It is possible a bio request can not be submitted as a single NVMe IO
command if the bio_vec is not mergeable with the NVMe PRP alignement
constraints. This condition was handled by submitting an IO for the
mergeable portion then submitting a follow on IO for the remaining data
after the previous IO completes. The remainder to be sent was tracked
by manipulating the bio->bi_idx and bio->bi_sector. This patch splits
the request as many times as necessary and submits the bios together.

Since submitting the bio may cause it to be requeued on split,
nvme_resubmit_bios had to be modified to remove the wait queue when
the bio list is empty prior to submitting the bio since a split would
have added the wait queue a second time, corrupting the wait queue head
task list.

There are a few other benefits from doing this: it fixes a potential
issue with the previous handling of a non-mergeable bio as the requeuing
method could would use an unlocked nvme_queue if the callback isn't
invoked on the queue's associated cpu; it will be possible to retry a
failed bio if desired at some later time since it does not manipulate
the original bio; the bio integrity extensions require the bio to be in
its original condition for the checks to work correctly if we implement
the end-to-end data protection in the future.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

There is no situation that could occur where we could error out of this
function and require cleaning up allocated namespaces.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

The nvme_queue's depth is not set if we fail to allocate submission queue
entries, which was being used to determine how much coherent memory to
free on error. Use the depth variable instead.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Fixes a potential memory leak if requesting the admin queue irq fails.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Translates a scsi unmap request from SG_IO ioctl to NVMe
data-set-management deallocate.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Acked-by: Vishal Verma <vishal.l.verma@linux.intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Fixes nvme queue usages in scsi-to-nvme translation code to not get
a queue more often than it is being put, and not use the queue in an
unsafe way without it being locked.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Acked-by: Vishal Verma <vishal.l.verma@linux.intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

The kthread has two tasks; handling timeouts (for which it runs once per
second), and submitting queued BIOs.  If a BIO happens to be queued after
the thread has processed the queue but before it calls schedule_timeout(),
the thread will sleep for a second before submitting it, which can cause
performance problems in some rare cases (that will become more common in
a subsequent patch).
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Tested-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Registers a miscellaneous device for each nvme controller probed. This
creates character device files as /dev/nvmeN, where N is the device
instance, and supports nvme admin ioctl commands so devices without
namespaces can be managed.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

When constructing the command, dsmgmt needs to be treated as a 32-bit
value, not a 16-bit value.  reftag, apptag and appmask all need to be
converted from native-endian to little-endian.  Again, sparse's bitwise
warnings caught this problem.  Thanks to Keith for pointing out the
correct way to fix the reftag.
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>
Acked-by: Keith Busch <keith.busch@intel.com>

The sparse bitwise checks pointed out that I needed to shift the status
before changing its endianness, not after.
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Sparse produced warnings for some instances of
mismatched types and direct userspace dereferences.
This patch fixes those for the scsi emulation layer.
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

The nvme_dev_add() function currently returns the last error code that it
saw, which (if everything else succeeds) happens to be the result of an
optional command, so it can legitimately fail.  Looking at the error path
more closely reveals that we should return success from this function,
even if no device namespaces are added.  So once the queues are created
and the device has responded to Identify, make sure that this function
succeeds.
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>
Acked-by: Keith Busch <keith.busch@intel.com>

Introduce nvme_block_nr() to help convert sectors to block numbers.
This fixes an integer overflow in the SCSI conversion layer, and it's
slightly less typing than opencoding it.
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>
Acked-by: Keith Busch <keith.busch@intel.com>

The nvme driver has a "once per second" event where the management kthread
wakes up the system and then reschedules itself for 1 second later.
For power efficiency reasons, I'd like this timer to happen together
with other wakeups in the system.

This patch makes the schedule_timeout() call in the kthread use
round_jiffies_relative(), causing the wakeup to at least align with other
"once per X seconds" events in the kernel.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Tested-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Translates SCSI commands in SG_IO ioctl to NVMe commands.
Uses the scsi-nvme translation spec from nvmexpress.org as reference.
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

The SCSI emulation has the ability to send format commands, so we need
to add the definition of the command.  Also add a missing error code.
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

nvme-scsi.c uses several data structures and definitions that were
previously private to nvme-core.c.  Move the definitions to nvme.h,
protected by __KERNEL__.
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

In preparation for adding nvme-scsi.c
It is preferable to retain the module name 'nvme'
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

This adds discard support to block queues if the nvme device is capable of
deallocating blocks as indicated by the controller's optional command support.
A discard flagged bio request will submit an NVMe deallocate Data Set
Management command for the requested blocks.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

Pull timer fix from Thomas Gleixner:
 "A single bugfix which prevents that a non functional timer device is
  selected to provide the fallback device, which is supposed to serve
  timer interrupts on behalf of non functional devices ..."

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  clockevents: Don't allow dummy broadcast timers

Pull infiniband/rdma fixes from Roland Dreier:
 "Small batch of InfiniBand/RDMA fixes for 3.9:

   - Fix for TX lockup in IPoIB
   - QLogic -> Intel update for qib driver
   - Small static checker fix for qib
   - Fix error path return value in cxgb4"

* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  IB/qib: change QLogic to Intel
  IB/ipath: Silence a static checker warning
  IPoIB: Fix send lockup due to missed TX completion
  RDMA/cxgb4: Fix error return code in create_qp()

Pull ARM SoC bug fixes from Arnd Bergmann:
 "Four patches for arm-soc this week:

   - Kevin Hilman is no longer reachable under his previous email
     address.  He submitted the patch earlier, but nobody felt
     responsible to pick it up.

   - One Tegra fix for an incorect register address in device tree.

   - IMX multiplatform support exposes a configuration option that leads
     to unbootable kernels on all other machines and that needs to
     depend on that platform.

   - A nontrivial bug fix for the setup of the mxs video output."

* tag 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  MAINTAINERS: update email address for Kevin Hilman
  ARM: tegra: fix register address of slink controller
  ARM: imx: add dependency check for DEBUG_IMX_UART_PORT
  ARM: video: mxs: Fix mxsfb misconfiguring VDCTRL0

Pull nfsd bugfixes from J Bruce Fields:
 "Fixes for a couple mistakes in the new DRC code.  And thanks to Kent
  Overstreet for noticing we've been sync'ing the wrong range on stable
  writes since 3.8."

* 'for-3.9' of git://linux-nfs.org/~bfields/linux:
  nfsd: fix bad offset use
  nfsd: fix startup order in nfsd_reply_cache_init
  nfsd: only unhash DRC entries that are in the hashtable

Pull drm fixes from Dave Airlie:
 "Exynos and Intel fixes.

  The intel fixes are fairly straightforward, mostly reverts due to bugs
  found.  The exynos one is a big larger since they found some issues
  with the G2D engine and iommu interaction, and needed to verify the
  operations a lot better than they were previously, otherwise a user
  app can just crash the kernel with an iommu fault."

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  Revert "drm/i915: write backlight harder"
  drm/i915: don't disable the power well yet
  Revert "drm/i915: set TRANSCODER_EDP even earlier"
  drm/exynos: Check g2d cmd list for g2d restrictions
  drm/exynos: Add a new function to get gem buffer size
  drm/exynos: Deal with g2d buffer info more efficiently
  drm/exynos: Clean up some G2D codes for readability
  drm/exynos: Fix G2D core malfunctioning issue
  drm/exynos: clear node object type at gem unmap
  drm/exynos: Fix error routine to getting dma addr.
  drm/exynos: Replaced kzalloc & memcpy with kmemdup
  drm/exynos: fimd: calculate the correct address offset
  drm/exynos: Make mixer_check_timing static
  drm/exynos: modify the compatible string for exynos fimd

Daniel writes:
"Just three revert/disable by default patches, one of them cc: stable
(since the offending commit was cc: stable, too)."

* 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel:
  Revert "drm/i915: write backlight harder"
  drm/i915: don't disable the power well yet
  Revert "drm/i915: set TRANSCODER_EDP even earlier"

Merge branch 'exynos-drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into HEAD

Inki writes:
Includes bug fixes and code cleanups.
And it considers some restrictions to G2D hardware.
With this, the malfunction and page fault issues to g2d driver
would be fixed.

* 'exynos-drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos:
  drm/exynos: Check g2d cmd list for g2d restrictions
  drm/exynos: Add a new function to get gem buffer size
  drm/exynos: Deal with g2d buffer info more efficiently
  drm/exynos: Clean up some G2D codes for readability
  drm/exynos: Fix G2D core malfunctioning issue
  drm/exynos: clear node object type at gem unmap
  drm/exynos: Fix error routine to getting dma addr.
  drm/exynos: Replaced kzalloc & memcpy with kmemdup
  drm/exynos: fimd: calculate the correct address offset
  drm/exynos: Make mixer_check_timing static
  drm/exynos: modify the compatible string for exynos fimd

Merge tag 'pinctrl-fixes-for-v3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl

Pull pinctrl fixes from Linus Walleij:
 "Here are a few pinctrl fixes for the v3.9 rc series:
   - Usecount bounds checking so we do not go below zero on mux
     usecounts.
   - Loop range checking in GPIO ranges in the DT range parser.
   - Proper print in debugfs for pinconf state.
   - Fix compilation bug in generic pinconf code.
   - Minor bugfixes to abx500 and mvebu drivers."

* tag 'pinctrl-fixes-for-v3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  pinmux: forbid mux_usecount to be set at UINT_MAX
  pinctrl: mvebu: fix checking for SoC specific controls
  pinctrl: generic: Fix compilation error
  pinctrl: Print the correct information in debugfs pinconf-state file
  pinctrl: abx500: Fix checking if pin use AlternateFunction register
  gpio: fix wrong checking condition for gpio range

Pull x86 fixes from Peter Anvin:
 "A collection of minor fixes, more EFI variables paranoia
  (anti-bricking) plus the ability to disable the pstore either as a
  runtime default or completely, due to bricking concerns."

* 'x86/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  efivars: Fix check for CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE
  x86, microcode_intel_early: Mark apply_microcode_early() as cpuinit
  efivars: Handle duplicate names from get_next_variable()
  efivars: explicitly calculate length of VariableName
  efivars: Add module parameter to disable use as a pstore backend
  efivars: Allow disabling use as a pstore backend
  x86-32, microcode_intel_early: Fix crash with CONFIG_DEBUG_VIRTUAL
  x86-64: Fix the failure case in copy_user_handle_tail()

This reverts commit cf0a6584.

Turns out that cargo-culting breaks systems. Note that we can't revert
further, since

commit 770c1231
Author: Takashi Iwai <tiwai@suse.de>
Date:   Sat Aug 11 08:56:42 2012 +0200

    drm/i915: Fix blank panel at reopening lid

fixed a regression in 3.6-rc kernels for which we've never figured out
the exact root cause. But some further inspection of the backlight
code reveals that it's seriously lacking locking. And especially the
asle backlight update is know to get fired (through some smm magic)
when writing specific backlight control registers. So the possibility
of suffering from races is rather real.

Until those races are fixed I don't think it makes sense to try
further hacks. Which sucks a bit, but sometimes that's how it is :(

References: http://www.mail-archive.com/intel-gfx@lists.freedesktop.org/msg18788.html
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=47941Tested-by: Takashi Iwai <tiwai@suse.de>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: stable@vger.kernel.org (the reverted commit was cc: stable, too)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

We're still not 100% ready to disable the power well, so don't disable
it for now. When we disable it we break the audio driver (because some
of the audio registers are on the power well) and machines with eDP on
port D (because it doesn't use TRANSCODER_EDP).

Also, instead of just reverting the code, add a Kernel option to let
us disable it if we want. This will allow us to keep developing and
testing the feature while it's not enabled.

This fixes problems caused by the following commit:
  commit d6dd9eb1
  Author: Daniel Vetter <daniel.vetter@ffwll.ch>
  Date:   Tue Jan 29 16:35:20 2013 -0200
       drm/i915: dynamic Haswell display power well support

References: http://www.mail-archive.com/intel-gfx@lists.freedesktop.org/msg18788.html
Cc: Takashi Iwai <tiwai@suse.de>
Cc: Mengdong Lin <mengdong.lin@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

This reverts commit cc464b2a.

The reason is that Takashi Iwai reported a regression bisected to this
commit:

http://www.mail-archive.com/intel-gfx@lists.freedesktop.org/msg18788.html

His machine has eDP on port D (usual desktop all-in-on setup), which
intel_dp.c identifies as an eDP panel, but the hsw ddi code
mishandles.

Closer inspection of the code reveals that haswell_crtc_mode_set also
checks intel_encoder_is_pch_edp when setting is_cpu_edp. On haswell
that doesn't make much sense (since there's no edp on the pch), but
what this function _really_ checks is whether that edp connector is on
port A or port D. It's just that on ilk-ivb port D was on the pch ...

So that explains why this seemingly innocent change killed eDP on port
D. Furthermore it looks like everything else accidentally works, since
we've never enabled eDP on port D support for hsw intentionally (e.g.
we still register the HDMI output for port D in that case).

But in retrospective I also don't like that this leaks highly platform
specific details into common code, and the reason is that the drm
vblank layer sucks. So instead I think we should:
- move the cpu_transcoder into the dynamic pipe_config tracking (once
  that's merged).
- fix up the drm vblank layer to finally deal with kms crtc objects
  instead of int pipes.

v2: Pimp commit message with the better diagnosis as discussed with
Paulo on irc.

Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Takashi Iwai <tiwai@suse.de>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>

Pull SCSI target fixes from Nicholas Bellinger:
 "These are mostly minor fixes this time around.  The iscsi-target CHAP
  big-endian bugfix and bump FD_MAX_SECTORS=2048 default patch to allow
  1MB sized I/Os for FILEIO backends on >= v3.5 code are both CC'ed to
  stable.

  Also, there is a persistent reservations regression that has recently
  been reported for >= v3.8.x code, that is currently being tracked down
  for v3.9."

* git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:
  target/pscsi: Reject cross page boundary case in pscsi_map_sg
  target/file: Bump FD_MAX_SECTORS to 2048 to handle 1M sized I/Os
  tcm_vhost: Flush vhost_work in vhost_scsi_flush()
  tcm_vhost: Add missed lock in vhost_scsi_clear_endpoint()
  target: fix possible memory leak in core_tpg_register()
  target/iscsi: Fix mutual CHAP auth on big-endian arches
  target_core_sbc: use noop for SYNCHRONIZE_CACHE

Pull md fixes from NeilBrown:
 "A few bugfixes for md

   - recent regressions in raid5
   - recent regressions in dmraid
   - a few instances of CONFIG_MULTICORE_RAID456 linger

  Several tagged for -stable"

* tag 'md-3.9-fixes' of git://neil.brown.name/md:
  md: remove CONFIG_MULTICORE_RAID456 entirely
  md/raid5: ensure sync and DISCARD don't happen at the same time.
  MD: Prevent sysfs operations on uninitialized kobjects
  MD RAID5: Avoid accessing gendisk or queue structs when not available
  md/raid5: schedule_construction should abort if nothing to do.

Pull libata updates from Jeff Garzik:
 "Simple stuff.  See one-line summaries."

* tag 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  pata_samsung_cf: use module_platform_driver_probe()
  [libata] Avoid specialized TLA's in ZPODD's Kconfig
  libata-acpi.c: fix copy and paste mistake in ata_acpi_register_power_resource
  sata_fsl: Remove redundant NULL check before kfree
  ahci: Add Device IDs for Intel Wellsburg PCH
  ata_piix: Add MODULE_PARM_DESC to prefer_ms_hyperv

Pull i2c fixes from Wolfram Sang:
 "One bugfix for the tegra driver.  Two updates regarding email
  addresses and MAINTAINERS which I like to have up-to-date so people
  can be reached immediately.  While we are here, there is on PCI_ID
  addition."

* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
  MAINTAINERS: add maintainer entry for atmel i2c driver
  i2c: Fix my e-mail address in drivers and documentation
  i2c: iSMT: add Intel Avoton DeviceIDs
  i2c: tegra: check the clk_prepare_enable() return value

Pull watchdog fixes from Wim Van Sebroeck:
 "Fix a boot issues and correct the AcpiMmioSel bitmask in the
  sp5100_tco watchdog device driver"

* git://www.linux-watchdog.org/linux-watchdog:
  watchdog: sp5100_tco: Set the AcpiMmioSel bitmask value to 1 instead of 2
  watchdog: sp5100_tco: Remove code that may cause a boot failure

When KMS has parsed an EDID "detailed timing", it leaves the frame rate
zeroed.  Consecutive (debug-) output of that mode thus yields 0 for
vsync.  This simple fix also speeds up future invocations of
drm_mode_vrefresh().

While it is debatable whether this qualifies as a -stable fix I'd apply
it for consistency's sake; drm_helper_probe_single_connector_modes()
does the same thing already for all probed modes.

Cc: stable@vger.kernel.org
Signed-off-by: Torsten Duwe <duwe@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>