GitHub/moto-9609/android_kernel_motorola_exynos9610.git
10 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 12 Jul 2014 00:10:05 +0000 (17:10 -0700)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Peter Anvin:
 "A couple of further build fixes for the VDSO code.

  This is turning into a bit of a headache, and Andy has already come up
  with a more ultimate cleanup, but most likely that is 3.17 material"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86-32, vdso: Fix vDSO build error due to missing align_vdso_addr()
  x86-64, vdso: Fix vDSO build breakage due to empty .rela.dyn

10 years agoMerge branch 'for-3.16' of git://linux-nfs.org/~bfields/linux
Linus Torvalds [Fri, 11 Jul 2014 22:10:04 +0000 (15:10 -0700)]
Merge branch 'for-3.16' of git://linux-nfs.org/~bfields/linux

Pull nfsd bugfix from Bruce Fields:
 "Another xdr encoding regression that may cause incorrect encoding on
  failures of certain readdirs"

* 'for-3.16' of git://linux-nfs.org/~bfields/linux:
  nfsd: Fix bad reserving space for encoding rdattr_error

10 years agoMerge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Linus Torvalds [Fri, 11 Jul 2014 22:09:15 +0000 (15:09 -0700)]
Merge tag 'arm64-fixes' of git://git./linux/kernel/git/arm64/linux

Pull arm64 fixes from Catalin Marinas:
 "ARM64 implementation of TASK_SIZE_OF and exporting two functions to
  modules"

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64: implement TASK_SIZE_OF
  arm64: export __cpu_{clear,copy}_user_page functions

10 years agoMerge tag 'dm-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device...
Linus Torvalds [Fri, 11 Jul 2014 16:33:36 +0000 (09:33 -0700)]
Merge tag 'dm-3.16-fixes' of git://git./linux/kernel/git/device-mapper/linux-dm

Pull device mapper fixes from Mike Snitzer:

 - Fix DM multipath IO hang regression from 3.15 due to logic bug in
   multipath_busy.  This impacted cable-pull testing and also the
   ability to boot with IPR SCSI on a POWER8 box.

 - Fix possible deadlock with deferred device removal by using a new
   dedicated workqueue rather than using the system workqueue.

 - Fix NULL pointer crash due to race condition in dm-io's wake up code
   for sync_io by using a completion.

 - Update dm-crypt and dm-zero author name following legal name change;
   this is important to Jana so I didn't see any reason to hold it back.

* tag 'dm-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
  dm mpath: fix IO hang due to logic bug in multipath_busy
  dm io: fix a race condition in the wake up code for sync_io
  dm crypt, dm zero: update author name following legal name change
  dm: allocate a special workqueue for deferred device removal

10 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Fri, 11 Jul 2014 16:32:39 +0000 (09:32 -0700)]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

Pull powerpc fixes from Ben Herrenschmidt:
 "Here are a few more powerpc fixes for 3.16

  There's a small series of 3 patches that fix saving/restoring MMUCR2
  when using KVM without which perf goes completely bonkers in the host
  system.  Another perf fix from Anton that's been rotting away in
  patchwork due to my poor eyesight, a couple of compile fixes, a little
  addition to the WSP removal by Michael (removing a bit more dead
  stuff) and a fix for an embarassing regression with our soft irq
  masking"

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  powerpc/perf: Never program book3s PMCs with values >= 0x80000000
  powerpc: Disable RELOCATABLE for COMPILE_TEST with PPC64
  powerpc/perf: Clear MMCR2 when enabling PMU
  powerpc/perf: Add PPMU_ARCH_207S define
  powerpc/kvm: Remove redundant save of SIER AND MMCR2
  powerpc/powernv: Check for IRQHAPPENED before sleeping
  powerpc: Clean up MMU_FTRS_A2 and MMU_FTR_TYPE_3E
  powerpc/cell: Fix compilation with CONFIG_COREDUMP=n

10 years agopowerpc/perf: Never program book3s PMCs with values >= 0x80000000
Anton Blanchard [Wed, 28 May 2014 22:15:38 +0000 (08:15 +1000)]
powerpc/perf: Never program book3s PMCs with values >= 0x80000000

We are seeing a lot of PMU warnings on POWER8:

    Can't find PMC that caused IRQ

Looking closer, the active PMC is 0 at this point and we took a PMU
exception on the transition from negative to 0. Some versions of POWER8
have an issue where they edge detect and not level detect PMC overflows.

A number of places program the PMC with (0x80000000 - period_left),
where period_left can be negative. We can either fix all of these or
just ensure that period_left is always >= 1.

This patch takes the second option.

Cc: <stable@vger.kernel.org>
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Fri, 11 Jul 2014 03:37:33 +0000 (20:37 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 "Nothing too scary, we have one outstanding i915 regression but Daniel
  has promised the fix as soon as he's finished testing it a bit.

  Fixes for the main x86 drivers:
   - radeon: dpm fixes, displayport regression fix
   - i915: quirks for backlight regression, edp reboot fix, valleyview
     black screen fixes
   - nouveau: display port regression fixes, fix for memory reclocking"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/radeon/dpm: Reenabling SS on Cayman
  drm/radeon: fix typo in ci_stop_dpm()
  drm/radeon: fix typo in golden register setup on evergreen
  drm/radeon: only print meaningful VM faults
  drm/radeon/dp: return -EIO for flags not zero case
  drm/i915/vlv: T12 eDP panel timing enforcement during reboot
  drm/i915: Only unbind vgacon, not other console drivers
  drm/i915: Don't clobber the GTT when it's within stolen memory
  drm/i915/vlv: Update the DSI ULPS entry/exit sequence
  drm/i915/vlv: DPI FIFO empty check is not needed
  drm/i915: Toshiba CB35 has a controllable backlight
  drm/i915: Acer C720 and C720P have controllable backlights
  drm/i915: quirk asserts controllable backlight presence, overriding VBT
  drm/nouveau/ram: fix test for gpio presence
  drm/nouveau/dp: workaround broken display
  drm/nouveau/dp: fix required link bandwidth calculations
  drm/nouveau/kms: restore fbcon after display has been resumed
  drm/nv50-/kms: pass a non-zero value for head to sor dpms methods
  drm/nouveau/fb: Prevent inlining of ramfuc_reg
  drm/gk104/ram: bash mpll bit 31 on

10 years agopowerpc: Disable RELOCATABLE for COMPILE_TEST with PPC64
Guenter Roeck [Mon, 30 Jun 2014 18:45:30 +0000 (11:45 -0700)]
powerpc: Disable RELOCATABLE for COMPILE_TEST with PPC64

powerpc:allmodconfig has been failing for some time with the following
error.

arch/powerpc/kernel/exceptions-64s.S: Assembler messages:
arch/powerpc/kernel/exceptions-64s.S:1312: Error: attempt to move .org backwards
make[1]: *** [arch/powerpc/kernel/head_64.o] Error 1

A number of attempts to fix the problem by moving around code have been
unsuccessful and resulted in failed builds for some configurations and
the discovery of toolchain bugs.

Fix the problem by disabling RELOCATABLE for COMPILE_TEST builds instead.
While this is less than perfect, it avoids substantial code changes
which would otherwise be necessary just to make COMPILE_TEST builds
happy and might have undesired side effects.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agopowerpc/perf: Clear MMCR2 when enabling PMU
Joel Stanley [Tue, 8 Jul 2014 06:38:22 +0000 (16:08 +0930)]
powerpc/perf: Clear MMCR2 when enabling PMU

On POWER8 when switching to a KVM guest we set bits in MMCR2 to freeze
the PMU counters. Aside from on boot they are then never reset,
resulting in stuck perf counters for any user in the guest or host.

We now set MMCR2 to 0 whenever enabling the PMU, which provides a sane
state for perf to use the PMU counters under either the guest or the
host.

This was manifesting as a bug with ppc64_cpu --frequency:

    $ sudo ppc64_cpu --frequency
    WARNING: couldn't run on cpu 0
    WARNING: couldn't run on cpu 8
      ...
    WARNING: couldn't run on cpu 144
    WARNING: couldn't run on cpu 152
    min:    18446744073.710 GHz (cpu -1)
    max:    0.000 GHz (cpu -1)
    avg:    0.000 GHz

The command uses a perf counter to measure CPU cycles over a fixed
amount of time, in order to approximate the frequency of the machine.
The counters were returning zero once a guest was started, regardless of
weather it was still running or had been shut down.

By dumping the value of MMCR2, it was observed that once a guest is
running MMCR2 is set to 1s - which stops counters from running:

    $ sudo sh -c 'echo p > /proc/sysrq-trigger'
    CPU: 0 PMU registers, ppmu = POWER8 n_counters = 6
    PMC1:  5b635e38 PMC2: 00000000 PMC3: 00000000 PMC4: 00000000
    PMC5:  1bf5a646 PMC6: 5793d378 PMC7: deadbeef PMC8: deadbeef
    MMCR0: 0000000080000000 MMCR1: 000000001e000000 MMCRA: 0000040000000000
    MMCR2: fffffffffffffc00 EBBHR: 0000000000000000
    EBBRR: 0000000000000000 BESCR: 0000000000000000
    SIAR:  00000000000a51cc SDAR:  c00000000fc40000 SIER:  0000000001000000

This is done unconditionally in book3s_hv_interrupts.S upon entering the
guest, and the original value is only save/restored if the host has
indicated it was using the PMU. This is okay, however the user of the
PMU needs to ensure that it is in a defined state when it starts using
it.

Fixes: e05b9b9e5c10 ("powerpc/perf: Power8 PMU support")
Cc: stable@vger.kernel.org
Signed-off-by: Joel Stanley <joel@jms.id.au>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agopowerpc/perf: Add PPMU_ARCH_207S define
Joel Stanley [Tue, 8 Jul 2014 06:38:21 +0000 (16:08 +0930)]
powerpc/perf: Add PPMU_ARCH_207S define

Instead of separate bits for every POWER8 PMU feature, have a single one
for v2.07 of the architecture.

This saves us adding a MMCR2 define for a future patch.

Cc: stable@vger.kernel.org
Signed-off-by: Joel Stanley <joel@jms.id.au>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agopowerpc/kvm: Remove redundant save of SIER AND MMCR2
Joel Stanley [Tue, 8 Jul 2014 06:38:20 +0000 (16:08 +0930)]
powerpc/kvm: Remove redundant save of SIER AND MMCR2

These two registers are already saved in the block above. Aside from
being unnecessary, by the time we get down to the second save location
r8 no longer contains MMCR2, so we are clobbering the saved value with
PMC5.

MMCR2 primarily consists of counter freeze bits. So restoring the value
of PMC5 into MMCR2 will most likely have the effect of freezing
counters.

Fixes: 72cde5a88d37 ("KVM: PPC: Book3S HV: Save/restore host PMU registers that are new in POWER8")
Cc: stable@vger.kernel.org
Signed-off-by: Joel Stanley <joel@jms.id.au>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Acked-by: Paul Mackerras <paulus@samba.org>
Reviewed-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agopowerpc/powernv: Check for IRQHAPPENED before sleeping
Preeti U Murthy [Wed, 2 Jul 2014 03:49:35 +0000 (09:19 +0530)]
powerpc/powernv: Check for IRQHAPPENED before sleeping

Commit 8d6f7c5a: "powerpc/powernv: Make it possible to skip the IRQHAPPENED
check in power7_nap()" added code that prevents cpus from checking for
pending interrupts just before entering sleep state, which is wrong. These
interrupts are delivered during the soft irq disabled state of the cpu.

A cpu cannot enter any idle state with pending interrupts because they will
never be serviced until the next time the cpu is woken up by some other
interrupt. Its only then that the pending interrupts are replayed. This can result
in device timeouts or warnings about this cpu being stuck.

This patch fixes ths issue by ensuring that cpus check for pending interrupts
just before entering any idle state as long as they are not in the path of split
core operations.

Signed-off-by: Preeti U Murthy <preeti@linux.vnet.ibm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agopowerpc: Clean up MMU_FTRS_A2 and MMU_FTR_TYPE_3E
Michael Ellerman [Tue, 8 Jul 2014 07:10:45 +0000 (17:10 +1000)]
powerpc: Clean up MMU_FTRS_A2 and MMU_FTR_TYPE_3E

In fb5a515704d7 "powerpc: Remove platforms/wsp and associated pieces",
we removed the last user of MMU_FTRS_A2. So remove it.

MMU_FTRS_A2 was the last user of MMU_FTR_TYPE_3E, so remove it also.
This leaves some unreachable code in mmu_context_nohash.c, so remove
that also.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agopowerpc/cell: Fix compilation with CONFIG_COREDUMP=n
Michael Ellerman [Tue, 8 Jul 2014 06:00:09 +0000 (16:00 +1000)]
powerpc/cell: Fix compilation with CONFIG_COREDUMP=n

Commit 046d662f4818 "coredump: make core dump functionality optional"
made the coredump optional, but didn't update the spufs code that
depends on it. That leads to build errors such as:

  arch/powerpc/platforms/built-in.o: In function `.spufs_arch_write_note':
  coredump.c:(.text+0x22cd4): undefined reference to `.dump_emit'
  coredump.c:(.text+0x22cf4): undefined reference to `.dump_emit'
  coredump.c:(.text+0x22d0c): undefined reference to `.dump_align'
  coredump.c:(.text+0x22d48): undefined reference to `.dump_emit'
  coredump.c:(.text+0x22e7c): undefined reference to `.dump_skip'

Fix it by adding some ifdefs in the cell code.

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
10 years agoMerge branch 'rc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild
Linus Torvalds [Fri, 11 Jul 2014 01:40:47 +0000 (18:40 -0700)]
Merge branch 'rc-fixes' of git://git./linux/kernel/git/mmarek/kbuild

Pull kbuild fixes from Michal Marek:
 "Three more fixes for the relative build dir feature:

   - Shut up make -s again
   - Fix for rpm/deb/tar-pkg with O=<subdir>
   - Fix for CONFIG_EXTRA_FIRMWARE"

* 'rc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
  firmware: Create directories for external firmware
  kbuild: Fix packaging targets with relative $(srctree)
  kbuild: Do not print the build directory with make -s

10 years agoMerge tag 'pm+acpi-3.16-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Fri, 11 Jul 2014 01:36:30 +0000 (18:36 -0700)]
Merge tag 'pm+acpi-3.16-rc5' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI and power management fixes from Rafael Wysocki:
 "These are a few regression fixes for ACPI device enumeration and
  resources management, intel_pstate and cpufreq, a revert of an ACPI
  commit removing user space interfaces in /proc that we incorrectly
  thought were not used any more, fixes for some long-standing
  concurrency issues in the ACPI EC driver, two ACPI battery driver
  fixes, stable-candidate fixes for intel_pstate, an ACPI-related fix
  for i915 and two new ACPI video blacklist entries for Win8-oriented
  BIOSes.

  Specifics:

   - Missing device ID for ACPI enumeration of PNP devices that we
     overlooked during the recent rework of that code from Zhang Rui.

   - Fix for a problem introduced during the 3.14 cycle in the ACPI
     device resources management code and causing it to reject all
     resources of length 0 although some of them are actually valid
     which affects serial ports detection on a number of systems.  From
     Andy Whitcroft.

   - intel_pstate fix for a boot problem on some BayTrail-based systems
     introduced by a previous fix related to that platform during the
     3.13 cycle from Dirk Brandewie.

   - Revert of a 3.13 commit that removed the ACPI AC /proc interface
     which turns out to be still needed by some old utilities
     (kpowersave from kde 3.5.10 in particular) from Lan Tianyu.

   - cpufreq build fix for the davinci ARM platform from Prabhakar Lad
     (the breakage was introduced during the 3.10 cycle).

   - ACPI-related i915 fix preventing firmware on some Thinkpad laptops
     from setting backlight levels incorrectly during AC plug/unplug.
     From Aaron Lu.

   - Fixes for two nasty race conditions in the ACPI embedded controller
     driver that may be responsible for a number of past bug reports
     related to the EC from Lv Zhang and a fix for two memory leaks in
     error code paths in that driver from Colin Ian King.

   - Fixes for a couple of corner-case issues in the intel_pstate driver
     (all candidates for -stable) from Dirk Brandewie and Vincent Minet.

   - Fixes for two corner-case issues in the ACPI battery driver from
     Josef Gajdusek and Lan Tianyu.

   - Two new ACPI video blacklist entries for Acer TravelMate B113 and
     Dell Inspiron 5737 from Edward Lin and Martin Kepplinger"

* tag 'pm+acpi-3.16-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI / PNP: add soc_button_array device ID to PNP IDs list
  cpufreq: Makefile: fix compilation for davinci platform
  ACPI / video: Add Acer TravelMate B113 to native backlight blacklist
  ACPI / video: Add Dell Inspiron 5737 to the blacklist
  ACPI / i915: ignore firmware requests for backlight change
  ACPI / battery: fix wrong value of capacity_now reported when fully charged
  ACPI / resources: only reject zero length resources based at address zero
  ACPI / battery: Retry to get battery information if failed during probing
  ACPI / EC: Free saved_ec on error exit path
  ACPI / EC: Add detailed fields debugging support of EC_SC(R).
  ACPI / EC: Update revision due to recent changes
  ACPI / EC: Fix race condition in ec_transaction_completed()
  ACPI / EC: Remove duplicated ec_wait_ibf0() waiter
  ACPI / EC: Add asynchronous command byte write support
  ACPI / EC: Avoid race condition related to advance_transaction()
  intel_pstate: Set CPU number before accessing MSRs
  intel_pstate: Update documentation of {max,min}_perf_pct sysfs files
  intel_pstate: don't touch turbo bit if turbo disabled or unavailable.
  intel_pstate: Fix setting VID
  Revert "ACPI / AC: Remove AC's proc directory."

10 years agoMerge branch 'drm-fixes-3.16' of git://people.freedesktop.org/~agd5f/linux into drm...
Dave Airlie [Fri, 11 Jul 2014 01:24:13 +0000 (11:24 +1000)]
Merge branch 'drm-fixes-3.16' of git://people.freedesktop.org/~agd5f/linux into drm-fixes

This new request drops the pageflipping fixes for now.  Just a few small
fixes for dpm, DP, and a fix for a hang on boot evergreen.

* 'drm-fixes-3.16' of git://people.freedesktop.org/~agd5f/linux:
  drm/radeon/dpm: Reenabling SS on Cayman
  drm/radeon: fix typo in ci_stop_dpm()
  drm/radeon: fix typo in golden register setup on evergreen
  drm/radeon: only print meaningful VM faults
  drm/radeon/dp: return -EIO for flags not zero case

10 years agox86-32, vdso: Fix vDSO build error due to missing align_vdso_addr()
Jan Beulich [Thu, 3 Jul 2014 14:35:07 +0000 (15:35 +0100)]
x86-32, vdso: Fix vDSO build error due to missing align_vdso_addr()

Relying on static functions used just once to get inlined (and
subsequently have dead code paths eliminated) is wrong: Compilers are
free to decide whether they do this, regardless of optimization level.
With this not happening for vdso_addr() (observed with gcc 4.1.x), an
unresolved reference to align_vdso_addr() causes the build to fail.

[ hpa: vdso_addr() is never actually used on x86-32, as calculate_addr
  in map_vdso() is always false.  It ought to be possible to clean
  this up further, but this fixes the immediate problem. ]

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Link: http://lkml.kernel.org/r/53B5863B02000078000204D5@mail.emea.novell.com
Acked-by: Andy Lutomirski <luto@amacapital.net>
Tested-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Tested-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
10 years agox86-64, vdso: Fix vDSO build breakage due to empty .rela.dyn
Jan Beulich [Thu, 3 Jul 2014 14:34:38 +0000 (15:34 +0100)]
x86-64, vdso: Fix vDSO build breakage due to empty .rela.dyn

Certain ld versions (observed with 2.20.0) put an empty .rela.dyn
section into shared object files, breaking the assumption on the number
of sections to be copied to the final output. Simply discard any empty
SHT_REL and SHT_RELA sections to address this.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Link: http://lkml.kernel.org/r/53B5861E02000078000204D1@mail.emea.novell.com
Acked-by: Andy Lutomirski <luto@amacapital.net>
Tested-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Tested-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
10 years agodrm/radeon/dpm: Reenabling SS on Cayman
Alexandre Demers [Wed, 9 Jul 2014 02:27:36 +0000 (22:27 -0400)]
drm/radeon/dpm: Reenabling SS on Cayman

It reverts commit c745fe611ca42295c9d91d8e305d27983e9132ef now that
Cayman is stable since VDDCI fix. Spread spectrum was not the culprit.

This depends on b0880e87c1fd038b84498944f52e52c3e86ebe59
(drm/radeon/dpm: fix vddci setup typo on cayman).

Signed-off-by: Alexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agodrm/radeon: fix typo in ci_stop_dpm()
Alex Deucher [Tue, 8 Jul 2014 22:25:25 +0000 (18:25 -0400)]
drm/radeon: fix typo in ci_stop_dpm()

Need to use the RREG32_SMC() accessor since the register
is an smc indirect index.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agodrm/radeon: fix typo in golden register setup on evergreen
Alex Deucher [Mon, 7 Jul 2014 21:59:37 +0000 (17:59 -0400)]
drm/radeon: fix typo in golden register setup on evergreen

Fixes hangs on driver load on some cards.

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=76998

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agodrm/radeon: only print meaningful VM faults
Christian König [Mon, 7 Jul 2014 09:16:29 +0000 (11:16 +0200)]
drm/radeon: only print meaningful VM faults

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
10 years agodrm/radeon/dp: return -EIO for flags not zero case
Alex Deucher [Thu, 3 Jul 2014 15:17:55 +0000 (11:17 -0400)]
drm/radeon/dp: return -EIO for flags not zero case

If there are error flags in the aux transaction return
-EIO rather than -EBUSY.  -EIO restarts the whole transaction
while -EBUSY jus retries.  Fixes problematic aux transfers.

Bug:
https://bugs.freedesktop.org/show_bug.cgi?id=80684

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agodm mpath: fix IO hang due to logic bug in multipath_busy
Jun'ichi Nomura [Tue, 8 Jul 2014 00:55:14 +0000 (00:55 +0000)]
dm mpath: fix IO hang due to logic bug in multipath_busy

Commit e80991773 ("dm mpath: push back requests instead of queueing")
modified multipath_busy() to return true if !pg_ready().  pg_ready()
checks the current state of the multipath device and may return false
even if a new IO is needed to change the state.

Bart Van Assche reported that he had multipath IO lockup when he was
performing cable pull tests.  Analysis showed that the multipath
device had a single path group with both paths active, but that the
path group itself was not active.  During the multipath device state
transitions 'queue_io' got set but nothing could clear it.  Clearing
'queue_io' only happens in __choose_pgpath(), but it won't be called
if multipath_busy() returns true due to pg_ready() returning false
when 'queue_io' is set.

As such the !pg_ready() check in multipath_busy() is wrong because new
IO will not be sent to multipath target and the multipath state change
won't happen.  That results in multipath IO lockup.

The intent of multipath_busy() is to avoid unnecessary cycles of
dequeue + request_fn + requeue if it is known that the multipath
device will requeue.

Such "busy" situations would be:
  - path group is being activated
  - there is no path and the multipath is setup to requeue if no path

Fix multipath_busy() to return "busy" early only for these specific
situations.

Reported-by: Bart Van Assche <bvanassche@acm.org>
Tested-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org # v3.15
10 years agodm io: fix a race condition in the wake up code for sync_io
Joe Thornber [Fri, 27 Jun 2014 19:29:04 +0000 (15:29 -0400)]
dm io: fix a race condition in the wake up code for sync_io

There's a race condition between the atomic_dec_and_test(&io->count)
in dec_count() and the waking of the sync_io() thread.  If the thread
is spuriously woken immediately after the decrement it may exit,
making the on stack io struct invalid, yet the dec_count could still
be using it.

Fix this race by using a completion in sync_io() and dec_count().

Reported-by: Minfei Huang <huangminfei@ucloud.cn>
Signed-off-by: Joe Thornber <thornber@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Acked-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
10 years agodm crypt, dm zero: update author name following legal name change
Jana Saout [Tue, 24 Jun 2014 18:27:04 +0000 (14:27 -0400)]
dm crypt, dm zero: update author name following legal name change

Signed-off-by: Jana Saout <jana@saout.de>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
10 years agodm: allocate a special workqueue for deferred device removal
Mikulas Patocka [Sat, 14 Jun 2014 17:44:31 +0000 (13:44 -0400)]
dm: allocate a special workqueue for deferred device removal

The commit 2c140a246dc ("dm: allow remove to be deferred") introduced a
deferred removal feature for the device mapper.  When this feature is
used (by passing a flag DM_DEFERRED_REMOVE to DM_DEV_REMOVE_CMD ioctl)
and the user tries to remove a device that is currently in use, the
device will be removed automatically in the future when the last user
closes it.

Device mapper used the system workqueue to perform deferred removals.
However, some targets (dm-raid1, dm-mpath, dm-stripe) flush work items
scheduled for the system workqueue from their destructor.  If the
destructor itself is called from the system workqueue during deferred
removal, it introduces a possible deadlock - the workqueue tries to flush
itself.

Fix this possible deadlock by introducing a new workqueue for deferred
removals.  We allocate just one workqueue for all dm targets.  The
ability of dm targets to process IOs isn't dependent on deferred removal
of unused targets, so a deadlock due to shared workqueue isn't possible.

Also, cleanup local_init() to eliminate potential for returning success
on failure.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org # 3.13+
10 years agoMerge branch 'pm-cpufreq'
Rafael J. Wysocki [Thu, 10 Jul 2014 19:04:37 +0000 (21:04 +0200)]
Merge branch 'pm-cpufreq'

* pm-cpufreq:
  cpufreq: Makefile: fix compilation for davinci platform
  intel_pstate: Set CPU number before accessing MSRs
  intel_pstate: Update documentation of {max,min}_perf_pct sysfs files
  intel_pstate: don't touch turbo bit if turbo disabled or unavailable.
  intel_pstate: Fix setting VID

10 years agoMerge branch 'acpi-video'
Rafael J. Wysocki [Thu, 10 Jul 2014 19:04:25 +0000 (21:04 +0200)]
Merge branch 'acpi-video'

* acpi-video:
  ACPI / video: Add Acer TravelMate B113 to native backlight blacklist
  ACPI / video: Add Dell Inspiron 5737 to the blacklist
  ACPI / i915: ignore firmware requests for backlight change

10 years agoMerge branch 'acpi-ec'
Rafael J. Wysocki [Thu, 10 Jul 2014 19:03:57 +0000 (21:03 +0200)]
Merge branch 'acpi-ec'

* acpi-ec:
  ACPI / EC: Free saved_ec on error exit path
  ACPI / EC: Add detailed fields debugging support of EC_SC(R).
  ACPI / EC: Update revision due to recent changes
  ACPI / EC: Fix race condition in ec_transaction_completed()
  ACPI / EC: Remove duplicated ec_wait_ibf0() waiter
  ACPI / EC: Add asynchronous command byte write support
  ACPI / EC: Avoid race condition related to advance_transaction()

10 years agoMerge branches 'acpi-resources', 'acpi-ac', 'acpi-scan' and 'acpi-battery'
Rafael J. Wysocki [Thu, 10 Jul 2014 19:03:25 +0000 (21:03 +0200)]
Merge branches 'acpi-resources', 'acpi-ac', 'acpi-scan' and 'acpi-battery'

* acpi-resources:
  ACPI / resources: only reject zero length resources based at address zero

* acpi-ac:
  Revert "ACPI / AC: Remove AC's proc directory."

* acpi-scan:
  ACPI / PNP: add soc_button_array device ID to PNP IDs list

* acpi-battery:
  ACPI / battery: fix wrong value of capacity_now reported when fully charged
  ACPI / battery: Retry to get battery information if failed during probing

10 years agoMerge branch 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj...
Linus Torvalds [Thu, 10 Jul 2014 18:39:23 +0000 (11:39 -0700)]
Merge branch 'for-3.16-fixes' of git://git./linux/kernel/git/tj/libata

Pull libata fixes from Tejun Heo:
 "Nothing too interesting.  All fixes are ahci_platform related"

* 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/libata:
  ahci_xgene: fix the dma state machine lockup for the IDENTIFY DEVICE PIO mode command.
  libahci: export ahci_qc_issue() and ahci_start_fix_rx()
  ata: ahci_imx: warn when disabling ahci link
  ahci: disable ncq feature for hisilicon sata
  ahci: imx: manage only sata_ref_clk in imx_sata_enable[disable]
  libahci_platform: Fail when PHY required but PHY support disabled

10 years agoMerge branch 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj...
Linus Torvalds [Thu, 10 Jul 2014 18:38:23 +0000 (11:38 -0700)]
Merge branch 'for-3.16-fixes' of git://git./linux/kernel/git/tj/cgroup

Pull cgroup fixes from Tejun Heo:
 "Mostly fixes for the fallouts from the recent cgroup core changes.

  The decoupled nature of cgroup dynamic hierarchy management
  (hierarchies are created dynamically on mount but may or may not be
  reused once unmounted depending on remaining usages) led to more
  ugliness being added to kernfs.

  Hopefully, this is the last of it"

* 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cpuset: break kernfs active protection in cpuset_write_resmask()
  cgroup: fix a race between cgroup_mount() and cgroup_kill_sb()
  kernfs: introduce kernfs_pin_sb()
  cgroup: fix mount failure in a corner case
  cpuset,mempolicy: fix sleeping function called from invalid context
  cgroup: fix broken css_has_online_children()

10 years agoMerge branch 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq
Linus Torvalds [Thu, 10 Jul 2014 18:37:41 +0000 (11:37 -0700)]
Merge branch 'for-3.16-fixes' of git://git./linux/kernel/git/tj/wq

Pull workqueue fixes from Tejun Heo:
 "Two workqueue fixes.  Both are one liners.  One fixes missing uevent
  for workqueue files on sysfs.  The other one fixes missing zeroing of
  NUMA cpu masks which can lead to oopses among other things"

* 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
  workqueue: zero cpumask of wq_numa_possible_cpumask on init
  workqueue: fix dev_set_uevent_suppress() imbalance

10 years agoMerge branch 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj...
Linus Torvalds [Thu, 10 Jul 2014 18:35:47 +0000 (11:35 -0700)]
Merge branch 'for-3.16-fixes' of git://git./linux/kernel/git/tj/percpu

Pull percpu fix from Tejun Heo:
 "One patch to fix a typo in percpu section name.  Given how long the
  bug has been there and that there hasn't been any report of brekage,
  it's unlikely to cause actual issues"

* 'for-3.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu:
  core: fix typo in percpu read_mostly section

10 years agoMerge branch 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa...
Linus Torvalds [Thu, 10 Jul 2014 18:31:39 +0000 (11:31 -0700)]
Merge branch 'i2c/for-current' of git://git./linux/kernel/git/wsa/linux

Pull i2c build fix and cleanup from Wolfram Sang:
 "One buildfix for the i2c subsystem.  While here, I decided to add the
  header cleanup as well"

* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
  i2c: sun6i-p2wi: Remove duplicate inclusion of module.h
  i2c: mux: pca954x: fix dependencies

10 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Thu, 10 Jul 2014 18:30:57 +0000 (11:30 -0700)]
Merge git://git./linux/kernel/git/herbert/crypto-2.6

Pull crypto fixes from Herbert Xu:
 "This push fixes an error in sha512_ssse3 that leads to incorrect
  output as well as a memory leak in caam_jr when the module is
  unloaded"

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: caam - fix memleak in caam_jr module
  crypto: sha512_ssse3 - fix byte count to bit count conversion

10 years agocrypto: caam - fix memleak in caam_jr module
Cristian Stoica [Mon, 7 Jul 2014 08:52:41 +0000 (11:52 +0300)]
crypto: caam - fix memleak in caam_jr module

This patch fixes a memory leak that appears when caam_jr module is unloaded.

Cc: <stable@vger.kernel.org> # 3.13+
Signed-off-by: Cristian Stoica <cristian.stoica@freescale.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
10 years agoMerge tag 'drm-intel-fixes-2014-07-09' of git://anongit.freedesktop.org/drm-intel...
Dave Airlie [Thu, 10 Jul 2014 00:10:06 +0000 (10:10 +1000)]
Merge tag 'drm-intel-fixes-2014-07-09' of git://anongit.freedesktop.org/drm-intel into drm-fixes

Fixes for regressions and black screens, cc: stable where applicapable
(the last minute rebase was to sprinkle missing stable tags). A bit more
than what I'd wish for, but excluding vlv and that the first 3 patches are
just quirks for 1 regression it looks much better.

There's still a "oops, lost dithering" issue on older platforms open. I'm
working on a fix for that now but didn't want to delay this pile.

* tag 'drm-intel-fixes-2014-07-09' of git://anongit.freedesktop.org/drm-intel:
  drm/i915/vlv: T12 eDP panel timing enforcement during reboot
  drm/i915: Only unbind vgacon, not other console drivers
  drm/i915: Don't clobber the GTT when it's within stolen memory
  drm/i915/vlv: Update the DSI ULPS entry/exit sequence
  drm/i915/vlv: DPI FIFO empty check is not needed
  drm/i915: Toshiba CB35 has a controllable backlight
  drm/i915: Acer C720 and C720P have controllable backlights
  drm/i915: quirk asserts controllable backlight presence, overriding VBT

10 years agoMerge branch 'drm-nouveau-next' of git://anongit.freedesktop.org/git/nouveau/linux...
Dave Airlie [Thu, 10 Jul 2014 00:09:33 +0000 (10:09 +1000)]
Merge branch 'drm-nouveau-next' of git://anongit.freedesktop.org/git/nouveau/linux-2.6 into drm-fixes

A couple of DP regression fixes, kepler memory reclocking fixes, and a fix for an annoying display issue that can pop up on resume.

* 'drm-nouveau-next' of git://anongit.freedesktop.org/git/nouveau/linux-2.6:
  drm/nouveau/ram: fix test for gpio presence
  drm/nouveau/dp: workaround broken display
  drm/nouveau/dp: fix required link bandwidth calculations
  drm/nouveau/kms: restore fbcon after display has been resumed
  drm/nv50-/kms: pass a non-zero value for head to sor dpms methods
  drm/nouveau/fb: Prevent inlining of ramfuc_reg
  drm/gk104/ram: bash mpll bit 31 on

10 years agoMerge tag 'pci-v3.16-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaa...
Linus Torvalds [Wed, 9 Jul 2014 23:18:18 +0000 (16:18 -0700)]
Merge tag 'pci-v3.16-fixes-1' of git://git./linux/kernel/git/helgaas/pci

Pull PCI fixes from Bjorn Helgaas:
 "Just a fix for the device reset path and an email address update.

  Virtualization
    - Fix "wait for pending transactions" for PCI AF reset (Alex
      Williamson)

  Miscellaneous
    - Update mx6 PCI driver maintainer email (Fabio Estevam)"

* tag 'pci-v3.16-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  MAINTAINERS: Update mx6 PCI driver maintainer's email
  PCI: Fix unaligned access in AF transaction pending test

10 years agoMerge tag 'f2fs-fixes-3.16' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk...
Linus Torvalds [Wed, 9 Jul 2014 16:46:58 +0000 (09:46 -0700)]
Merge tag 'f2fs-fixes-3.16' of git://git./linux/kernel/git/jaegeuk/f2fs

Pull f2fs bugfixes from Jaegeuk Kim:
 "This includes a couple of bug fixes found by xfstests.  In addition,
  one critical bug was reported by Brian Chadwick, which is falling into
  the infinite loop in balance_dirty_pages.  And it turned out due to
  the IO merging policy in f2fs, which was newly merged in 3.16.

   - fix normal and recovery path for fallocated regions
   - fix error case mishandling
   - recover renamed fsync inodes correctly
   - fix to get out of infinite loops in balance_dirty_pages
   - fix kernel NULL pointer error"

* tag 'f2fs-fixes-3.16' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs:
  f2fs: avoid to access NULL pointer in issue_flush_thread
  f2fs: check bdi->dirty_exceeded when trying to skip data writes
  f2fs: do checkpoint for the renamed inode
  f2fs: release new entry page correctly in error path of f2fs_rename
  f2fs: fix error path in init_inode_metadata
  f2fs: check lower bound nid value in check_nid_range
  f2fs: remove unused variables in f2fs_sm_info
  f2fs: fix not to allocate unnecessary blocks during fallocate
  f2fs: recover fallocated data and its i_size together
  f2fs: fix to report newly allocate region as extent

10 years agofirmware: Create directories for external firmware
Michal Marek [Tue, 8 Jul 2014 12:38:33 +0000 (14:38 +0200)]
firmware: Create directories for external firmware

Commit 5180d5f4 ("firmware: Simplify directory creation") broke
including firmware specified in CONFIG_EXTRA_FIRMWARE:

  MK_FW   firmware/amd-ucode/microcode_amd.bin.gen.S
/bin/sh: firmware/amd-ucode/microcode_amd.bin.gen.S: No such file or directory
...
firmware/Makefile:185: recipe for target
'firmware/amd-ucode/microcode_amd.bin.gen.S' failed

It works with O= builds, because the directory is created by
Makefile.build. Create the directory in firmware/Makefile in non-O
builds.

Reported-and-tested-by: Ronald Uitermark <ronald645@gmail.com>
Reported-and-tested-by: Torsten Kaiser <just.for.lkml@googlemail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
10 years agof2fs: avoid to access NULL pointer in issue_flush_thread
Chao Yu [Mon, 7 Jul 2014 01:39:32 +0000 (09:39 +0800)]
f2fs: avoid to access NULL pointer in issue_flush_thread

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=75861

Denis 2014-05-10 11:28:59 UTC reported:
"F2FS-fs (mmcblk0p28): mounting..
 Unable to handle kernel NULL pointer dereference at virtual address 00000018
 ...
 [<c0a2f678>] (_raw_spin_lock+0x3c/0x70) from [<c03a0330>] (issue_flush_thread+0x50/0x17c)
 [<c03a0330>] (issue_flush_thread+0x50/0x17c) from [<c01b4064>] (kthread+0x98/0xa4)
 [<c01b4064>] (kthread+0x98/0xa4) from [<c0108060>] (kernel_thread_exit+0x0/0x8)"

This patch assign cmd_control_info in sm_info before issue_flush_thread is being
created, so this make sure that issue flush thread will have no chance to access
invalid info in fcc.

Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Reviewed-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agof2fs: check bdi->dirty_exceeded when trying to skip data writes
Jaegeuk Kim [Fri, 27 Jun 2014 16:00:41 +0000 (01:00 +0900)]
f2fs: check bdi->dirty_exceeded when trying to skip data writes

If we don't check the current backing device status, balance_dirty_pages can
fall into infinite pausing routine.

This can be occurred when a lot of directories make a small number of dirty
dentry pages including files.

Reported-by: Brian Chadwick <brianchad@westnet.com.au>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agof2fs: do checkpoint for the renamed inode
Jaegeuk Kim [Mon, 30 Jun 2014 09:09:55 +0000 (18:09 +0900)]
f2fs: do checkpoint for the renamed inode

If an inode is renamed, it should be registered as file_lost_pino to conduct
checkpoint at f2fs_sync_file.
Otherwise, the inode cannot be recovered due to no dent_mark in the following
scenario.

Note that, this scenario is from xfstests/322.

1. create "a"
2. fsync "a"
3. rename "a" to "b"
4. fsync "b"
5. Sudden power-cut

After recovery is done, "b" should be seen.
However, the result shows "a", since the recovery procedure does not enter
recover_dentry due to no dent_mark.

The reason is like below.
- The nid of "a" is checkpointed during #2, f2fs_sync_file.
- The inode page for "b" produced by #3 is written without dent_mark by
sync_node_pages.

So, this patch fixes this bug by assinging file_lost_pino to the "a"'s inode.
If the pino is lost, f2fs_sync_file conducts checkpoint, and then recovers
the latest pino and its dentry information for further recovery.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agof2fs: release new entry page correctly in error path of f2fs_rename
Chao Yu [Tue, 24 Jun 2014 06:13:13 +0000 (14:13 +0800)]
f2fs: release new entry page correctly in error path of f2fs_rename

This patch correct releasing code of new_page to avoid BUG_ON in error patch of
f2fs_rename.

Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agof2fs: fix error path in init_inode_metadata
Chao Yu [Tue, 24 Jun 2014 02:34:00 +0000 (10:34 +0800)]
f2fs: fix error path in init_inode_metadata

If we fail in this path:
->init_inode_metadata
  ->make_empty_dir
    ->get_new_data_page
      ->grab_cache_page return -ENOMEM

We will bug on in error path of init_inode_metadata when call remove_inode_page
because i_block = 2 (one inode block will be released later & one dentry block).

We should release the dentry block in init_inode_metadata to avoid this BUG_ON,
and avoid leak of dentry block resource, because we never have second chance to
release that block in ->evict_inode as in upper error path we make this inode
'bad'.

Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agof2fs: check lower bound nid value in check_nid_range
Chao Yu [Thu, 12 Jun 2014 05:23:41 +0000 (13:23 +0800)]
f2fs: check lower bound nid value in check_nid_range

This patch add lower bound verification for nid in check_nid_range, so nids
reserved like 0, node, meta passed by caller could be checked there.

And then check_nid_range could be used in f2fs_nfs_get_inode for simplifying
code.

Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agof2fs: remove unused variables in f2fs_sm_info
Chao Yu [Wed, 11 Jun 2014 10:32:23 +0000 (18:32 +0800)]
f2fs: remove unused variables in f2fs_sm_info

Remove unused variables in struct f2fs_sm_info.

Signed-off-by: Chao Yu <chao2.yu@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
10 years agoACPI / PNP: add soc_button_array device ID to PNP IDs list
Zhang Rui [Wed, 9 Jul 2014 12:31:04 +0000 (14:31 +0200)]
ACPI / PNP: add soc_button_array device ID to PNP IDs list

The soc_button_array PNP driver was introduced in 3.15.
But in commit eec15edbb0e1 (ACPI / PNP: use device ID list for
PNPACPI device enumeration), when reworking the PNPACPI device
enumeration, we missed the soc_button_array device ID.

This results in a regression in 3.16-rc1 that soc_button_array
pnp device fails to be enumerated.

Fix the problem by adding soc_button_array device ID into the
acpi_pnp scan handler's ID list.

Fixes: eec15edbb0e1 (ACPI / PNP: use device ID list for PNPACPI device enumeration)
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agodrm/i915/vlv: T12 eDP panel timing enforcement during reboot
Clint Taylor [Mon, 7 Jul 2014 20:01:46 +0000 (13:01 -0700)]
drm/i915/vlv: T12 eDP panel timing enforcement during reboot

The panel power sequencer on vlv doesn't appear to accept changes to its
T12 power down duration during warm reboots. This change forces a delay
for warm reboots to the T12 panel timing as defined in the VBT table for
the connected panel.

Ver2: removed redundant pr_crit(), commented magic value for pp_div_reg

Ver3: moved SYS_RESTART check earlier, new name for pp_div.

Ver4: Minor issue changes

Ver5: Move registration of reboot notifier to edp_connector_init,
      Added warning comment to handler about lack of PM notification.

Signed-off-by: Clint Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
10 years agodrm/i915: Only unbind vgacon, not other console drivers
Daniel Vetter [Tue, 8 Jul 2014 08:02:43 +0000 (10:02 +0200)]
drm/i915: Only unbind vgacon, not other console drivers

The console subsystem only provides a function to switch to a given
console, but we want to actually only switach away from vgacon.
Unconditionally switching to the dummy console resulted in switching
away from fbcon in multi-gpu setups when other gpu drivers are loaded
before i915.

Then either the reinitialization of fbcon when i915 registers its
fbdev emulation or the teardown of the fbcon driver killed the
machine. So only switch to the dummy console when it's required.

Kudos to Chris for the original idea, I've only refined it a bit to
still unregister vgacon even when it's currently unused.

This regression has been introduced in

commit a4de05268e674e8ed31df6348269e22d6c6a1803
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Jun 5 16:20:46 2014 +0200

    drm/i915: Kick out vga console

Reported-and-tested-by: Ed Tomlinson <edt@aei.ca>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: David Herrmann <dh.herrmann@gmail.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
10 years agodrm/i915: Don't clobber the GTT when it's within stolen memory
Ville Syrjälä [Thu, 5 Jun 2014 17:02:59 +0000 (20:02 +0300)]
drm/i915: Don't clobber the GTT when it's within stolen memory

On most gen2-4 platforms the GTT can be (or maybe always is?)
inside the stolen memory region. If that's the case, reduce the
size of the stolen memory appropriately to make make sure we
don't clobber the GTT.

v2: Deal with gen4 36 bit physical address

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80151
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
10 years agodrm/i915/vlv: Update the DSI ULPS entry/exit sequence
Shobhit Kumar [Thu, 3 Jul 2014 11:05:41 +0000 (16:35 +0530)]
drm/i915/vlv: Update the DSI ULPS entry/exit sequence

We should keep DEVICE_READY bit set in the ULPS enter sequence. In
exit sequence also we should set DEVICE_READY, but thats causing
blankout for me. Also exit sequence is simplified as per hw team
recommendation.

This should fix -
[drm:intel_dsi_clear_device_ready] *ERROR* DSI LP not going Low

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80818
Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com>
Tested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
10 years agodrm/i915/vlv: DPI FIFO empty check is not needed
Shobhit Kumar [Thu, 3 Jul 2014 11:05:40 +0000 (16:35 +0530)]
drm/i915/vlv: DPI FIFO empty check is not needed

While sending DPI SHUTDOWN command, we cannot wait for FIFO empty as
pipes are not disabled at that time. In case of MIPI we disable port
first and send SHUTDOWN command while pipe is still running and FIFOs
will not be empty, causing spurious error log

Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com>
Tested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
10 years agodrm/i915: Toshiba CB35 has a controllable backlight
Scot Doyle [Thu, 3 Jul 2014 23:27:52 +0000 (23:27 +0000)]
drm/i915: Toshiba CB35 has a controllable backlight

The Toshiba CB35 Chromebook (with Celeron 2955U CPU) has a controllable
backlight although its VBT reports otherwise. Apply quirk to ignore the
backlight presence check during backlight setup.

Patch tested by author on Toshiba CB35.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=79813
Signed-off-by: Scot Doyle <lkml14@scotdoyle.com>
CC: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Cc: stable@vger.kernel.org # 3.15 only
[danvet: Add cc: stable because the regressing commit is in 3.15.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
10 years agodrm/i915: Acer C720 and C720P have controllable backlights
Scot Doyle [Thu, 3 Jul 2014 23:27:51 +0000 (23:27 +0000)]
drm/i915: Acer C720 and C720P have controllable backlights

The Acer C720 and C720P Chromebooks (with Celeron 2955U CPU) have a
controllable backlight although their VBT reports otherwise. Apply quirk
to ignore the backlight presence check during backlight setup.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=79813
Tested-by: James Duley <jagduley@gmail.com>
Tested-by: Michael Mullin <masmullin@gmail.com>
Signed-off-by: Scot Doyle <lkml14@scotdoyle.com>
CC: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Cc: stable@vger.kernel.org # 3.15 only
[danvet: Add cc: stable because the regressing commit is in 3.15.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
10 years agodrm/i915: quirk asserts controllable backlight presence, overriding VBT
Scot Doyle [Thu, 3 Jul 2014 23:27:50 +0000 (23:27 +0000)]
drm/i915: quirk asserts controllable backlight presence, overriding VBT

commit c675949ec58ca50d5a3ae3c757892f1560f6e896
Author: Jani Nikula <jani.nikula@intel.com>
Date:   Wed Apr 9 11:31:37 2014 +0300

    drm/i915: do not setup backlight if not available according to VBT

caused a regression on machines with a misconfigured VBT. Add a quirk to
assert the presence of a controllable backlight. Use it to ignore the VBT
backlight presence check during backlight setup.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=79813
Tested-by: James Duley <jagduley@gmail.com>
Tested-by: Michael Mullin <masmullin@gmail.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Scot Doyle <lkml14@scotdoyle.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Cc: stable@vger.kernel.org # 3.15 only
[danvet: Add cc: stable because the regressing commit is in 3.15.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
10 years agocpufreq: Makefile: fix compilation for davinci platform
Prabhakar Lad [Tue, 8 Jul 2014 15:25:38 +0000 (16:25 +0100)]
cpufreq: Makefile: fix compilation for davinci platform

Since commtit 8a7b1227e303 (cpufreq: davinci: move cpufreq driver to
drivers/cpufreq) this added dependancy only for CONFIG_ARCH_DAVINCI_DA850
where as davinci_cpufreq_init() call is used by all davinci platform.

This patch fixes following build error:

arch/arm/mach-davinci/built-in.o: In function `davinci_init_late':
:(.init.text+0x928): undefined reference to `davinci_cpufreq_init'
make: *** [vmlinux] Error 1

Fixes: 8a7b1227e303 (cpufreq: davinci: move cpufreq driver to drivers/cpufreq)
Signed-off-by: Lad, Prabhakar <prabhakar.csengg@gmail.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 3.10+ <stable@vger.kernel.org> # 3.10+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoMerge tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck...
Linus Torvalds [Tue, 8 Jul 2014 19:08:33 +0000 (12:08 -0700)]
Merge tag 'hwmon-for-linus' of git://git./linux/kernel/git/groeck/linux-staging

Pull hwmon fixes from Guenter Roeck:
 "Fix a number of zero-day bugs, mostly found with module test scripts.
  Nothing major, but potentially annoying and worthwhile to fix"

* tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (adc128d818) Drop write support on inX_input attributes
  hwmon: (emc2103) Clamp limits instead of bailing out
  hwmon: (adm1031) Fix writes to limit registers
  hwmon: (adm1021) Fix cache problem when writing temperature limits
  hwmon: (adm1029) Ensure the fan_div cache is updated in set_fan_div
  hwmon: (amc6821) Fix permissions for temp2_input
  hwmon: (ntc_thermistor) Correct information printed during probe

10 years agoarm64: implement TASK_SIZE_OF
Colin Cross [Wed, 18 Jun 2014 20:10:09 +0000 (21:10 +0100)]
arm64: implement TASK_SIZE_OF

include/linux/sched.h implements TASK_SIZE_OF as TASK_SIZE if it
is not set by the architecture headers.  TASK_SIZE uses the
current task to determine the size of the virtual address space.
On a 64-bit kernel this will cause reading /proc/pid/pagemap of a
64-bit process from a 32-bit process to return EOF when it reads
past 0xffffffff.

Implement TASK_SIZE_OF exactly the same as TASK_SIZE with
test_tsk_thread_flag instead of test_thread_flag.

Cc: stable@vger.kernel.org
Signed-off-by: Colin Cross <ccross@android.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
10 years agoarm64: export __cpu_{clear,copy}_user_page functions
Mark Salter [Tue, 17 Jun 2014 17:14:26 +0000 (18:14 +0100)]
arm64: export __cpu_{clear,copy}_user_page functions

The __cpu_clear_user_page() and __cpu_copy_user_page() functions
are not currently exported. This prevents modules from using
clear_user_page() and copy_user_page().

Signed-off-by: Mark Salter <msalter@redhat.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
10 years agoahci_xgene: fix the dma state machine lockup for the IDENTIFY DEVICE PIO mode command.
Suman Tripathi [Mon, 7 Jul 2014 17:03:05 +0000 (22:33 +0530)]
ahci_xgene: fix the dma state machine lockup for the IDENTIFY DEVICE PIO mode command.

This patch fixes the dma state machine lockup due to the processing
of IDENTIFY DEVICE PIO mode command. The X-Gene AHCI controller
has an errata in which it cannot clear the BSY bit after the PIO setup
FIS. The dma state machine enters CMFatalErrorUpdate state and locks
up. This patch also removes the dma restart workaround from the read_id
function as the read_id function is only called by libata layer for
ATA_INTERNAL commands. But for some cases eg: PORT MULTIPLIER
and udev, the framework will enumerate using SCSI commands and it will
not call read_id function.

Signed-off-by: Loc Ho <lho@apm.com>
Signed-off-by: Suman Tripathi <stripathi@apm.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
10 years agolibahci: export ahci_qc_issue() and ahci_start_fix_rx()
Suman Tripathi [Mon, 7 Jul 2014 17:03:04 +0000 (22:33 +0530)]
libahci: export ahci_qc_issue() and ahci_start_fix_rx()

The subsequent patch will make use of them.

Signed-off-by: Loc Ho <lho@apm.com>
Signed-off-by: Suman Tripathi <stripathi@apm.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
10 years agodrm/nouveau/ram: fix test for gpio presence
Ben Skeggs [Tue, 8 Jul 2014 00:50:36 +0000 (10:50 +1000)]
drm/nouveau/ram: fix test for gpio presence

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agodrm/nouveau/dp: workaround broken display
Ben Skeggs [Wed, 2 Jul 2014 22:53:50 +0000 (08:53 +1000)]
drm/nouveau/dp: workaround broken display

The display in fdo#76483 pulses the hotplug line for link retraining
after we cut power to the main link on the source, even while it's
in D3.

fdo#76483

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agodrm/nouveau/dp: fix required link bandwidth calculations
Ben Skeggs [Tue, 1 Jul 2014 00:54:52 +0000 (10:54 +1000)]
drm/nouveau/dp: fix required link bandwidth calculations

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agodrm/nouveau/kms: restore fbcon after display has been resumed
Ben Skeggs [Mon, 30 Jun 2014 03:04:14 +0000 (13:04 +1000)]
drm/nouveau/kms: restore fbcon after display has been resumed

Under some complicated circumstances (boot, suspend, resume, attach
second display, suspend, resume, suspend, detach second display,
resume, suspend, attach second display, resume), the fb_set_suspend()
call can somehow result in a modeset being attempted before we're
ready for it and things blow up in fun ways.

Running display init first fixes the issue.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agodrm/nv50-/kms: pass a non-zero value for head to sor dpms methods
Ben Skeggs [Mon, 30 Jun 2014 01:10:02 +0000 (11:10 +1000)]
drm/nv50-/kms: pass a non-zero value for head to sor dpms methods

There's Apple machines out there which (probably completely arbitrarily)
restrict each output path to a particular head.  This causes us to not
be able to locate the output data needed to power on/off the DP output
correctly.

We fix this by passing in a head index we know is valid (as opposed to
"head 0").

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agodrm/nouveau/fb: Prevent inlining of ramfuc_reg
Stéphane Marchesin [Fri, 27 Jun 2014 20:17:25 +0000 (13:17 -0700)]
drm/nouveau/fb: Prevent inlining of ramfuc_reg

When gcc 4.8 inlines this function, it eats up 16 bytes on the stack
every time. Eventually we hit warnings because our stack grew too
much:

ramnve0.c:1383:1: error: the frame size of 1496 bytes is larger than
1024 bytes

We fix this by preventing inlining for this function.

Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agodrm/gk104/ram: bash mpll bit 31 on
Ben Skeggs [Thu, 12 Jun 2014 04:26:37 +0000 (14:26 +1000)]
drm/gk104/ram: bash mpll bit 31 on

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
10 years agoACPI / video: Add Acer TravelMate B113 to native backlight blacklist
Martin Kepplinger [Mon, 23 Jun 2014 20:30:46 +0000 (22:30 +0200)]
ACPI / video: Add Acer TravelMate B113 to native backlight blacklist

Fix backlight control for Acer TravelMate B113 Laptop by adding
it to the video_dmi_table.

A workaround before that was to use acpi_osi=Linux or
acpi_backlight=vendor on boot but even then, only the function-
keys worked.

With this change there is no need for boot parameters and DE's
controls work as well.

Signed-off-by: Martin Kepplinger <martink@posteo.de>
[rjw: Subject]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / video: Add Dell Inspiron 5737 to the blacklist
Edward Lin [Fri, 20 Jun 2014 08:13:42 +0000 (16:13 +0800)]
ACPI / video: Add Dell Inspiron 5737 to the blacklist

With win8 capabiltiy, the ACPI backlight control is broken.
The system also loses backlight setting when resuming from S3.

Add this model to the the ACPI video detect blacklist to make backlight
functionality work.

Although backlight functionality works via video.use_native_backlight=1,
this approach may be safer.

Signed-off-by: Edward Lin <yidi.lin@canonical.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / i915: ignore firmware requests for backlight change
Aaron Lu [Mon, 7 Jul 2014 07:43:51 +0000 (15:43 +0800)]
ACPI / i915: ignore firmware requests for backlight change

Some Thinkpad laptops' firmware will initiate a backlight level change
request through operation region on the events of AC plug/unplug, but
since we are not using firmware's interface to do the backlight setting
on these affected laptops, we do not want the firmware to use some
arbitrary value from its ASL variable to set the backlight level on
AC plug/unplug either.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=76491
Link: https://bugzilla.kernel.org/show_bug.cgi?id=77091
Reported-and-tested-by: Igor Gnatenko <i.gnatenko.brain@gmail.com>
Reported-and-tested-by: Anton Gubarkov <anton.gubarkov@gmail.com>
Signed-off-by: Aaron Lu <aaron.lu@intel.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / battery: fix wrong value of capacity_now reported when fully charged
Josef Gajdusek [Tue, 17 Jun 2014 21:15:49 +0000 (23:15 +0200)]
ACPI / battery: fix wrong value of capacity_now reported when fully charged

It seems that some batteries (noticed on DELL JYPJ136) assume
capacity_now = design_capacity when fully charged. This causes
reported capacity to suddenly jump to >full_charge_capacity (and that
means capacity reported to userspace is >100% and incorrect)
values after 99%. This patch detects capacity_now > full_charge_capacity,
notifies userspace (unless it is the known bug where capacity_now ==
design_capacity) and trims the value to full_charge_capacity.

Signed-off-by: Josef Gajdusek <atx@atx.name>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoMerge branch 'for-rc' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux
Linus Torvalds [Mon, 7 Jul 2014 20:23:13 +0000 (13:23 -0700)]
Merge branch 'for-rc' of git://git./linux/kernel/git/rzhang/linux

Pull thermal fixes from Zhang Rui:
 "Specifics:

   - update Email address of Thermal subsystem maintainer Eduardo
     Valentin.

   - fix a problem that unloading thermal module results in kernel crash
     because a non-exist device file is removed on thermal unload.

   - fix a problem that critical trip point is set wrongly on latest
     i.MX6 SOC and results in system critical shutdown.

   - a couple of fixes to Tmon tool, of-thermal code and ti thermal
     driver"

* 'for-rc' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux:
  tmon: set umask to a reasonable value
  tmon: Check log file for common secuirty issues
  tools/thermal: tmon: fix compilation errors when building statically
  thermal: ti-soc-thermal: ti-bandgap.c: Cleaning up wrong address is checked
  Thermal: imx: correct critical trip temperature setting
  thermal: Bind cooling devices with the correct arguments
  thermal: Add braces around suspect code
  thermal: hwmon: Make the check for critical temp valid consistent
  MAINTAINERS: Update Eduardo Valentin's email address

10 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid
Linus Torvalds [Mon, 7 Jul 2014 20:13:03 +0000 (13:13 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/jikos/hid

Pull HID fixes from Jiri Kosina:
 "A few tiny HID subsystem fixes for 3.16"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid:
  HID: use multi input quirk for 22b9:2968
  HID: sensor-hub: fix potential memory leak
  HID: usbhid: quirk for PM1610 and PM1640 Touchscreen.
  HID: rmi: Protect PM-only functions by #ifdef CONFIG_PM
  HID: sensor-hub: introduce Kconfig dependency on IOMEM
  HID: sensor-hub: make dyn_callback_lock IRQ-safe

10 years agoMerge tag 'pinctrl-v3.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw...
Linus Torvalds [Mon, 7 Jul 2014 20:11:36 +0000 (13:11 -0700)]
Merge tag 'pinctrl-v3.16-2' of git://git./linux/kernel/git/linusw/linux-pinctrl

Pull pin control fixes from Linus Walleij:
 "Two fixes for the pin control subsystem, both relating to the error
  path in probe()

  I'm a bit snowed under by mail but these have boiled in linux-next and
  should propagate to you"

* tag 'pinctrl-v3.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  pinctrl: berlin: fix an error code in berlin_pinctrl_probe()
  pinctrl: sunxi: Fix potential null pointer dereference

10 years agoACPI / resources: only reject zero length resources based at address zero
Andy Whitcroft [Thu, 19 Jun 2014 10:19:16 +0000 (11:19 +0100)]
ACPI / resources: only reject zero length resources based at address zero

The recently merged change (in v3.14-rc6) to ACPI resource detection
(below) causes all zero length ACPI resources to be elided from the
table:

  commit b355cee88e3b1a193f0e9a81db810f6f83ad728b
  Author: Zhang Rui <rui.zhang@intel.com>
  Date:   Thu Feb 27 11:37:15 2014 +0800

    ACPI / resources: ignore invalid ACPI device resources

This change has caused a regression in (at least) serial port detection
for a number of machines (see LP#1313981 [1]).  These seem to represent
their IO regions (presumably incorrectly) as a zero length region.
Reverting the above commit restores these serial devices.

Only elide zero length resources which lie at address 0.

Fixes: b355cee88e3b (ACPI / resources: ignore invalid ACPI device resources)
Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Zhang Rui <rui.zhang@intel.com>
Cc: 3.14+ <stable@vger.kernel.org> # 3.14+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agonfsd: Fix bad reserving space for encoding rdattr_error
Kinglong Mee [Sun, 6 Jul 2014 03:34:43 +0000 (11:34 +0800)]
nfsd: Fix bad reserving space for encoding rdattr_error

Introduced by commit 561f0ed498 (nfsd4: allow large readdirs).

Signed-off-by: Kinglong Mee <kinglongmee@gmail.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
10 years agohwmon: (adc128d818) Drop write support on inX_input attributes
Guenter Roeck [Mon, 7 Jul 2014 14:10:10 +0000 (07:10 -0700)]
hwmon: (adc128d818) Drop write support on inX_input attributes

Writes into input registers doesn't make sense, even more so since
the writes actually ended up writing into the maximum limit registers.
Drop it.

Cc: stable@vger.kernel.org
Reviewed-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
10 years agohwmon: (emc2103) Clamp limits instead of bailing out
Guenter Roeck [Sun, 6 Jul 2014 18:39:24 +0000 (11:39 -0700)]
hwmon: (emc2103) Clamp limits instead of bailing out

It is customary to clamp limits instead of bailing out with an error
if a configured limit is out of the range supported by the driver.
This simplifies limit configuration, since the user will not typically
know chip and/or driver specific limits.

Reviewed-by: Jean Delvare <jdelvare@suse.de>
Cc: stable@vger.kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
10 years agoworkqueue: zero cpumask of wq_numa_possible_cpumask on init
Yasuaki Ishimatsu [Mon, 7 Jul 2014 13:56:48 +0000 (09:56 -0400)]
workqueue: zero cpumask of wq_numa_possible_cpumask on init

When hot-adding and onlining CPU, kernel panic occurs, showing following
call trace.

  BUG: unable to handle kernel paging request at 0000000000001d08
  IP: [<ffffffff8114acfd>] __alloc_pages_nodemask+0x9d/0xb10
  PGD 0
  Oops: 0000 [#1] SMP
  ...
  Call Trace:
   [<ffffffff812b8745>] ? cpumask_next_and+0x35/0x50
   [<ffffffff810a3283>] ? find_busiest_group+0x113/0x8f0
   [<ffffffff81193bc9>] ? deactivate_slab+0x349/0x3c0
   [<ffffffff811926f1>] new_slab+0x91/0x300
   [<ffffffff815de95a>] __slab_alloc+0x2bb/0x482
   [<ffffffff8105bc1c>] ? copy_process.part.25+0xfc/0x14c0
   [<ffffffff810a3c78>] ? load_balance+0x218/0x890
   [<ffffffff8101a679>] ? sched_clock+0x9/0x10
   [<ffffffff81105ba9>] ? trace_clock_local+0x9/0x10
   [<ffffffff81193d1c>] kmem_cache_alloc_node+0x8c/0x200
   [<ffffffff8105bc1c>] copy_process.part.25+0xfc/0x14c0
   [<ffffffff81114d0d>] ? trace_buffer_unlock_commit+0x4d/0x60
   [<ffffffff81085a80>] ? kthread_create_on_node+0x140/0x140
   [<ffffffff8105d0ec>] do_fork+0xbc/0x360
   [<ffffffff8105d3b6>] kernel_thread+0x26/0x30
   [<ffffffff81086652>] kthreadd+0x2c2/0x300
   [<ffffffff81086390>] ? kthread_create_on_cpu+0x60/0x60
   [<ffffffff815f20ec>] ret_from_fork+0x7c/0xb0
   [<ffffffff81086390>] ? kthread_create_on_cpu+0x60/0x60

In my investigation, I found the root cause is wq_numa_possible_cpumask.
All entries of wq_numa_possible_cpumask is allocated by
alloc_cpumask_var_node(). And these entries are used without initializing.
So these entries have wrong value.

When hot-adding and onlining CPU, wq_update_unbound_numa() is called.
wq_update_unbound_numa() calls alloc_unbound_pwq(). And alloc_unbound_pwq()
calls get_unbound_pool(). In get_unbound_pool(), worker_pool->node is set
as follow:

3592         /* if cpumask is contained inside a NUMA node, we belong to that node */
3593         if (wq_numa_enabled) {
3594                 for_each_node(node) {
3595                         if (cpumask_subset(pool->attrs->cpumask,
3596                                            wq_numa_possible_cpumask[node])) {
3597                                 pool->node = node;
3598                                 break;
3599                         }
3600                 }
3601         }

But wq_numa_possible_cpumask[node] does not have correct cpumask. So, wrong
node is selected. As a result, kernel panic occurs.

By this patch, all entries of wq_numa_possible_cpumask are allocated by
zalloc_cpumask_var_node to initialize them. And the panic disappeared.

Signed-off-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: stable@vger.kernel.org
Fixes: bce903809ab3 ("workqueue: add wq_numa_tbl_len and wq_numa_possible_cpumask[]")

10 years agoHID: use multi input quirk for 22b9:2968
Wen-chien Jesse Sung [Wed, 2 Jul 2014 13:06:59 +0000 (21:06 +0800)]
HID: use multi input quirk for 22b9:2968

This device generates ABS_Z and ABS_RX events instead of ABS_X and
ABS_Y.

Signed-off-by: Wen-chien Jesse Sung <jesse.sung@canonical.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
10 years agohwmon: (adm1031) Fix writes to limit registers
Guenter Roeck [Thu, 3 Jul 2014 20:44:23 +0000 (13:44 -0700)]
hwmon: (adm1031) Fix writes to limit registers

Upper limit for write operations to temperature limit registers
was clamped to a fractional value. However, limit registers do
not support fractional values. As a result, upper limits of 127.5
degrees C or higher resulted in a rounded limit of 128 degrees C.
Since limit registers are signed, this was stored as -128 degrees C.
Clamp limits to (-55, +127) degrees C to solve the problem.

Value on writes to auto_temp[12]_min and auto_temp[12]_max were not
clamped at all, but masked. As a result, out-of-range writes resulted
in a more or less arbitrary limit. Clamp those attributes to (0, 127)
degrees C for more predictable results.

Cc: Axel Lin <axel.lin@ingics.com>
Cc: stable@vger.kernel.org
Reviewed-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
10 years agoACPI / battery: Retry to get battery information if failed during probing
Lan Tianyu [Mon, 7 Jul 2014 07:47:12 +0000 (15:47 +0800)]
ACPI / battery: Retry to get battery information if failed during probing

Some machines (eg. Lenovo Z480) ECs are not stable during boot up
and causes battery driver fails to be loaded due to failure of getting
battery information from EC sometimes. After several retries, the
operation will work. This patch is to retry to get battery information 5
times if the first try fails.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=75581
Reported-and-tested-by: naszar <naszar@ya.ru>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Lan Tianyu <tianyu.lan@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Free saved_ec on error exit path
Colin Ian King [Wed, 2 Jul 2014 23:35:09 +0000 (00:35 +0100)]
ACPI / EC: Free saved_ec on error exit path

Smatch detected two memory leaks on saved_ec:

drivers/acpi/ec.c:1070 acpi_ec_ecdt_probe() warn: possible
  memory leak of 'saved_ec'
drivers/acpi/ec.c:1109 acpi_ec_ecdt_probe() warn: possible
  memory leak of 'saved_ec'

Free saved_ec on these two error exit paths to stop the memory
leak.  Note that saved_ec maybe null, but kfree on null is allowed.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Lan Tianyu <tianyu.lan@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Add detailed fields debugging support of EC_SC(R).
Lv Zheng [Sun, 15 Jun 2014 00:42:42 +0000 (08:42 +0800)]
ACPI / EC: Add detailed fields debugging support of EC_SC(R).

Developers really don't need to translate EC_SC(R) in mind as long as the
field details are decoded in the debugging message.

Tested-by: Gareth Williams <gareth@garethwilliams.me.uk>
Tested-by: Steffen Weber <steffen.weber@gmail.com>
Tested-by: Hans de Goede <jwrdegoede@fedoraproject.org>
Tested-by: Arthur Chen <axchen@nvidia.com>
Signed-off-by: Lv Zheng <lv.zheng@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Update revision due to recent changes
Lv Zheng [Sun, 15 Jun 2014 00:42:19 +0000 (08:42 +0800)]
ACPI / EC: Update revision due to recent changes

The bug fixes and asynchronous improvements have been done to the EC driver
by the previous commits. This patch increases the revision to 2.2 to
indicate the behavior differences between the old and the new drivers. The
copyright/authorship notices are also updated.

Signed-off-by: Lv Zheng <lv.zheng@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Fix race condition in ec_transaction_completed()
Lv Zheng [Sun, 15 Jun 2014 00:42:07 +0000 (08:42 +0800)]
ACPI / EC: Fix race condition in ec_transaction_completed()

There is a race condition in ec_transaction_completed().

When ec_transaction_completed() is called in the GPE handler, it could
return true because of (ec->curr == NULL). Then the wake_up() invocation
could complete the next command unexpectedly since there is no lock between
the 2 invocations. With the previous cleanup, the IBF=0 waiter race need
not be handled any more. It's now safe to return a flag from
advance_condition() to indicate the requirement of wakeup, the flag is
returned from a locked context.

The ec_transaction_completed() is now only invoked by the ec_poll() where
the ec->curr is ensured to be different from NULL.

After cleaning up, the EVT_SCI=1 check should be moved out of the wakeup
condition so that an EVT_SCI raised with (ec->curr == NULL) can trigger a
QR_SC command.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=70891
Link: https://bugzilla.kernel.org/show_bug.cgi?id=63931
Link: https://bugzilla.kernel.org/show_bug.cgi?id=59911
Reported-and-tested-by: Gareth Williams <gareth@garethwilliams.me.uk>
Reported-and-tested-by: Hans de Goede <jwrdegoede@fedoraproject.org>
Reported-by: Barton Xu <tank.xuhan@gmail.com>
Tested-by: Steffen Weber <steffen.weber@gmail.com>
Tested-by: Arthur Chen <axchen@nvidia.com>
Signed-off-by: Lv Zheng <lv.zheng@intel.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Remove duplicated ec_wait_ibf0() waiter
Lv Zheng [Sun, 15 Jun 2014 00:41:48 +0000 (08:41 +0800)]
ACPI / EC: Remove duplicated ec_wait_ibf0() waiter

After we've added the first command byte write into advance_transaction(),
the IBF=0 waiter is duplicated with the command completion waiter
implemented in the ec_poll() because:
   If IBF=1 blocked the first command byte write invoked in the task
   context ec_poll(), it would be kicked off upon IBF=0 interrupt or timed
   out and retried again in the task context.

Remove this seperate and duplicate IBF=0 waiter.  By doing so we can
reduce the overall number of times to access the EC_SC(R) status
register.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=70891
Link: https://bugzilla.kernel.org/show_bug.cgi?id=63931
Link: https://bugzilla.kernel.org/show_bug.cgi?id=59911
Reported-and-tested-by: Gareth Williams <gareth@garethwilliams.me.uk>
Reported-and-tested-by: Hans de Goede <jwrdegoede@fedoraproject.org>
Reported-by: Barton Xu <tank.xuhan@gmail.com>
Tested-by: Steffen Weber <steffen.weber@gmail.com>
Tested-by: Arthur Chen <axchen@nvidia.com>
Signed-off-by: Lv Zheng <lv.zheng@intel.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Add asynchronous command byte write support
Lv Zheng [Sun, 15 Jun 2014 00:41:35 +0000 (08:41 +0800)]
ACPI / EC: Add asynchronous command byte write support

Move the first command byte write into advance_transaction() so that all
EC register accesses that can affect the command processing state machine
can happen in this asynchronous state machine advancement function.

The advance_transaction() function then can be a complete implementation
of an asyncrhonous transaction for a single command so that:
 1. The first command byte can be written in the interrupt context;
 2. The command completion waiter can also be used to wait the first command
    byte's timeout;
 3. In BURST mode, the follow-up command bytes can be written in the
    interrupt context directly, so that it doesn't need to return to the
    task context. Returning to the task context reduces the throughput of
    the BURST mode and in the worst cases where the system workload is very
    high, this leads to the hardware driven automatic BURST mode exit.

In order not to increase memory consumption, convert 'done' into 'flags'
to contain multiple indications:
 1. ACPI_EC_COMMAND_COMPLETE: converting from original 'done' condition,
    indicating the completion of the command transaction.
 2. ACPI_EC_COMMAND_POLL: indicating the availability of writing the first
    command byte. A new command can utilize this flag to compete for the
    right of accessing the underlying hardware. There is a follow-up bug
    fix that has utilized this new flag.

The 2 flags are important because it also reflects a key concept of IO
programs' design used in the system softwares. Normally an IO program
running in the kernel should first be implemented in the asynchronous way.
And the 2 flags are the most common way to implement its synchronous
operations on top of the asynchronous operations:
1. POLL: This flag can be used to block until the asynchronous operations
         can happen.
2. COMPLETE: This flag can be used to block until the asynchronous
             operations have completed.
By constructing code cleanly in this way, many difficult problems can be
solved smoothly.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=70891
Link: https://bugzilla.kernel.org/show_bug.cgi?id=63931
Link: https://bugzilla.kernel.org/show_bug.cgi?id=59911
Reported-and-tested-by: Gareth Williams <gareth@garethwilliams.me.uk>
Reported-and-tested-by: Hans de Goede <jwrdegoede@fedoraproject.org>
Reported-by: Barton Xu <tank.xuhan@gmail.com>
Tested-by: Steffen Weber <steffen.weber@gmail.com>
Tested-by: Arthur Chen <axchen@nvidia.com>
Signed-off-by: Lv Zheng <lv.zheng@intel.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoACPI / EC: Avoid race condition related to advance_transaction()
Lv Zheng [Sun, 15 Jun 2014 00:41:17 +0000 (08:41 +0800)]
ACPI / EC: Avoid race condition related to advance_transaction()

The advance_transaction() will be invoked from the IRQ context GPE handler
and the task context ec_poll(). The handling of this function is locked so
that the EC state machine are ensured to be advanced sequentially.

But there is a problem. Before invoking advance_transaction(), EC_SC(R) is
read. Then for advance_transaction(), there could be race condition around
the lock from both contexts. The first one reading the register could fail
this race and when it passes the stale register value to the state machine
advancement code, the hardware condition is totally different from when
the register is read. And the hardware accesses determined from the wrong
hardware status can break the EC state machine. And there could be cases
that the functionalities of the platform firmware are seriously affected.
For example:
 1. When 2 EC_DATA(W) writes compete the IBF=0, the 2nd EC_DATA(W) write may
    be invalid due to IBF=1 after the 1st EC_DATA(W) write. Then the
    hardware will either refuse to respond a next EC_SC(W) write of the next
    command or discard the current WR_EC command when it receives a EC_SC(W)
    write of the next command.
 2. When 1 EC_SC(W) write and 1 EC_DATA(W) write compete the IBF=0, the
    EC_DATA(W) write may be invalid due to IBF=1 after the EC_SC(W) write.
    The next EC_DATA(R) could never be responded by the hardware. This is
    the root cause of the reported issue.

Fix this issue by moving the EC_SC(R) access into the lock so that we can
ensure that the state machine is advanced consistently.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=70891
Link: https://bugzilla.kernel.org/show_bug.cgi?id=63931
Link: https://bugzilla.kernel.org/show_bug.cgi?id=59911
Reported-and-tested-by: Gareth Williams <gareth@garethwilliams.me.uk>
Reported-and-tested-by: Hans de Goede <jwrdegoede@fedoraproject.org>
Reported-by: Barton Xu <tank.xuhan@gmail.com>
Tested-by: Steffen Weber <steffen.weber@gmail.com>
Tested-by: Arthur Chen <axchen@nvidia.com>
Signed-off-by: Lv Zheng <lv.zheng@intel.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agointel_pstate: Set CPU number before accessing MSRs
Vincent Minet [Fri, 4 Jul 2014 23:51:33 +0000 (01:51 +0200)]
intel_pstate: Set CPU number before accessing MSRs

Ensure that cpu->cpu is set before writing MSR_IA32_PERF_CTL during CPU
initialization. Otherwise only cpu0 has its P-state set and all other
cores are left with their values unchanged.

In most cases, this is not too serious because the P-states will be set
correctly when the timer function is run.  But when the default governor
is set to performance, the per-CPU current_pstate stays the same forever
and no attempts are made to write the MSRs again.

Signed-off-by: Vincent Minet <vincent@vincent-minet.net>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agointel_pstate: Update documentation of {max,min}_perf_pct sysfs files
Dirk Brandewie [Fri, 20 Jun 2014 14:28:00 +0000 (07:28 -0700)]
intel_pstate: Update documentation of {max,min}_perf_pct sysfs files

Update documentation to make the interpretation of the values clearer

Link: https://bugzilla.kernel.org/show_bug.cgi?id=64251
Cc: 3.13+ <stable@vger.kernel.org> # 3.13+
Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agointel_pstate: don't touch turbo bit if turbo disabled or unavailable.
Dirk Brandewie [Fri, 20 Jun 2014 14:27:59 +0000 (07:27 -0700)]
intel_pstate: don't touch turbo bit if turbo disabled or unavailable.

If turbo is disabled in the BIOS bit 38 should be set in
MSR_IA32_MISC_ENABLE register per section 14.3.2.1 of the SDM Vol 3
document 325384-050US Feb 2014.  If this bit is set do *not* attempt
to disable trubo via the MSR_IA32_PERF_CTL register.  On some systems
trying to disable turbo via MSR_IA32_PERF_CTL will cause subsequent
writes to MSR_IA32_PERF_CTL not take affect, in fact reading
MSR_IA32_PERF_CTL will not show the IDA/Turbo DISENGAGE bit(32) as
set. A write of bit 32 to zero returns to normal operation.

Also deal with the case where the processor does not support
turbo and the BIOS does not report the fact in MSR_IA32_MISC_ENABLE
but does report the max and turbo P states as the same value.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=64251
Cc: 3.13+ <stable@vger.kernel.org> # 3.13+
Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agointel_pstate: Fix setting VID
Dirk Brandewie [Fri, 20 Jun 2014 14:27:58 +0000 (07:27 -0700)]
intel_pstate: Fix setting VID

Commit 21855ff5 (intel_pstate: Set turbo VID for BayTrail) introduced
setting the turbo VID which is required to prevent a machine check on
some Baytrail SKUs under heavy graphics based workloads.  The
docmumentation update that brought the requirement to light also
changed the bit mask used for enumerating P state and VID values from
0x7f to 0x3f.

This change returns the mask value to 0x7f.

Tested with the Intel NUC DN2820FYK,
BIOS version FYBYT10H.86A.0034.2014.0513.1413 with v3.16-rc1 and
v3.14.8 kernel versions.

Fixes: 21855ff5 (intel_pstate: Set turbo VID for BayTrail)
Link: https://bugzilla.kernel.org/show_bug.cgi?id=77951
Reported-and-tested-by: Rune Reterson <rune@megahurts.dk>
Reported-and-tested-by: Eric Eickmeyer <erich@ericheickmeyer.com>
Cc: 3.13+ <stable@vger.kernel.org> # 3.13+
Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoRevert "ACPI / AC: Remove AC's proc directory."
Lan Tianyu [Sun, 6 Jul 2014 23:13:46 +0000 (01:13 +0200)]
Revert "ACPI / AC: Remove AC's proc directory."

Revert commit ab0fd674d6ce (ACPI / AC: Remove AC's proc directory.),
because some old tools (e.g. kpowersave from kde 3.5.10) are still
using /proc/acpi/ac_adapter.

Fixes: ab0fd674d6ce (ACPI / AC: Remove AC's proc directory.)
Reported-and-tested-by: Sorin Manolache <sorinm@gmail.com>
Signed-off-by: Lan Tianyu <tianyu.lan@intel.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>