 silicon requires updates to the phy tuning recipe and also support
for user provided cable selects (per-phy) for short, medium, and long
cables.  Default to 'short' awaiting support for selecting the cable via
oem parameters.
Reviewed-by: NJiangbi Liu <jiangbi.liu@intel.com>
Signed-off-by: NMarcin Tomczak <marcin.tomczak@intel.com>
Signed-off-by: NJeff Skirvin <jeffrey.d.skirvin@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Before updating the code to support the latest platform updates and
silicon revision cleanup some of the long deref chains.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

This parameter blob and generator program have been moved to the
linux-firmware.git repository.
Signed-off-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Signed-off-by: NAdam Radford <aradford@gmail.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

This code has never worked correctly, doesn't disable interrupts when
set as a module parameter, doesn't disable interrupts when set after
driver load time in sysfs node, etc.
Signed-off-by: NAdam Radford <aradford@gmail.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Mask off flags in the ioctl path to prevent memory scribble with older MegaCLI
versions.
Signed-off-by: NAdam Radford <aradford@gmail.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

The following patch for megaraid_sas fixes the reglockFlags field for
degraded raid5/6 for MR9360/9380, which will result in a performance
improvement.
Signed-off-by: NAdam Radford <aradford@gmail.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Don't call kthread_stop with a spin lock held and interrupts
disabled because kthread_stop will sleep waiting for the thread
to stop.
Signed-off-by: NStephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Welcome to Arvind Kumar, our new pvscsi maintainer.
Signed-off-by: NAlok N Kataria <akataria@vmware.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

This patch fixes a bug where devloss is not called on fc_host teardown.
The issue is seen if the LLDD uses rport_rolechg to add the target role
to an rport.

When an rport goes away, the LLDD will call fc_remote_port_delete, which
will start the devloss timer.  If the timer expires, the transport will
call the devloss callback and set the FC_RPORT_DEVLOSS_CALLBK_DONE flag.
However, the rport structure is not deleted, it is retained to store the
SCSI id mappings for the rport in case it comes back. In the scenario
where it does come back, and the driver calls fc_remote_port_add, but does
not indicate the "target" role for the rport - the create will clear the
structure, but forgets to clear FC_RPORT_DEVLOSS_CALLBK_DONE flag (which
is cleared if it's added with the target role). The secondary call, of
fc_remote_port_rolechg to add the target role also does not clear the flag.
Thus, the next time the rport goes away, the resulting devloss timer
expiration will not call the driver callback as the flag is still set.

This patch adds the FC_RPORT_DEVLOSS_CALLBK_DONE flags to the list of
those that are cleared upon reuse of the rport structure.
Signed-off-by: NAlex Iannicelli <alex.iannicelli@emulex.com>
Signed-off-by: NJames Smart <james.smart@emulex.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

This patch adds a check-condition in scsi_dh_alua handler for a retry.
Sometimes, I have seen attach failing due to this check-condition with
following error messages on NetApp E series storage.

Dec  7 15:31:01 nilgiris kernel: [102979.696673] scsi 3:0:2:9: alua: port group 00 rel port 01
Dec  7 15:31:01 nilgiris kernel: [102979.697082] scsi 3:0:2:9: alua: rtpg failed with 8000002
Dec  7 15:31:01 nilgiris kernel: [102979.697086] scsi 3:0:2:9: alua: rtpg sense code 06/2a/01
Dec  7 15:31:01 nilgiris kernel: [102979.697088] scsi 3:0:2:9: alua: not attached
Signed-off-by: NBabu Moger <babu.moger@netapp.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

This patch re-implements LUN Masking feature using SCSI Slave Callouts.  With
the new design in the slave_alloc entry point; for each new LUN discovered we
check with our internal LUN Masking config whether to expose or to mask this
particular LUN. We return -ENXIO (No such device or address) from slave_alloc
for the LUNs we don't want to be exposed.  We also notify the SCSI mid-layer
to do a sequential LUN scan rather than REPORT_LUNS based scan if LUN masking
is enabled on our HBA port, since a -ENXIO from any LUN in REPORT_LUNS based
scan translates to a scan abort.  This patch also handles the dynamic lun
masking config change from enable to disable or vice-versa by resetting
sdev_bflags of LUN 0 appropriately.
Signed-off-by: NKrishna Gudipati <kgudipat@brocade.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

This patch reverts the current LUN Masking Implementation.  We re-implemented
this feature using the SCSI Slave Callout's as per the review comments.
Signed-off-by: NKrishna Gudipati <kgudipat@brocade.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Patch fixes the possible NULL pointer dereference when we try to add or delete
a rpwwn to the lunmask config which is not zoned to this port.  Check if the
FCS rport is not NULL before de-referencing it.
Signed-off-by: NKrishna Gudipati <kgudipat@brocade.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

skb_linearize already has a check for skb_is_nonlinear,
there is no need to duplicate the check in fcoe.c. This
patch simply removes the unnecessary check and calls
skb_linearize unconditionally.
Reported-by: Npatrick kelle <patrick.kelle81@gmail.com>
Signed-off-by: NRobert Love <robert.w.love@intel.com>
Acked-by: Npatrick kelle <patrick.kelle81@gmail.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Signed-off-by: NVikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

If ql4xdisablesysfsboot = 0 and sendtargets entry as boot index then
driver does export sendtarget entries in sysfs but iscsistart does not
do discovery. So in this case let driver do the discovery and
login to the targets.
Signed-off-by: NManish Rangankar <manish.rangankar@qlogic.com>
Signed-off-by: NVikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Signed-off-by: NNilesh Javali <nilesh.javali@qlogic.com>
Signed-off-by: NVikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

A wrong default timeout value programmed in the adapter causes driver
to wait for that much time while waiting for target discoveries to complete.
This could add huge delays during the driver load time. To avoid this,
limit the default timeout value to 12 seconds if the default timeout value
set in adapter is less than 12 seconds and greater than 120 seconds.
Signed-off-by: NNilesh Javali <nilesh.javali@qlogic.com>
Signed-off-by: NVikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

Make qla4xxx_build_ddb_list shorter by adding more helper functions.
Signed-off-by: NLalit Chandivade <lalit.chandivade@qlogic.com>
Signed-off-by: NVikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

In alloc_pdu, libcxgbi tries to allocate a skb with GFP_ATOMIC, which
may potentially fail. When it happens, the current code prints a warning
message.

When the system is under IO stress, this failure may happen lots of
times and it usually scares users.

Instead of printing the warning message, the code now increases the
tx_dropped statistics for the ethernet interface wich is doing the iscsi
task.
Signed-off-by: NThadeu Lima de Souza Cascardo <cascardo@linux.vnet.ibm.com>
Acked-by: NKaren Xie <kxie@chelsio.com>
Reviewed-by: NMike Christie <michaelc@cs.wisc.edu>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

sym53c8xx_slave_destroy unconditionally assumes that sym53c8xx_slave_alloc has
succesesfully allocated a sym_lcb. This can lead to a NULL pointer dereference
(exposed by commit 4e6c82b3).
Signed-off-by: NStratos Psomadakis <psomas@gentoo.org>
Cc: stable@vger.kernel.org
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

If create_singlethread_workqueue() failes, rdac_init should fail too.
Signed-off-by: NRichard Weinberger <richard@nod.at>
Acked-by: N"Moger, Babu" <Babu.Moger@netapp.com>
Signed-off-by: NJames Bottomley <JBottomley@Parallels.com>

MTD pull for 3.3

* tag 'for-linus-3.3' of git://git.infradead.org/mtd-2.6: (113 commits)
  mtd: Fix dependency for MTD_DOC200x
  mtd: do not use mtd->block_markbad directly
  logfs: do not use 'mtd->block_isbad' directly
  mtd: introduce mtd_can_have_bb helper
  mtd: do not use mtd->suspend and mtd->resume directly
  mtd: do not use mtd->lock, unlock and is_locked directly
  mtd: do not use mtd->sync directly
  mtd: harmonize mtd_writev usage
  mtd: do not use mtd->lock_user_prot_reg directly
  mtd: mtd->write_user_prot_reg directly
  mtd: do not use mtd->read_*_prot_reg directly
  mtd: do not use mtd->get_*_prot_info directly
  mtd: do not use mtd->read_oob directly
  mtd: mtdoops: do not use mtd->panic_write directly
  romfs: do not use mtd->get_unmapped_area directly
  mtd: do not use mtd->get_unmapped_area directly
  mtd: do use mtd->point directly
  mtd: introduce mtd_has_oob helper
  mtd: mtdcore: export symbols cleanup
  mtd: clean-up the default_mtd_writev function
  ...

Fix up trivial edit/remove conflict in drivers/staging/spectra/lld_mtd.c

* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu: (53 commits)
  iommu/amd: Set IOTLB invalidation timeout
  iommu/amd: Init stats for iommu=pt
  iommu/amd: Remove unnecessary cache flushes in amd_iommu_resume
  iommu/amd: Add invalidate-context call-back
  iommu/amd: Add amd_iommu_device_info() function
  iommu/amd: Adapt IOMMU driver to PCI register name changes
  iommu/amd: Add invalid_ppr callback
  iommu/amd: Implement notifiers for IOMMUv2
  iommu/amd: Implement IO page-fault handler
  iommu/amd: Add routines to bind/unbind a pasid
  iommu/amd: Implement device aquisition code for IOMMUv2
  iommu/amd: Add driver stub for AMD IOMMUv2 support
  iommu/amd: Add stat counter for IOMMUv2 events
  iommu/amd: Add device errata handling
  iommu/amd: Add function to get IOMMUv2 domain for pdev
  iommu/amd: Implement function to send PPR completions
  iommu/amd: Implement functions to manage GCR3 table
  iommu/amd: Implement IOMMUv2 TLB flushing routines
  iommu/amd: Add support for IOMMUv2 domain mode
  iommu/amd: Add amd_iommu_domain_direct_map function
  ...