8 years agoce4100: fix a build error
Zhang Rui [Thu, 10 Nov 2011 13:21:10 +0000]
ce4100: fix a build error

Fix a build error. CE4100 with no serial errors because the alternate
function is only a prototype not a null function as intended.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoMerge branch 'upstream-linus' of git://github.com/jgarzik/libata-dev
Linus Torvalds [Sat, 12 Nov 2011 01:55:01 +0000]
Merge branch 'upstream-linus' of git://github.com/jgarzik/libata-dev

* 'upstream-linus' of git://github.com/jgarzik/libata-dev:
  pata_of_platform: Don't use NO_IRQ
  [libata] ahci: Add ASMedia ASM1061 support
  [libata] Issue SRST to Sil3726 PMP
  sata_sis.c: trivial spelling fix
  ahci_platform: use dev_get_platdata()
  [libata] libata-scsi.c: Add function parameter documentation

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux...
Linus Torvalds [Sat, 12 Nov 2011 01:47:06 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/mason/linux-btrfs

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  btrfs: rename the option to nospace_cache
  Btrfs: handle bio_add_page failure gracefully in scrub
  Btrfs: fix deadlock caused by the race between relocation
  Btrfs: only map pages if we know we need them when reading the space cache
  Btrfs: fix orphan backref nodes
  Btrfs: Abstract similar code for btrfs_block_rsv_add{, _noflush}
  Btrfs: fix unreleased path in btrfs_orphan_cleanup()
  Btrfs: fix no reserved space for writing out inode cache
  Btrfs: fix nocow when deleting the item
  Btrfs: tweak the delayed inode reservations again
  Btrfs: rework error handling in btrfs_mount()
  Btrfs: close devices on all error paths in open_ctree()
  Btrfs: avoid null dereference and leaks when bailing from open_ctree()
  Btrfs: fix subvol_name leak on error in btrfs_mount()
  Btrfs: fix memory leak in btrfs_parse_early_options()
  Btrfs: fix our reservations for updating an inode when completing io
  Btrfs: fix oops on NULL trans handle in btrfs_truncate
  btrfs: fix double-free 'tree_root' in 'btrfs_mount()'

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Linus Torvalds [Sat, 12 Nov 2011 01:45:42 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: usb-audio: Use kmemdup rather than duplicating its implementation
  ALSA: hda - Re-enable the check NO_PRESENCE misc bit
  ALSA: vmaster - Free slave-links when freeing the master element
  ALSA: hda - Don't add elements of other codecs to vmaster slave
  ALSA: intel8x0: improve virtual environment detection
  ALSA: intel8x0: move virtual environment detection code into one place
  ALSA: snd_usb_audio: add Logitech HD Webcam c510 to quirk-384
  ALSA: hda - fix internal mic on Dell Vostro 3500 laptop
  ALSA: HDA: Remove quirk for Toshiba T110
  ALSA: usb-audio - Fix the missing volume quirks at delayed init
  ALSA: hda - Mute unused capture sources for Realtek codecs
  ALSA: intel8x0: Improve comments for VM optimization
  ASoC: Ensure we get an impedence reported for WM8958 jack detect
  ASoC: Don't use wm8994->control_data when requesting IRQs
  ASoC: Don't use wm8994->control_data in wm8994_readable_register()
  ASoC: Update git repository URL

8 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Sat, 12 Nov 2011 01:43:51 +0000]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (42 commits)
  drm/radeon/kms/pm: switch to dynamically allocating clock mode array
  drm/radeon/kms: optimize r600_pm_profile_init
  drm/radeon/kms/pm: add a proper pm profile init function for fusion
  drm/radeon/kms: remove extraneous calls to radeon_pm_compute_clocks()
  drm/exynos: added padding to be 64-bit align.
  drm: fix kconfig unmet dependency warning
  drm: add some comments to drm_wait_vblank and drm_queue_vblank_event
  drm/radeon/benchmark: signedness bug in radeon_benchmark_move()
  drm: do not sleep on vblank while holding a mutex
  MAINTAINERS: exynos: Add EXYNOS DRM maintainer entry
  drm: try to restore previous CRTC config if mode set fails
  drm/radeon/kms: make an aux failure debug only
  drm: drop select of SLOW_WORK
  drm: serialize access to list of debugfs files
  drm/radeon/kms: fix use of vram scratch page on evergreen/ni
  drm/radeon: Make sure CS mutex is held across GPU reset.
  drm: Ensure string is null terminated.
  vmwgfx: Only allow 64x64 cursors
  vmwgfx: Initialize clip rect loop correctly in surface dirty
  vmwgfx: Close screen object system
  ...

8 years agoMerge git://github.com/herbertx/crypto
Linus Torvalds [Sat, 12 Nov 2011 01:40:02 +0000]
Merge git://github.com/herbertx/crypto

* git://github.com/herbertx/crypto:
  crypto: algapi - Fix build problem with NET disabled
  crypto: user - Fix rwsem leak in crypto_user

8 years agoMerge branch 'for-linus' of git://oss.sgi.com/xfs/xfs
Linus Torvalds [Sat, 12 Nov 2011 01:37:17 +0000]
Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs

* 'for-linus' of git://oss.sgi.com/xfs/xfs:
  xfs: fix force shutdown handling in xfs_end_io
  xfs: constify xfs_item_ops
  xfs: Fix possible memory corruption in xfs_readlink

8 years agobtrfs: rename the option to nospace_cache
David Sterba [Fri, 11 Nov 2011 15:14:57 +0000]
btrfs: rename the option to nospace_cache

Rename no_space_cache option to nospace_cache to be more consistent with
the rest, where the simple prefix 'no' is used to negate an option.

The option has been introduced during the -rc1 cycle and there are has not been
widely used, so it's safe.

Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agodrm/radeon/kms/pm: switch to dynamically allocating clock mode array
Alex Deucher [Fri, 4 Nov 2011 14:09:43 +0000]
drm/radeon/kms/pm: switch to dynamically allocating clock mode array

On newer chips the number of clock modes per power state varies.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon/kms: optimize r600_pm_profile_init
Alex Deucher [Fri, 4 Nov 2011 14:09:42 +0000]
drm/radeon/kms: optimize r600_pm_profile_init

Avoid a lot of extra loops through the pm state array.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon/kms/pm: add a proper pm profile init function for fusion
Alex Deucher [Fri, 4 Nov 2011 14:09:41 +0000]
drm/radeon/kms/pm: add a proper pm profile init function for fusion

The new power tables need to be handled differently when setting
up the profiles.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon/kms: remove extraneous calls to radeon_pm_compute_clocks()
Alex Deucher [Fri, 4 Nov 2011 14:09:40 +0000]
drm/radeon/kms: remove extraneous calls to radeon_pm_compute_clocks()

It's already called via the DPMS functions.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agoBtrfs: handle bio_add_page failure gracefully in scrub
Arne Jansen [Fri, 11 Nov 2011 13:17:10 +0000]
Btrfs: handle bio_add_page failure gracefully in scrub

Currently scrub fails with ENOMEM when bio_add_page fails. Unfortunately
dm based targets accept only one page per bio, thus making scrub always
fails. This patch just submits the current bio when an error is encountered
and starts a new one.

Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agodrm/exynos: added padding to be 64-bit align.
Inki Dae [Fri, 4 Nov 2011 12:09:50 +0000]
drm/exynos: added padding to be 64-bit align.

Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm: fix kconfig unmet dependency warning
Randy Dunlap [Mon, 31 Oct 2011 19:51:30 +0000]
drm: fix kconfig unmet dependency warning

Fix kconfig unmet dependency warning.  BACKLIGHT_CLASS_DEVICE depends on
BACKLIGHT_LCD_SUPPORT, so select the latter along with the former.

warning: (DRM_RADEON_KMS && DRM_I915 && STUB_POULSBO && FB_BACKLIGHT && PANEL_SHARP_LS037V7DW01 && PANEL_ACX565AKM && USB_APPLEDISPLAY && FB_OLPC_DCON && ASUS_LAPTOP && SONY_LAPTOP && THINKPAD_ACPI && EEEPC_LAPTOP && ACPI_ASUS && ACPI_CMPC && SAMSUNG_Q10) selects BACKLIGHT_CLASS_DEVICE which has unmet direct dependencies (HAS_IOMEM && BACKLIGHT_LCD_SUPPORT)

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Cc: David Airlie <airlied@linux.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agoMerge branch 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2...
Dave Airlie [Fri, 11 Nov 2011 11:17:43 +0000]
Merge branch 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2.6 into drm-fixes

* 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2.6:
  drm/nouveau: Fix bandwidth calculation for DisplayPort
  drm/nouveau: by default use low bpp framebuffer on low memory cards
  drm/nv10: Change the BO size threshold determining the memory placement range.
  drm/nvc0: enable acceleration for nvc1 by default
  drm/nvc0/gr: fixup the mmio list register writes for 0xc1
  drm/nvc1: hacky workaround to fix accel issues
  drm/nvc0/gr: fix some bugs in grctx generation
  drm/nvc0: enable acceleration on 0xc8 by default
  drm/nvc0/vram: skip disabled PBFB subunits
  drm/nv40/pm: fix issues on igp chipsets, which don't have memory
  drm/nouveau: testing the wrong variable
  drm/nvc0/vram: storage type 0xc3 is not compressed
  drm/nv50: fix stability issue on NV86.
  drm/nouveau: initialize chan->fence.lock before use
  drm/nv50/vram: fix incorrect detection of bank count on newer chipsets
  drm/nv50/gr: typo fix, how about we not reset fifo during graph init?
  drm/nv50/bios: fixup mpll programming from the init table parser
  drm/nouveau: fix oops if i2c bus not found in nouveau_i2c_identify()
  drm: make sure drm_vblank_init() has been called before touching vbl_lock

8 years agodrm: add some comments to drm_wait_vblank and drm_queue_vblank_event
Ilija Hadzic [Mon, 31 Oct 2011 17:11:57 +0000]
drm: add some comments to drm_wait_vblank and drm_queue_vblank_event

during the review of the fix for locks problems in drm_wait_vblank,
a couple of false concerns were raised about how the drm_vblank_get
and drm_vblank_put are used in this function; it turned out that the
code is correct and that it cannot be simplified

add a few comments to explain non-obvious flows in the code,
to prevent "false alarms" in the future

v2: incorporate comments received from Daniel Vetter

Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon/benchmark: signedness bug in radeon_benchmark_move()
Dan Carpenter [Sat, 29 Oct 2011 07:21:28 +0000]
drm/radeon/benchmark: signedness bug in radeon_benchmark_move()

radeon_benchmark_do_move() returns an int so "time" should be int
too.  Making it unsigned breaks the error handling.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm: do not sleep on vblank while holding a mutex
Ilija Hadzic [Mon, 31 Oct 2011 21:46:18 +0000]
drm: do not sleep on vblank while holding a mutex

drm_wait_vblank must be DRM_UNLOCKED because otherwise it
will grab the drm_global_mutex and then go to sleep until the vblank
event it is waiting for. That can wreck havoc in the windowing system
because if one process issues this ioctl, it will block all other
processes for the duration of all vblanks between the current and the
one it is waiting for. In some cases it can block the entire windowing
system.

v2: incorporate comments received from Daniel Vetter and
    Michel Daenzer.

v3/v4: after a lengty discussion with Daniel Vetter, it was concluded
       that the only thing not yet protected with locks and atomic
       ops is the write to dev->last_vblank_wait. It's only used in a
       debug file in proc, and the current code already employs no
       correct locking: the proc file only takes dev->struct_mutex,
       whereas drm_wait_vblank implicitly took the drm_global_mutex.
       Given all this, it's not worth bothering to try to fix
       the locks at this time.

Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agoMAINTAINERS: exynos: Add EXYNOS DRM maintainer entry
Kyungmin Park [Wed, 2 Nov 2011 02:33:16 +0000]
MAINTAINERS: exynos: Add EXYNOS DRM maintainer entry

As Exynos DRM is merged at mainline. Also update the maintainer entry.

Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm: try to restore previous CRTC config if mode set fails
Jesse Barnes [Mon, 7 Nov 2011 18:39:57 +0000]
drm: try to restore previous CRTC config if mode set fails

We restore the CRTC, encoder, and connector configurations, but if the
mode set failed, the attached display may have been turned off, so we
need to try set_config again to restore things to the way they were.

Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon/kms: make an aux failure debug only
Alex Deucher [Tue, 8 Nov 2011 15:09:58 +0000]
drm/radeon/kms: make an aux failure debug only

Can happen when there is no DP panel attached, confusing
users.  Make it debug only.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm: drop select of SLOW_WORK
Paul Bolle [Wed, 9 Nov 2011 00:16:50 +0000]
drm: drop select of SLOW_WORK

slow-work got killed in commit 181a51f6e0. This means that since v2.6.36
there is no Kconfig symbol SLOW_WORK. Apparently selecting that symbol
is a nop. Drop that select.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm: serialize access to list of debugfs files
Marcin Slusarz [Wed, 9 Nov 2011 21:20:35 +0000]
drm: serialize access to list of debugfs files

Nouveau, when configured with debugfs, creates debugfs files for every
channel, so structure holding list of files needs to be protected from
simultaneous changes by multiple threads.

Without this patch it's possible to hit kernel oops in
drm_debugfs_remove_files just by running a couple of xterms with
looped glxinfo.

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon/kms: fix use of vram scratch page on evergreen/ni
Alex Deucher [Thu, 10 Nov 2011 13:59:39 +0000]
drm/radeon/kms: fix use of vram scratch page on evergreen/ni

This hunk seems to have gotten lost when I rebased the patch.

Reported-by: Sylvain Bertrand <sylvain.bertrand@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm/radeon: Make sure CS mutex is held across GPU reset.
Michel Dänzer [Thu, 10 Nov 2011 17:57:26 +0000]
drm/radeon: Make sure CS mutex is held across GPU reset.

This was only the case if the GPU reset was triggered from the CS ioctl,
otherwise other processes could happily enter the CS ioctl and wreak havoc
during the GPU reset.

This is a little complicated because the GPU reset can be triggered from the
CS ioctl, in which case we're already holding the mutex, or from other call
paths, in which case we need to lock the mutex. AFAICT the mutex API doesn't
allow recursive locking or finding out the mutex owner, so we need to handle
this with helper functions which allow recursive locking from the same
process.

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agodrm: Ensure string is null terminated.
Vinson Lee [Thu, 10 Nov 2011 19:55:40 +0000]
drm: Ensure string is null terminated.

Fixes Coverity buffer not null terminated defect.

Signed-off-by: Vinson Lee <vlee@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agovmwgfx: Only allow 64x64 cursors
Jakob Bornecrantz [Wed, 9 Nov 2011 09:25:28 +0000]
vmwgfx: Only allow 64x64 cursors

Snooping code expects this to be the case.

Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agovmwgfx: Initialize clip rect loop correctly in surface dirty
Jakob Bornecrantz [Wed, 9 Nov 2011 09:25:27 +0000]
vmwgfx: Initialize clip rect loop correctly in surface dirty

Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agovmwgfx: Close screen object system
Jakob Bornecrantz [Wed, 9 Nov 2011 09:25:26 +0000]
vmwgfx: Close screen object system

Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

8 years agoBtrfs: fix deadlock caused by the race between relocation
Miao Xie [Fri, 11 Nov 2011 01:45:05 +0000]
Btrfs: fix deadlock caused by the race between relocation

We can not do flushable reservation for the relocation when we create snapshot,
because it may make the transaction commit task and the flush task wait for
each other and the deadlock happens.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: only map pages if we know we need them when reading the space cache
Josef Bacik [Fri, 11 Nov 2011 01:45:05 +0000]
Btrfs: only map pages if we know we need them when reading the space cache

People have been running into a warning when loading space cache because the
page is already mapped when trying to read in a bitmap.  The way we read in
entries and pages is kind of convoluted, so fix it so that io_ctl_read_entry
maps the entries if it needs to, and if it hits the end of the page it simply
unmaps the page.  That way we can unconditionally unmap the io_ctl before
reading in the bitmap and we should stop hitting these warnings.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: fix orphan backref nodes
Miao Xie [Fri, 11 Nov 2011 01:45:05 +0000]
Btrfs: fix orphan backref nodes

If the root node of a fs/file tree is in the block group that is
being relocated, but the others are not in the other block groups.
when we create a snapshot for this tree between the relocation tree
creation ends and ->create_reloc_tree is set to 0, Btrfs will create
some backref nodes that are the lowest nodes of the backrefs cache.
But we forget to add them into ->leaves list of the backref cache
and deal with them, and at last, they will triggered BUG_ON().

  kernel BUG at fs/btrfs/relocation.c:239!

This patch fixes it by adding them into ->leaves list of backref cache.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: Abstract similar code for btrfs_block_rsv_add{, _noflush}
Miao Xie [Fri, 11 Nov 2011 01:45:05 +0000]
Btrfs: Abstract similar code for btrfs_block_rsv_add{, _noflush}

btrfs_block_rsv_add{, _noflush}() have similar code, so abstract that code.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: fix unreleased path in btrfs_orphan_cleanup()
Miao Xie [Fri, 11 Nov 2011 01:45:05 +0000]
Btrfs: fix unreleased path in btrfs_orphan_cleanup()

When we did stress test for the space relocation, the deadlock happened.
By debugging, We found it was caused by the carelessness that we forgot
to unlock the read lock of the extent buffers in btrfs_orphan_cleanup()
before we end the transaction handle, so the transaction commit task waited
the task, which called btrfs_orphan_cleanup(), to unlock the extent buffer,
but that task waited the commit task to end the transaction commit, and
the deadlock happened. Fix it.

Signed-ff-by: Miao Xie <miaox@cn.fujitsu.com>

Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: fix no reserved space for writing out inode cache
Miao Xie [Fri, 11 Nov 2011 01:45:04 +0000]
Btrfs: fix no reserved space for writing out inode cache

I-node cache forgets to reserve the space when writing out it. And when
we do some stress test, such as synctest, it will trigger WARN_ON() in
use_block_rsv().

WARNING: at fs/btrfs/extent-tree.c:5718 btrfs_alloc_free_block+0xbf/0x281 [btrfs]()
...
Call Trace:
 [<ffffffff8104df86>] warn_slowpath_common+0x80/0x98
 [<ffffffff8104dfb3>] warn_slowpath_null+0x15/0x17
 [<ffffffffa0369c60>] btrfs_alloc_free_block+0xbf/0x281 [btrfs]
 [<ffffffff810cbcb8>] ? __set_page_dirty_nobuffers+0xfe/0x108
 [<ffffffffa035c040>] __btrfs_cow_block+0x118/0x3b5 [btrfs]
 [<ffffffffa035c7ba>] btrfs_cow_block+0x103/0x14e [btrfs]
 [<ffffffffa035e4c4>] btrfs_search_slot+0x249/0x6a4 [btrfs]
 [<ffffffffa036d086>] btrfs_lookup_inode+0x2a/0x8a [btrfs]
 [<ffffffffa03788b7>] btrfs_update_inode+0xaa/0x141 [btrfs]
 [<ffffffffa036d7ec>] btrfs_save_ino_cache+0xea/0x202 [btrfs]
 [<ffffffffa03a761e>] ? btrfs_update_reloc_root+0x17e/0x197 [btrfs]
 [<ffffffffa0373867>] commit_fs_roots+0xaa/0x158 [btrfs]
 [<ffffffffa03746a6>] btrfs_commit_transaction+0x405/0x731 [btrfs]
 [<ffffffff810690df>] ? wake_up_bit+0x25/0x25
 [<ffffffffa039d652>] ? btrfs_log_dentry_safe+0x43/0x51 [btrfs]
 [<ffffffffa0381c5f>] btrfs_sync_file+0x16a/0x198 [btrfs]
 [<ffffffff81122806>] ? mntput+0x21/0x23
 [<ffffffff8112d150>] vfs_fsync_range+0x18/0x21
 [<ffffffff8112d170>] vfs_fsync+0x17/0x19
 [<ffffffff8112d316>] do_fsync+0x29/0x3e
 [<ffffffff8112d348>] sys_fsync+0xb/0xf
 [<ffffffff81468352>] system_call_fastpath+0x16/0x1b

Sometimes it causes BUG_ON() in the reservation code of the delayed inode
is triggered.

So we must reserve enough space for inode cache.

Note: If we can not reserve the enough space for inode cache, we will
give up writing out it.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: fix nocow when deleting the item
Miao Xie [Fri, 11 Nov 2011 01:45:04 +0000]
Btrfs: fix nocow when deleting the item

btrfs_previous_item() just search the b+ tree, do not COW the nodes or leaves,
if we modify the result of it, the meta-data will be broken. fix it.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoMerge branch 'mount-fixes' of git://github.com/idryomov/btrfs-unstable into integration
Chris Mason [Fri, 11 Nov 2011 01:42:53 +0000]
Merge branch 'mount-fixes' of git://github.com/idryomov/btrfs-unstable into integration

8 years agoBtrfs: tweak the delayed inode reservations again
Chris Mason [Fri, 11 Nov 2011 01:39:08 +0000]
Btrfs: tweak the delayed inode reservations again

Josef sent along an incremental to the inode reservation
code to make sure we try and fall back to directly updating
the inode item if things go horribly wrong.

This reworks that patch slightly, adding a fallback function
that will always try to update the inode item directly without
going through the delayed_inode code.

Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agocrypto: algapi - Fix build problem with NET disabled
Herbert Xu [Thu, 3 Nov 2011 12:46:07 +0000]
crypto: algapi - Fix build problem with NET disabled

The report functions use NLA_PUT so we need to ensure that NET
is enabled.

Reported-by: Luis Henriques <henrix@camandro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

8 years agopata_of_platform: Don't use NO_IRQ
Anton Vorontsov [Thu, 10 Nov 2011 16:28:59 +0000]
pata_of_platform: Don't use NO_IRQ

Drivers should not use NO_IRQ; moreover, some architectures don't
have it nowadays. '0' is the 'no irq' case.

Signed-off-by: Anton Vorontsov <cbouatmailru@gmail.com>
Acked-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agoMerge branch 'fix/misc' into for-linus
Takashi Iwai [Thu, 10 Nov 2011 18:52:08 +0000]
Merge branch 'fix/misc' into for-linus

8 years agoALSA: usb-audio: Use kmemdup rather than duplicating its implementation
Thomas Meyer [Thu, 10 Nov 2011 18:38:43 +0000]
ALSA: usb-audio: Use kmemdup rather than duplicating its implementation

Use kmemdup rather than duplicating its implementation

The semantic patch that makes this change is available
in scripts/coccinelle/api/memdup.cocci.

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: hda - Re-enable the check NO_PRESENCE misc bit
Takashi Iwai [Thu, 10 Nov 2011 11:36:46 +0000]
ALSA: hda - Re-enable the check NO_PRESENCE misc bit

We disabled the check of NO_PRESENCE bit of the default pin-config
in commit f4419172 temporarily.  One problem was that the first
implementation was wrong -- the bit after the shift must be checked.
However, this would still give many regressions on machines with broken
BIOS.  They set this bit wrongly even on active pins.

A workaround is to check whether all pins contain this bit.  As far as
I've checked, broken BIOSen set this bit on all pins, no matter whether
active or not.  In such a case, the driver should ignore this bit check.

Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: vmaster - Free slave-links when freeing the master element
Takashi Iwai [Thu, 10 Nov 2011 11:34:24 +0000]
ALSA: vmaster - Free slave-links when freeing the master element

When freeing the vmaster master element, we should release slave-links
properly, not only assumig that slaves will be freed soon later.

Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: hda - Don't add elements of other codecs to vmaster slave
Takashi Iwai [Thu, 10 Nov 2011 11:28:38 +0000]
ALSA: hda - Don't add elements of other codecs to vmaster slave

When a virtual mater control is created, the driver looks for slave
elements from the assigned card instance.  But this may include the
elements of other codecs when multiple codecs are on the same HD-audio
bus.  This works at the first time, but it'll give Oops when it's once
freed and re-created via reconfig sysfs.

This patch changes the element-look-up strategy to limit only to the
mixer elements of the same codec.

Reported-by: David Henningsson <david.henningsson@canonical.com>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoRevert "proc: fix races against execve() of /proc/PID/fd**"
Linus Torvalds [Wed, 9 Nov 2011 23:16:00 +0000]
Revert "proc: fix races against execve() of /proc/PID/fd**"

This reverts commit aa6afca5bcaba8101f3ea09d5c3e4100b2b9f0e5.

It escalates of some of the google-chrome SELinux problems with ptrace
("Check failed: pid_ > 0.  Did not find zygote process"), and Andrew
says that it is also causing mystery lockdep reports.

Reported-by: Alex Villacís Lasso <a_villacis@palosanto.com>
Requested-by: James Morris <jmorris@namei.org>
Requested-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agodrm/nouveau: Fix bandwidth calculation for DisplayPort
Adam Jackson [Tue, 25 Oct 2011 17:09:43 +0000]
drm/nouveau: Fix bandwidth calculation for DisplayPort

Ported from the equivalent fix in drm-intel-next:

http://cgit.freedesktop.org/~keithp/linux/commit/?h=drm-intel-next&id=cd9dde44f47501394b9f0715b6a36a92aa74c0d0

Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nouveau: by default use low bpp framebuffer on low memory cards
Marcin Slusarz [Sun, 6 Nov 2011 19:32:04 +0000]
drm/nouveau: by default use low bpp framebuffer on low memory cards

Framebuffer's BPP is not that important but can waste significant part
of memory on low-VRAM cards. Lower it to 8bpp on < 32MB cards and to
16bpp on 64MB cards. It can still be overridden by video= option.

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nv10: Change the BO size threshold determining the memory placement range.
Francisco Jerez [Sun, 6 Nov 2011 20:21:28 +0000]
drm/nv10: Change the BO size threshold determining the memory placement range.

Fixes the framebuffer memory allocation failure seen on some
low-memory cards, followed by X refusing to start.

https://bugs.freedesktop.org/show_bug.cgi?id=42384

Reported-by: Chris Paulson-Ellis <chris@edesix.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc0: enable acceleration for nvc1 by default
Ben Skeggs [Fri, 28 Oct 2011 01:00:39 +0000]
drm/nvc0: enable acceleration for nvc1 by default

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc0/gr: fixup the mmio list register writes for 0xc1
Ben Skeggs [Fri, 28 Oct 2011 01:43:04 +0000]
drm/nvc0/gr: fixup the mmio list register writes for 0xc1

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc1: hacky workaround to fix accel issues
Ben Skeggs [Fri, 28 Oct 2011 00:59:45 +0000]
drm/nvc1: hacky workaround to fix accel issues

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc0/gr: fix some bugs in grctx generation
Ben Skeggs [Mon, 11 Jul 2011 05:40:43 +0000]
drm/nvc0/gr: fix some bugs in grctx generation

Most serious is for chips with only 1 TPC, we'd get stuck in an infinite
loop.  The fix here will slightly change the setup for all other chipsets
too, but, it shouldn't matter too much, and this all needs figuring out
and likely redone anyway.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc0: enable acceleration on 0xc8 by default
Ben Skeggs [Thu, 27 Oct 2011 01:31:49 +0000]
drm/nvc0: enable acceleration on 0xc8 by default

Worked well enough for glxgears and gnome-shell at least, no reason to
have this off anymore.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc0/vram: skip disabled PBFB subunits
Ben Skeggs [Thu, 27 Oct 2011 01:26:17 +0000]
drm/nvc0/vram: skip disabled PBFB subunits

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nv40/pm: fix issues on igp chipsets, which don't have memory
Ben Skeggs [Wed, 19 Oct 2011 04:06:59 +0000]
drm/nv40/pm: fix issues on igp chipsets, which don't have memory

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nouveau: testing the wrong variable
Dan Carpenter [Tue, 18 Oct 2011 06:07:51 +0000]
drm/nouveau: testing the wrong variable

memtimings is a valid pointer here, the intent was to test for
kcalloc() failure.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nvc0/vram: storage type 0xc3 is not compressed
Christoph Bumiller [Fri, 7 Oct 2011 16:10:45 +0000]
drm/nvc0/vram: storage type 0xc3 is not compressed

Signed-off-by: Christoph Bumiller <e0425955@student.tuwien.ac.at>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nv50: fix stability issue on NV86.
Maxim Levitsky [Sun, 9 Oct 2011 20:58:31 +0000]
drm/nv50: fix stability issue on NV86.

Confirmed to fix random hangs while running all Unegine demos on NV86.

Signed-off-by: Maxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nouveau: initialize chan->fence.lock before use
Marcin Slusarz [Fri, 9 Sep 2011 12:16:42 +0000]
drm/nouveau: initialize chan->fence.lock before use

Fence lock needs to be initialized before any call to nouveau_channel_put
because it calls nouveau_channel_idle->nouveau_fence_update which uses
fence lock.

BUG: spinlock bad magic on CPU#0, test/24134
 lock: ffff88019f90dba8, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
Pid: 24134, comm: test Not tainted 3.0.0-nv+ #800
Call Trace:
 spin_bug+0x9c/0xa3
 do_raw_spin_lock+0x29/0x13c
 _raw_spin_lock+0x1e/0x22
 nouveau_fence_update+0x2d/0xf1
 nouveau_channel_idle+0x22/0xa0
 nouveau_channel_put_unlocked+0x84/0x1bd
 nouveau_channel_put+0x20/0x24
 nouveau_channel_alloc+0x4ec/0x585
 nouveau_ioctl_fifo_alloc+0x50/0x130
 drm_ioctl+0x289/0x361
 do_vfs_ioctl+0x4dd/0x52c
 sys_ioctl+0x42/0x65
 system_call_fastpath+0x16/0x1b

It's easily triggerable from userspace.

Additionally remove double initialization of chan->fence.pending.

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nv50/vram: fix incorrect detection of bank count on newer chipsets
Ben Skeggs [Fri, 7 Oct 2011 06:00:31 +0000]
drm/nv50/vram: fix incorrect detection of bank count on newer chipsets

NVA3+ has an extra bit here compared to NV50:NVA3 chipsets.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nv50/gr: typo fix, how about we not reset fifo during graph init?
Ben Skeggs [Thu, 29 Sep 2011 03:15:17 +0000]
drm/nv50/gr: typo fix, how about we not reset fifo during graph init?

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nv50/bios: fixup mpll programming from the init table parser
Ben Skeggs [Fri, 23 Sep 2011 05:37:38 +0000]
drm/nv50/bios: fixup mpll programming from the init table parser

Reportedly this has been causing stability and corruption issues after
resuming from suspend for a few people.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm/nouveau: fix oops if i2c bus not found in nouveau_i2c_identify()
Ben Skeggs [Thu, 22 Sep 2011 02:51:31 +0000]
drm/nouveau: fix oops if i2c bus not found in nouveau_i2c_identify()

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agodrm: make sure drm_vblank_init() has been called before touching vbl_lock
Ben Skeggs [Mon, 4 Jul 2011 02:52:27 +0000]
drm: make sure drm_vblank_init() has been called before touching vbl_lock

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

8 years agoBtrfs: rework error handling in btrfs_mount()
Ilya Dryomov [Wed, 9 Nov 2011 12:41:22 +0000]
Btrfs: rework error handling in btrfs_mount()

Commits 6c41761f and 45ea6095 introduced the possibility of NULL pointer
dereference on error paths, also we would leave all devices busy and
leak fs_info with all sub-structures on error when trying to mount an
already mounted fs to a different directory.

Fix this by doing all allocations before trying to open any of the
devices, adjust error path for mount-already-mounted-fs case.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

8 years agoBtrfs: close devices on all error paths in open_ctree()
Ilya Dryomov [Wed, 9 Nov 2011 11:26:37 +0000]
Btrfs: close devices on all error paths in open_ctree()

Fix a bug introduced by 7e662854 where we would leave devices busy on
certain error paths in open_ctree().  fs_info is guaranteed to be
non-NULL now so it's safe to dereference it on all error paths.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

8 years agoBtrfs: avoid null dereference and leaks when bailing from open_ctree()
Ilya Dryomov [Tue, 8 Nov 2011 22:08:15 +0000]
Btrfs: avoid null dereference and leaks when bailing from open_ctree()

Fix bugs introduced by 6c41761f.  Firstly, after failing to allocate any
of the tree roots (first 'goto fail' in open_ctree()) we would
dereference a NULL fs_info pointer in free_fs_info().  Secondly, after
failures from init_srcu_struct(), setup_bdi() and new_inode() we would
leak all earlier allocated roots: fs_info fields haven't been
initialized yet so free_fs_info() is rendered useless.

Fix this by initializing fs_info pointer and fs_info fields before any
allocations happen.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

8 years agoBtrfs: fix subvol_name leak on error in btrfs_mount()
Ilya Dryomov [Tue, 8 Nov 2011 17:15:05 +0000]
Btrfs: fix subvol_name leak on error in btrfs_mount()

btrfs_parse_early_options() can fail due to error while scanning devices
(-o device= option), but still strdup() subvol_name string:

mount -o subvol=SUBV,device=BAD_DEVICE <dev> <mnt>

So free subvol_name string on error.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

8 years agoBtrfs: fix memory leak in btrfs_parse_early_options()
Ilya Dryomov [Tue, 8 Nov 2011 14:47:55 +0000]
Btrfs: fix memory leak in btrfs_parse_early_options()

Don't leak subvol_name string in case multiple subvol= options are
given.  "The lastest option is effective" behavior (consistent with
subvolid= and subvolrootid= options) is preserved.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

8 years agoALSA: intel8x0: improve virtual environment detection
Konstantin Ozerkov [Wed, 9 Nov 2011 15:28:55 +0000]
ALSA: intel8x0: improve virtual environment detection

Detection code improved by PCI SSID usage. VM optimization
now enabled only for known devcices (skip host devices forwarded
to VM by VT-d or same kind of technology).
For debug/troubleshooting purposes optimization can be
forced (on/off) by module parameter: "inside_vm" (boolean).

Known devices (PCI SSID):
1af4:1100: Reserved for KVM devices. Note this is not yet
           implemented for KVM's ICH/AC'97 emulation.
1ab8:xxxx: Parallels ICH/AC'97 emulated sound.

[ fixed a minor coding-style issue by tiwai]

Signed-off-by: Konstantin Ozerkov <kozerkov@parallels.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: intel8x0: move virtual environment detection code into one place
Konstantin Ozerkov [Wed, 9 Nov 2011 15:28:54 +0000]
ALSA: intel8x0: move virtual environment detection code into one place

This is refactoring patch: preparation for add improved detection code.
Now all detection code placed in one place.

Signed-off-by: Konstantin Ozerkov <kozerkov@parallels.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: snd_usb_audio: add Logitech HD Webcam c510 to quirk-384
Alexey Fisher [Wed, 9 Nov 2011 10:39:24 +0000]
ALSA: snd_usb_audio: add Logitech HD Webcam c510 to quirk-384

Logitech HD Webcam c510 provide wrong mixer resolution.
Add it to "res = 384" quirk.

Signed-off-by: Alexey Fisher <bug-track@fisher-privat.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: hda - fix internal mic on Dell Vostro 3500 laptop
Julian Wollrath [Wed, 9 Nov 2011 09:02:40 +0000]
ALSA: hda - fix internal mic on Dell Vostro 3500 laptop

Fix the not working internal mic on Dell Vostro 3500 laptop by introducing the
new model dell-vostro-3500.

Signed-off-by: Julian Wollrath <jwollrath@web.de>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoMerge branch 'fix/asoc' into for-linus
Takashi Iwai [Wed, 9 Nov 2011 07:52:22 +0000]
Merge branch 'fix/asoc' into for-linus

8 years agoALSA: HDA: Remove quirk for Toshiba T110
David Henningsson [Tue, 8 Nov 2011 19:37:26 +0000]
ALSA: HDA: Remove quirk for Toshiba T110

According to the bug reporter, model=auto is needed to make the
internal microphone work.

BugLink: https://bugs.launchpad.net/bugs/819699
Reported-by: Andrej (agno01)
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years ago[libata] ahci: Add ASMedia ASM1061 support
Keng-Yu Lin [Wed, 9 Nov 2011 06:47:36 +0000]
[libata] ahci: Add ASMedia ASM1061 support

Signed-off-by: Keng-Yu Lin <kengyu@canonical.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years ago[libata] Issue SRST to Sil3726 PMP
Gwendal Grignou [Thu, 20 Oct 2011 00:17:02 +0000]
[libata] Issue SRST to Sil3726 PMP

Reenable sending SRST to devices connected behind a Sil3726 PMP.
This allow staggered spinups and handles drives that spins up slowly.

While the drives spin up, the PMP will not accept SRST.
Most controller reissues the reset until the drive is ready, while
some [Sil3124] returns an error.
In ata_eh_error, wait 10s before reset the ATA port and try again.

Signed-off-by: Gwendal Grignou <gwendal@google.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agosata_sis.c: trivial spelling fix
Chris Dunlop [Sun, 23 Oct 2011 23:38:18 +0000]
sata_sis.c: trivial spelling fix

Trivial spelling fix.

Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agoahci_platform: use dev_get_platdata()
JiSheng Zhang [Mon, 31 Oct 2011 13:20:10 +0000]
ahci_platform: use dev_get_platdata()

Use dev_get_platdata() to retrieve the struct ahci_platform_data data
from the platform.

Signed-off-by: JiSheng Zhang <jszhang3@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years ago[libata] libata-scsi.c: Add function parameter documentation
Marcos Paulo de Souza [Tue, 1 Nov 2011 01:50:16 +0000]
[libata] libata-scsi.c: Add function parameter documentation

Add the documentation of parameters of ata_change_queue_depth to silence the warning of make xmldocs

Signed-off-by: Marcos paulo de Souza <marcos.mage@gmail.com>
Acked-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux
Linus Torvalds [Wed, 9 Nov 2011 02:34:16 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/paulg/linux

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux:
  drivers/media: video/a5k6aa is a module and so needs module.h
  mfd: fix build failures in recently added ab5500 code
  hwspinlock/u8500: include linux/module.h
  MTD: MAPS: bcm963xx-flash.c: explicitly include module.h

8 years agoMerge branch 'for-linus' of git://neil.brown.name/md
Linus Torvalds [Wed, 9 Nov 2011 02:33:53 +0000]
Merge branch 'for-linus' of git://neil.brown.name/md

* 'for-linus' of git://neil.brown.name/md:
  md/raid5: STRIPE_ACTIVE has lock semantics, add barriers
  md/raid5: abort any pending parity operations when array fails.

8 years agoMerge branch 'docs-move' of git://git.kernel.org/pub/scm/linux/kernel/git/rdunlap...
Linus Torvalds [Wed, 9 Nov 2011 02:33:11 +0000]
Merge branch 'docs-move' of git://git./linux/kernel/git/rdunlap/linux-docs

* 'docs-move' of git://git.kernel.org/pub/scm/linux/kernel/git/rdunlap/linux-docs: (45 commits)
  DocBook/drm: Clean up a todo-note
  DocBook/drm: `device aware' -> `device-aware'
  DocBook/drm: `(device|driver) specific' -> `(device|driver)-specific'
  DocBook/drm: Clean up the paragraph on framebuffer objects
  DocBook/drm: Use `; otherwise,'
  DocBook/drm: Better flow with `, and then'
  DocBook/drm: Refer to the domain-setting function as a device-specific ioctl
  DocBook/drm: Improve flow of GPU/CPU coherence sentence
  DocBook/drm: Use an <itemizelist> for fundamental GEM operations
  DocBook/drm: Insert a comma
  DocBook/drm: Use a <variablelist> for vblank ioctls
  DocBook/drm: Use an itemizedlist for what an encoder needs to provide
  DocBook/drm: Insert `the' for readability, and change `set' to `setting'
  DocBook/drm: Remove extraneous commas
  DocBook/drm: Use a colon
  DocBook/drm: Clarify `final initialization' via better formatting
  DocBook/drm: Remove redundancy
  DocBook/drm: Insert `it' for smooth reading
  DocBook/drm: The word `so-called'; I do not think it connotes what you think it connotes
  DocBook/drm: Use a singular subject for grammatical cleanliness
  ...

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt...
Linus Torvalds [Tue, 8 Nov 2011 23:41:10 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/rostedt/linux-ktest

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-ktest: (21 commits)
  ktest: Evaluate variables entered on the command line
  ktest: Add variable ${PWD}
  ktest: Add another monitor flush before installing kernel
  ktest: Do not opencode reboot in grub setting
  ktest: Add processing of complex conditionals
  ktest: Fix parsing of config section lines
  ktest: Sort make_min_config configs by dependecies
  ktest: Add DEFINED keyword for IF statements
  ktest: Add OVERRIDE keyword to DEFAULTS section
  ktest: Consolidate TEST_TYPE and DEFAULT code
  ktest: Add INCLUDE keyword to include other config files
  ktest: Let IF keyword take comparisons
  ktest: Add IF and ELSE to config sections
  ktest: Do not reboot on config or build issues
  ktest: Add option REBOOT_SUCCESS_LINE to stop waiting after a reboot
  ktest: Add NO_INSTALL option to not install for a test
  ktest: Fail when grub menu not found
  ktest: Include monitor in reboot code
  ktest: Only need to save .config when doing mrproper
  ktest: Create outputdir if it does not exist
  ...

8 years agoMerge branch 'pm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Linus Torvalds [Tue, 8 Nov 2011 22:24:18 +0000]
Merge branch 'pm-fixes' of git://git./linux/kernel/git/rafael/linux-pm

* 'pm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  PM / OPP: Use ERR_CAST instead of ERR_PTR(PTR_ERR())
  PM / devfreq: fix private_data
  Documentation: Fix typo in freezer-subsystem.txt
  PM / QoS: Set cpu_dma_pm_qos->name

8 years agoPM / OPP: Use ERR_CAST instead of ERR_PTR(PTR_ERR())
Thomas Meyer [Tue, 8 Nov 2011 21:34:00 +0000]
PM / OPP: Use ERR_CAST instead of ERR_PTR(PTR_ERR())

Use ERR_CAST inlined function instead of ERR_PTR(PTR_ERR(...))

[The semantic patch that makes this change is available
 in scripts/coccinelle/api/err_cast.cocci.

 More information about semantic patching is available at
 http://coccinelle.lip6.fr/]

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

8 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/ohad/hwspinlock
Linus Torvalds [Tue, 8 Nov 2011 20:50:54 +0000]
Merge branch 'fixes' of git://git./linux/kernel/git/ohad/hwspinlock

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/ohad/hwspinlock:
  hwspinlock/u8500: fix build error due to undefined label
  hwspinlock: Don't return a value in __hwspin_unlock

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc
Linus Torvalds [Tue, 8 Nov 2011 20:50:02 +0000]
Merge git://git./linux/kernel/git/davem/sparc

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc: Hook up process_vm_{readv,writev} syscalls.

8 years agox86 platform drivers: add POWER_SUPPLY to selected drivers for Dell
Linus Torvalds [Tue, 8 Nov 2011 20:49:29 +0000]
x86 platform drivers: add POWER_SUPPLY to selected drivers for Dell

The Kconfig loop detection goes crazy without this.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoBtrfs: fix our reservations for updating an inode when completing io
Josef Bacik [Tue, 8 Nov 2011 20:47:34 +0000]
Btrfs: fix our reservations for updating an inode when completing io

People have been reporting ENOSPC crashes in finish_ordered_io.  This is because
we try to steal from the delalloc block rsv to satisfy a reservation to update
the inode.  The problem with this is we don't explicitly save space for updating
the inode when doing delalloc.  This is kind of a problem and we've gotten away
with this because way back when we just stole from the delalloc reserve without
any questions, and this worked out fine because generally speaking the leaf had
been modified either by the mtime update when we did the original write or
because we just updated the leaf when we inserted the file extent item, only on
rare occasions had the leaf not actually been modified, and that was still ok
because we'd just use a block or two out of the over-reservation that is
delalloc.

Then came the delayed inode stuff.  This is amazing, except it wants a full
reservation for updating the inode since it may do it at some point down the
road after we've written the blocks and we have to recow everything again.  This
worked out because the delayed inode stuff just stole from the global reserve,
that is until recently when I changed that because it caused other problems.

So here we are, we're doing everything right and being screwed for it.  So take
an extra reservation for the inode at delalloc reservation time and carry it
through the life of the delalloc reservation.  If we need it we can steal it in
the delayed inode stuff.  If we have already stolen it try and do a normal
metadata reservation.  If that fails try to steal from the delalloc reservation.
If _that_ fails we'll get a WARN_ON() so I can start thinking of a better way to
solve this and in the meantime we'll steal from the global reserve.

With this patch I ran xfstests 13 in a loop for a couple of hours and didn't see
any problems.

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agox86 platform drivers: make Dell laptop driver select needed LED support
Linus Torvalds [Tue, 8 Nov 2011 20:17:25 +0000]
x86 platform drivers: make Dell laptop driver select needed LED support

Otherwise we get compile errors like this:

  ERROR: "led_classdev_unregister" [drivers/platform/x86/dell-laptop.ko] undefined!
  ERROR: "led_classdev_register" [drivers/platform/x86/dell-laptop.ko] undefined!
  make[1]: *** [__modpost] Error 1
  make: *** [modules] Error 2

when the dell-laptop support is enabled without the necessary LED
support being enabled.

Reported-by: Alessandro Suardi <alessandro.suardi@gmail.com>
Acked-by: Matthew Garrett <mjg59@srcf.ucam.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agodrivers/media: video/a5k6aa is a module and so needs module.h
Paul Gortmaker [Tue, 8 Nov 2011 19:56:50 +0000]
drivers/media: video/a5k6aa is a module and so needs module.h

This file uses core functions like module_init() and module_exit()
and so it explicitly needs to include the module.h header.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>

8 years agoBtrfs: fix oops on NULL trans handle in btrfs_truncate
Chris Mason [Tue, 8 Nov 2011 19:49:59 +0000]
Btrfs: fix oops on NULL trans handle in btrfs_truncate

If we fail to reserve space in the transaction during truncate, we can
error out with a NULL trans handle.  The cleanup code needs an extra
check to make sure we aren't trying to use the bad handle.

Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agon_gsm: Fix timings
Alan Cox [Tue, 8 Nov 2011 18:02:10 +0000]
n_gsm: Fix timings

Alek Du reported that the code erroneously applies time to jiffies
conversions twice to the t1 and t2 values. In normal use on a modem link
this cases no visible problem but on a slower link it will break as with
HZ=1000 as is typical we are running t1/t2 ten times too fast.

Alek's original patch removed the conversion from the timer setting but we
in fact have to be more careful as the contents of t1/t2 are visible via
the device API and we thus need to correct the constants.

Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoALSA: usb-audio - Fix the missing volume quirks at delayed init
Takashi Iwai [Tue, 8 Nov 2011 16:50:27 +0000]
ALSA: usb-audio - Fix the missing volume quirks at delayed init

In the recent usb-audio driver, the initialization of volume ranges
may be delayed when the device doesn't respond well at the probing time.
But the volume quirks for certain devices are applied only in
mixer_ctl_feature_info() thus only at the very first probe and will be
missing when the volume range is initialized later.

This patch moves the volume quirk code to be always called from the
volume-range extraction (get_min_max()), so that the quirks are properly
applied in the later init time.

Reported-and-tested-by: Alexey Fisher <bug-track@fisher-privat.net>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoxfs: fix force shutdown handling in xfs_end_io
Christoph Hellwig [Tue, 8 Nov 2011 08:56:15 +0000]
xfs: fix force shutdown handling in xfs_end_io

Ensure ioend->io_error gets propagated back to e.g. AIO completions.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>

8 years agoxfs: constify xfs_item_ops
Christoph Hellwig [Fri, 28 Oct 2011 09:54:24 +0000]
xfs: constify xfs_item_ops

The log item ops aren't nessecarily the biggest exploit vector, but marking
them const is easy enough.  Also remove the unused xfs_item_ops_t typedef
while we're at it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>

8 years agoxfs: Fix possible memory corruption in xfs_readlink
Carlos Maiolino [Mon, 7 Nov 2011 16:10:24 +0000]
xfs: Fix possible memory corruption in xfs_readlink

Fixes a possible memory corruption when the link is larger than
MAXPATHLEN and XFS_DEBUG is not enabled. This also remove the
S_ISLNK assert, since the inode mode is checked previously in
xfs_readlink_by_handle() and via VFS.

Updated to address concerns raised by Ben Hutchings about the loose
attention paid to 32- vs 64-bit values, and the lack of handling a
potentially negative pathlen value:
 - Changed type of "pathlen" to be xfs_fsize_t, to match that of
   ip->i_d.di_size
 - Added checking for a negative pathlen to the too-long pathlen
   test, and generalized the message that gets reported in that case
   to reflect the change
As a result, if a negative pathlen were encountered, this function
would return EFSCORRUPTED (and would fail an assertion for a debug
build)--just as would a too-long pathlen.

Signed-off-by: Alex Elder <aelder@sgi.com>
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>