firefly-linux-kernel-4.4.55.git
13 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block
Linus Torvalds [Thu, 3 Mar 2011 23:42:35 +0000 (15:42 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block

* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
  block: kill loop_mutex
  blktrace: Remove blk_fill_rwbs_rq.
  block: blk-flush shouldn't call directly into q->request_fn() __blk_run_queue()
  block: add @force_kblockd to __blk_run_queue()
  block: fix kernel-doc format for blkdev_issue_zeroout
  blk-throttle: Do not use kblockd workqueue for throtl work

13 years agoMerge branch 'i_nlink' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6
Linus Torvalds [Thu, 3 Mar 2011 23:37:59 +0000 (15:37 -0800)]
Merge branch 'i_nlink' of git://git./linux/kernel/git/viro/vfs-2.6

* 'i_nlink' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
  hfs: fix rename() over non-empty directory
  udf: fix i_nlink limit
  fix reiserfs mkdir() breakage
  exofs: i_nlink races in rename()
  nilfs2: i_nlink races in rename()
  minix: i_nlink races in rename()
  ufs: i_nlink races in rename()
  sysv: i_nlink races in rename()

13 years agoMerge branch 'for-linus' of git://oss.sgi.com/xfs/xfs
Linus Torvalds [Thu, 3 Mar 2011 20:44:22 +0000 (12:44 -0800)]
Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs

* 'for-linus' of git://oss.sgi.com/xfs/xfs:
  xfs: zero proper structure size for geometry calls

13 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke...
Linus Torvalds [Thu, 3 Mar 2011 20:42:48 +0000 (12:42 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/ryusuke/nilfs2

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
  nilfs2: fix regression that i-flag is not set on changeless checkpoints

13 years agoblock: kill loop_mutex
Petr Uzel [Thu, 3 Mar 2011 16:48:50 +0000 (11:48 -0500)]
block: kill loop_mutex

Following steps lead to deadlock in kernel:

dd if=/dev/zero of=img bs=512 count=1000
losetup -f img
mkfs.ext2 /dev/loop0
mount -t ext2 -o loop /dev/loop0 mnt
umount mnt/

Stacktrace:
[<c102ec04>] irq_exit+0x36/0x59
[<c101502c>] smp_apic_timer_interrupt+0x6b/0x75
[<c127f639>] apic_timer_interrupt+0x31/0x38
[<c101df88>] mutex_spin_on_owner+0x54/0x5b
[<fe2250e9>] lo_release+0x12/0x67 [loop]
[<c10c4eae>] __blkdev_put+0x7c/0x10c
[<c10a4da5>] fput+0xd5/0x1aa
[<fe2250cf>] loop_clr_fd+0x1a9/0x1b1 [loop]
[<fe225110>] lo_release+0x39/0x67 [loop]
[<c10c4eae>] __blkdev_put+0x7c/0x10c
[<c10a59d9>] deactivate_locked_super+0x17/0x36
[<c10b6f37>] sys_umount+0x27e/0x2a5
[<c10b6f69>] sys_oldumount+0xb/0xe
[<c1002897>] sysenter_do_call+0x12/0x26
[<ffffffff>] 0xffffffff

Regression since 2a48fc0ab24241755dc9, which introduced the private
loop_mutex as part of the BKL removal process.

As per [1], the mutex can be safely removed.

[1] http://www.gossamer-threads.com/lists/linux/kernel/1341930

Addresses: https://bugzilla.novell.com/show_bug.cgi?id=669394
Addresses: https://bugzilla.kernel.org/show_bug.cgi?id=29172

Signed-off-by: Petr Uzel <petr.uzel@suse.cz>
Cc: stable@kernel.org
Reviewed-by: Nikanth Karthikesan <knikanth@suse.de>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoblktrace: Remove blk_fill_rwbs_rq.
Tao Ma [Thu, 3 Mar 2011 15:53:20 +0000 (10:53 -0500)]
blktrace: Remove blk_fill_rwbs_rq.

If we enable trace events to trace block actions, We use
blk_fill_rwbs_rq to analyze the corresponding actions
in request's cmd_flags, but we only choose the minor 2 bits
from it, so most of other flags(e.g, REQ_SYNC) are missing.
For example, with a sync write we get:
write_test-2409  [001]   160.013869: block_rq_insert: 3,64 W 0 () 258135 + =
8 [write_test]

Since now we have integrated the flags of both bio and request,
it is safe to pass rq->cmd_flags directly to blk_fill_rwbs and
blk_fill_rwbs_rq isn't needed any more.

With this patch, after a sync write we get:
write_test-2417  [000]   226.603878: block_rq_insert: 3,64 WS 0 () 258135 +=
 8 [write_test]

Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Acked-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agohfs: fix rename() over non-empty directory
Al Viro [Thu, 3 Mar 2011 04:46:51 +0000 (23:46 -0500)]
hfs: fix rename() over non-empty directory

merge hfs_unlink() and hfs_rmdir(), while we are at it.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agoudf: fix i_nlink limit
Al Viro [Wed, 2 Mar 2011 15:15:26 +0000 (10:15 -0500)]
udf: fix i_nlink limit

(256 << sizeof(x)) - 1 is not the maximal possible value of x...
In reality, the maximal allowed value for UDF FileLinkCount is
65535.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agofix reiserfs mkdir() breakage
Al Viro [Wed, 2 Mar 2011 14:35:13 +0000 (09:35 -0500)]
fix reiserfs mkdir() breakage

if directory has so many subdirectories that its link count is set
to 1 (i.e. "can't tell accurately") and reiserfs_new_inode() fails,
we shouldn't decrement the parent's link count in cleanup path;
that's what DEC_DIR_INODE_NLINK() is for.  As it is, we end up
with parent suddenly getting zero i_nlink, with very unpleasant
effects.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agoexofs: i_nlink races in rename()
Al Viro [Wed, 2 Mar 2011 21:42:38 +0000 (16:42 -0500)]
exofs: i_nlink races in rename()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agonilfs2: i_nlink races in rename()
Al Viro [Wed, 2 Mar 2011 17:01:13 +0000 (12:01 -0500)]
nilfs2: i_nlink races in rename()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agominix: i_nlink races in rename()
Al Viro [Wed, 2 Mar 2011 14:41:38 +0000 (09:41 -0500)]
minix: i_nlink races in rename()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agoufs: i_nlink races in rename()
Al Viro [Wed, 2 Mar 2011 14:40:21 +0000 (09:40 -0500)]
ufs: i_nlink races in rename()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agosysv: i_nlink races in rename()
Al Viro [Wed, 2 Mar 2011 14:38:45 +0000 (09:38 -0500)]
sysv: i_nlink races in rename()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
13 years agoMerge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied...
Linus Torvalds [Thu, 3 Mar 2011 04:02:32 +0000 (20:02 -0800)]
Merge branch 'drm-fixes' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
  drm/i915: fix memory corruption with GM965 and >4GB RAM

13 years agoMerge branch 'devicetree/merge' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Thu, 3 Mar 2011 04:01:57 +0000 (20:01 -0800)]
Merge branch 'devicetree/merge' of git://git.secretlab.ca/git/linux-2.6

* 'devicetree/merge' of git://git.secretlab.ca/git/linux-2.6:
  of/promtree: allow DT device matching by fixing 'name' brokenness (v5)
  x86: OLPC: have prom_early_alloc BUG rather than return NULL
  of/flattree: Drop an uninteresting message to pr_debug level
  of: Add missing of_address.h to xilinx ehci driver

13 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mattst88...
Linus Torvalds [Thu, 3 Mar 2011 04:01:31 +0000 (20:01 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mattst88/alpha-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mattst88/alpha-2.6:
  alpha: Enable GENERIC_HARDIRQS_NO_DEPRECATED
  alpha: irq: Convert affinity to use irq_data
  alpha: irq: Remove pointless irq status manipulation
  alpha: titan: Convert irq_chip functions
  alpha: takara: Convert irq_chip functions
  alpha: sable: Convert irq_chip functions
  alpha: rx164: Convert irq_chip functions
  alpha: noritake: Convert irq_chip functions
  alpha: rawhide: Convert irq_chip functions
  alpha: mikasa: Convert irq_chip functions
  alpha: marvel: Convert irq_chip functions
  alpha: eiger: Convert irq_chip functions
  alpha: eb64p: Convert irq_chip functions
  alpha: dp264: Convert irq_chip functions
  alpha: cabriolet: Convert irq_chip functions
  alpha: i8259, alcor, jensen wildfire: Convert irq_chip
  alpha: srm: Convert irq_chip functions
  alpha: Pyxis convert irq_chip functions
  Fix typo in call to irq_to_desc()

13 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6
Linus Torvalds [Thu, 3 Mar 2011 04:01:07 +0000 (20:01 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/sameo/mfd-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6:
  mfd: Avoid tps6586x burst writes
  mfd: Don't suspend WM8994 if the CODEC is not suspended
  mfd: Fix DaVinci voice codec device name
  mfd: Fix NULL pointer due to non-initialized ucb1x00-ts absinfo
  mfd: Fix ASIC3 build with GENERIC_HARDIRQS_NO_DEPRECATED

13 years agoMerge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux...
Linus Torvalds [Thu, 3 Mar 2011 04:00:47 +0000 (20:00 -0800)]
Merge branch 'for_linus' of git://git./linux/kernel/git/jack/linux-fs-2.6

* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs-2.6:
  ext2: Fix link count corruption under heavy link+rename load

13 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/davej/cpufreq
Linus Torvalds [Thu, 3 Mar 2011 03:58:31 +0000 (19:58 -0800)]
Merge branch 'fixes' of git://git./linux/kernel/git/davej/cpufreq

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/davej/cpufreq:
  [CPUFREQ] p4-clockmod: print EST-capable warning message only once
  [CPUFREQ] fix BUG on cpufreq policy init failure
  [CPUFREQ] Fix another notifier leak in powernow-k8.
  [CPUFREQ] Missing "unregister_cpu_notifier" in powernow-k8.c

13 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Thu, 3 Mar 2011 03:58:14 +0000 (19:58 -0800)]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  powerpc: Fix call to flush_ptrace_hw_breakpoint()
  powerpc/kexec: Restore ppc_md.machine_kexec
  powerpc/mm: Make hpte_need_flush() safe for preemption

13 years agoMerge branch 'idle-release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb...
Linus Torvalds [Thu, 3 Mar 2011 02:08:03 +0000 (18:08 -0800)]
Merge branch 'idle-release' of git://git./linux/kernel/git/lenb/linux-idle-2.6

* 'idle-release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-idle-2.6:
  intel_idle: disable Atom/Lincroft HW C-state auto-demotion
  intel_idle: disable NHM/WSM HW C-state auto-demotion

13 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux...
Linus Torvalds [Wed, 2 Mar 2011 23:26:57 +0000 (15:26 -0800)]
Merge branch 'release' of git://git./linux/kernel/git/lenb/linux-acpi-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6:
  ACPI / ACPICA: Implicit notify for multiple devices
  ACPI / debugfs: Fix buffer overflows, double free

13 years agoof/promtree: allow DT device matching by fixing 'name' brokenness (v5)
Andres Salomon [Thu, 24 Feb 2011 06:38:22 +0000 (22:38 -0800)]
of/promtree: allow DT device matching by fixing 'name' brokenness (v5)

Commit e2f2a93b, "of/promtree: add package-to-path support to pdt"
changed dp->name from using the 'name' property to using
package-to-path.  This fixed /proc/device-tree creation by eliminating
conflicts between names (the 'name' property provides names like
'battery', whereas package-to-path provides names like
'/foo/bar/battery@0', which we stripped to 'battery@0').  However, it
also breaks of_device_id table matching.

The fix that we _really_ wanted was to keep dp->name based upon
the name property ('battery'), but based dp->full_name upon
package-to-path ('battery@0').  This patch does just that.

This changes all users (except SPARC) of promtree to use the full
result from package-to-path for full_name, rather than stripping the
directory out.  In practice, the strings end up being exactly the
same; this change saves time, code, and memory.

SPARC continues to use the existing build_path_component() code.

v2: combine two patches and revert of_pdt_node_name to original version
v3: use dp->phandle instead of passing around node
v4: warn/bail out for non-sparc archs if pkg2path is not set
v5: split of_pdt_build_full_name into sparc & non-sparc versions
v6: Pass NULL to pkg2path before buf gets assigned.
    Drop check for pkg2path hook on each and every node.
v7: Don't BUG() when unable to get the full_path; create a
    known-unique name instead.

Signed-off-by: Andres Salomon <dilinger@queued.net>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agox86: OLPC: have prom_early_alloc BUG rather than return NULL
Andres Salomon [Fri, 25 Feb 2011 04:06:31 +0000 (20:06 -0800)]
x86: OLPC: have prom_early_alloc BUG rather than return NULL

..similar to what sparc's prom_early_alloc does.

Signed-off-by: Andres Salomon <dilinger@queued.net>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoof/flattree: Drop an uninteresting message to pr_debug level
Paul Bolle [Mon, 14 Feb 2011 21:34:22 +0000 (22:34 +0100)]
of/flattree: Drop an uninteresting message to pr_debug level

This message looks like an error (which it isn't) when booting with a
flattened device tree.  Remove the message from normal kernel builds.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoof: Add missing of_address.h to xilinx ehci driver
Michal Simek [Mon, 14 Feb 2011 10:40:09 +0000 (11:40 +0100)]
of: Add missing of_address.h to xilinx ehci driver

Build log:
In file included from drivers/usb/host/ehci-hcd.c:1208:
drivers/usb/host/ehci-xilinx-of.c: In function 'ehci_hcd_xilinx_of_probe':
drivers/usb/host/ehci-xilinx-of.c:168: error: implicit declaration of function 'of_address_to_resource'

Signed-off-by: John Williams <john.williams@petalogix.com>
Signed-off-by: Michal Simek <monstr@monstr.eu>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agoalpha: Enable GENERIC_HARDIRQS_NO_DEPRECATED
Thomas Gleixner [Sun, 6 Feb 2011 14:33:02 +0000 (14:33 +0000)]
alpha: Enable GENERIC_HARDIRQS_NO_DEPRECATED

All interrupt chips are converted. Mark it clean.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: irq: Convert affinity to use irq_data
Thomas Gleixner [Sun, 6 Feb 2011 14:33:00 +0000 (14:33 +0000)]
alpha: irq: Convert affinity to use irq_data

affinity is moving to irq_data. Fix it up.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: irq: Remove pointless irq status manipulation
Thomas Gleixner [Sun, 6 Feb 2011 14:32:58 +0000 (14:32 +0000)]
alpha: irq: Remove pointless irq status manipulation

The irq descriptors are initialized IRQ_DISABLED in the generic
code. No need to fiddle with them.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: titan: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:56 +0000 (14:32 +0000)]
alpha: titan: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: takara: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:53 +0000 (14:32 +0000)]
alpha: takara: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: sable: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:51 +0000 (14:32 +0000)]
alpha: sable: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: rx164: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:49 +0000 (14:32 +0000)]
alpha: rx164: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: noritake: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:46 +0000 (14:32 +0000)]
alpha: noritake: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: rawhide: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:44 +0000 (14:32 +0000)]
alpha: rawhide: Convert irq_chip functions

Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: mikasa: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:42 +0000 (14:32 +0000)]
alpha: mikasa: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: marvel: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:39 +0000 (14:32 +0000)]
alpha: marvel: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: eiger: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:37 +0000 (14:32 +0000)]
alpha: eiger: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: eb64p: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:35 +0000 (14:32 +0000)]
alpha: eb64p: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: dp264: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:33 +0000 (14:32 +0000)]
alpha: dp264: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: cabriolet: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:30 +0000 (14:32 +0000)]
alpha: cabriolet: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: i8259, alcor, jensen wildfire: Convert irq_chip
Thomas Gleixner [Sun, 6 Feb 2011 14:32:28 +0000 (14:32 +0000)]
alpha: i8259, alcor, jensen wildfire: Convert irq_chip

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: srm: Convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:26 +0000 (14:32 +0000)]
alpha: srm: Convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoalpha: Pyxis convert irq_chip functions
Thomas Gleixner [Sun, 6 Feb 2011 14:32:23 +0000 (14:32 +0000)]
alpha: Pyxis convert irq_chip functions

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoFix typo in call to irq_to_desc()
Morten Holst Larsen [Sun, 23 Jan 2011 12:40:47 +0000 (13:40 +0100)]
Fix typo in call to irq_to_desc()

Fix typo in call to irq_to_desc()

Signed-off-by: Morten H. Larsen <m-larsen@post6.tele.dk>
Signed-off-by: Matt Turner <mattst88@gmail.com>
13 years agoblock: blk-flush shouldn't call directly into q->request_fn() __blk_run_queue()
Tejun Heo [Wed, 2 Mar 2011 13:48:06 +0000 (08:48 -0500)]
block: blk-flush shouldn't call directly into q->request_fn() __blk_run_queue()

blk-flush decomposes a flush into sequence of multiple requests.  On
completion of a request, the next one is queued; however, block layer
must not implicitly call into q->request_fn() directly from completion
path.  This makes the queue behave unexpectedly when seen from the
drivers and violates the assumption that q->request_fn() is called
with process context + queue_lock.

This patch makes blk-flush the following two changes to make sure
q->request_fn() is not called directly from request completion path.

- blk_flush_complete_seq_end_io() now asks __blk_run_queue() to always
  use kblockd instead of calling directly into q->request_fn().

- queue_next_fseq() uses ELEVATOR_INSERT_REQUEUE instead of
  ELEVATOR_INSERT_FRONT so that elv_insert() doesn't try to unplug the
  request queue directly.

Reported by Jan in the following threads.

 http://thread.gmane.org/gmane.linux.ide/48778
 http://thread.gmane.org/gmane.linux.ide/48786

stable: applicable to v2.6.37.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Jan Beulich <JBeulich@novell.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: stable@kernel.org
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoblock: add @force_kblockd to __blk_run_queue()
Tejun Heo [Wed, 2 Mar 2011 13:48:05 +0000 (08:48 -0500)]
block: add @force_kblockd to __blk_run_queue()

__blk_run_queue() automatically either calls q->request_fn() directly
or schedules kblockd depending on whether the function is recursed.
blk-flush implementation needs to be able to explicitly choose
kblockd.  Add @force_kblockd.

All the current users are converted to specify %false for the
parameter and this patch doesn't introduce any behavior change.

stable: This is prerequisite for fixing ide oops caused by the new
        blk-flush implementation.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Jan Beulich <JBeulich@novell.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: stable@kernel.org
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoext2: Fix link count corruption under heavy link+rename load
Josh Hunt [Thu, 24 Feb 2011 10:48:22 +0000 (11:48 +0100)]
ext2: Fix link count corruption under heavy link+rename load

vfs_rename_other() does not lock renamed inode with i_mutex. Thus changing
i_nlink in a non-atomic manner (which happens in ext2_rename()) can corrupt
it as reported and analyzed by Josh.

In fact, there is no good reason to mess with i_nlink of the moved file.
We did it presumably to simulate linking into the new directory and unlinking
from an old one. But the practical effect of this is disputable because fsck
can possibly treat file as being properly linked into both directories without
writing any error which is confusing. So we just stop increment-decrement
games with i_nlink which also fixes the corruption.

CC: stable@kernel.org
CC: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Josh Hunt <johunt@akamai.com>
Signed-off-by: Jan Kara <jack@suse.cz>
13 years agomfd: Avoid tps6586x burst writes
vwadekar@nvidia.com [Thu, 24 Feb 2011 04:48:13 +0000 (10:18 +0530)]
mfd: Avoid tps6586x burst writes

tps6586 does not support burst writes. i2c writes have to be
1 byte at a time.

Cc: stable@kernel.org
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
13 years agomfd: Don't suspend WM8994 if the CODEC is not suspended
Mark Brown [Fri, 4 Feb 2011 14:57:43 +0000 (14:57 +0000)]
mfd: Don't suspend WM8994 if the CODEC is not suspended

ASoC supports keeping the audio subsysetm active over suspend in order
to support use cases such as audio passthrough from a cellular modem
with the main CPU suspended. Ensure that we don't power down the CODEC
when this is happening by checking to see if VMID is up and skipping
suspend and resume when it is. If the CODEC has suspended then it'll
turn VMID off before the core suspend() gets called.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
13 years agomfd: Fix DaVinci voice codec device name
Manjunathappa, Prakash [Thu, 27 Jan 2011 13:28:36 +0000 (18:58 +0530)]
mfd: Fix DaVinci voice codec device name

Fix the device name in DaVinci Voice Codec MFD driver to load
davinci-vcif and cq93vc codec client drivers.

Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
13 years agomfd: Fix NULL pointer due to non-initialized ucb1x00-ts absinfo
Jochen Friedrich [Wed, 26 Jan 2011 10:30:01 +0000 (11:30 +0100)]
mfd: Fix NULL pointer due to non-initialized ucb1x00-ts absinfo

Call input_set_abs_params instead of manually setting absbit only.
This fixes this oops:

Unable to handle kernel NULL pointer dereference at virtual address 00000024
Internal error: Oops: 41b67017 [#1]
CPU: 0    Not tainted  (2.6.37 #4)
pc : [<c016d1fc>]    lr : [<00000000>]    psr: 20000093
sp : c19e5f30  ip : c19e5e6c  fp : c19e5f58
r10: 00000000  r9 : c19e4000  r8 : 00000003
r7 : 000001e4  r6 : 00000001  r5 : c1854400  r4 : 00000003
r3 : 00000018  r2 : 00000018  r1 : 00000018  r0 : c185447c
Flags: nzCv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
Control: c1b6717f  Table: c1b6717f  DAC: 00000017
Stack: (0xc19e5f30 to 0xc19e6000)
5f20:                                     00000003 00000003 c1854400 00000013
5f40: 00000001 000001e4 000001c5 c19e5f80 c19e5f5c c016d5e8 c016cf5c 000001e4
5f60: c1854400 c18b5860 00000000 00000171 000001e4 c19e5fc4 c19e5f84 c01559a4
5f80: c016d584 c18b5868 00000000 c1bb5c40 c0035afc c18b5868 c18b5868 c1a55d54
5fa0: c18b5860 c0155750 00000013 00000000 00000000 00000000 c19e5ff4 c19e5fc8
5fc0: c0050174 c015575c 00000000 c18b5860 00000000 c19e5fd4 c19e5fd4 c1a55d54
5fe0: c00500f0 c003b464 00000000 c19e5ff8 c003b464 c00500fc 04000400 04000400
Backtrace:
Function entered at [<c016cf50>] from [<c016d5e8>]
Function entered at [<c016d578>] from [<c01559a4>]
 r8:000001e4 r7:00000171 r6:00000000 r5:c18b5860 r4:c1854400
Function entered at [<c0155750>] from [<c0050174>]
Function entered at [<c00500f0>] from [<c003b464>]
 r6:c003b464 r5:c00500f0 r4:c1a55d54
Code: e59520fc e1a03286 e0433186 e0822003 (e592000c)

>>PC;  c016d1fc <input_handle_event+2ac/5a0>   <=====

Trace; c016cf50 <input_handle_event+0/5a0>
Trace; c016d5e8 <input_event+70/88>
Trace; c016d578 <input_event+0/88>
Trace; c01559a4 <ucb1x00_thread+254/2dc>
Trace; c0155750 <ucb1x00_thread+0/2dc>
Trace; c0050174 <kthread+84/8c>
Trace; c00500f0 <kthread+0/8c>
Trace; c003b464 <do_exit+0/624>

Signed-off-by: Jochen Friedrich <jochen@scram.de>
CC: stable@kernel.org
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
13 years agomfd: Fix ASIC3 build with GENERIC_HARDIRQS_NO_DEPRECATED
Lennert Buytenhek [Sat, 22 Jan 2011 01:08:30 +0000 (02:08 +0100)]
mfd: Fix ASIC3 build with GENERIC_HARDIRQS_NO_DEPRECATED

Signed-off-by: Lennert Buytenhek <buytenh@secretlab.ca>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
13 years agopowerpc: Fix call to flush_ptrace_hw_breakpoint()
K.Prasad [Thu, 10 Feb 2011 04:44:35 +0000 (04:44 +0000)]
powerpc: Fix call to flush_ptrace_hw_breakpoint()

Fix the error in spelling the config option for hw-breakpoints and fix
the build issue that follows.

Signed-off by: K.Prasad <prasad@linux.vnet.ibm.com>

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agopowerpc/kexec: Restore ppc_md.machine_kexec
Anton Blanchard [Wed, 23 Feb 2011 12:46:16 +0000 (12:46 +0000)]
powerpc/kexec: Restore ppc_md.machine_kexec

Kyle Moffett points out that mpc85xx has started using the
ppc_md.machine_kexec hook. As such, revert patch c94868788cf2
(powerpc/kexec: Remove ppc_md.machine_kexec).

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agopowerpc/mm: Make hpte_need_flush() safe for preemption
Peter Zijlstra [Thu, 24 Feb 2011 10:47:32 +0000 (10:47 +0000)]
powerpc/mm: Make hpte_need_flush() safe for preemption

hpte_need_flush() might be called outside of a preempt section
when manipulating the kernel page tables, so we need to use the
appopriate variants of per-cpu variable accesses. There should
be no risk of being in the middle of a batch and a context
switch will flush any pending batch.

[Patch extracted from a larger patch in Peter's preemptible
 mmu_gather series]

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
13 years agoxfs: zero proper structure size for geometry calls
Alex Elder [Tue, 1 Mar 2011 17:50:00 +0000 (17:50 +0000)]
xfs: zero proper structure size for geometry calls

Commit 493f3358cb289ccf716c5a14fa5bb52ab75943e5 added this call to
xfs_fs_geometry() in order to avoid passing kernel stack data back
to user space:

+       memset(geo, 0, sizeof(*geo));

Unfortunately, one of the callers of that function passes the
address of a smaller data type, cast to fit the type that
xfs_fs_geometry() requires.  As a result, this can happen:

Kernel panic - not syncing: stack-protector: Kernel stack is corrupted
in: f87aca93

Pid: 262, comm: xfs_fsr Not tainted 2.6.38-rc6-493f3358cb2+ #1
Call Trace:

[<c12991ac>] ? panic+0x50/0x150
[<c102ed71>] ? __stack_chk_fail+0x10/0x18
[<f87aca93>] ? xfs_ioc_fsgeometry_v1+0x56/0x5d [xfs]

Fix this by fixing that one caller to pass the right type and then
copy out the subset it is interested in.

Note: This patch is an alternative to one originally proposed by
Eric Sandeen.

Reported-by: Jeffrey Hundstad <jeffrey.hundstad@mnsu.edu>
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Tested-by: Jeffrey Hundstad <jeffrey.hundstad@mnsu.edu>
13 years agoMerge remote branch 'intel/drm-intel-fixes' of /ssd/git/drm-next into drm-fixes
Dave Airlie [Wed, 2 Mar 2011 02:48:23 +0000 (12:48 +1000)]
Merge remote branch 'intel/drm-intel-fixes' of /ssd/git/drm-next into drm-fixes

13 years agonilfs2: fix regression that i-flag is not set on changeless checkpoints
Ryusuke Konishi [Mon, 28 Feb 2011 04:41:11 +0000 (13:41 +0900)]
nilfs2: fix regression that i-flag is not set on changeless checkpoints

According to the report from Jiro SEKIBA titled "regression in
2.6.37?"  (Message-Id: <8739n8vs1f.wl%jir@sekiba.com>), on 2.6.37 and
later kernels, lscp command no longer displays "i" flag on checkpoints
that snapshot operations or garbage collection created.

This is a regression of nilfs2 checkpointing function, and it's
critical since it broke behavior of a part of nilfs2 applications.
For instance, snapshot manager of TimeBrowse gets to create
meaningless snapshots continuously; snapshot creation triggers another
checkpoint, but applications cannot distinguish whether the new
checkpoint contains meaningful changes or not without the i-flag.

This patch fixes the regression and brings that application behavior
back to normal.

Reported-by: Jiro SEKIBA <jir@unicus.jp>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Tested-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Tested-by: Jiro SEKIBA <jir@unicus.jp>
Cc: stable <stable@kernel.org> [2.6.37]
13 years ago[CPUFREQ] p4-clockmod: print EST-capable warning message only once
Naga Chumbalkar [Tue, 15 Feb 2011 17:44:11 +0000 (17:44 +0000)]
[CPUFREQ] p4-clockmod: print EST-capable warning message only once

Print the message only once. I see it 16 times on a 2P box with 16 logical CPUs.

Signed-off-by: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
13 years ago[CPUFREQ] fix BUG on cpufreq policy init failure
Jiri Slaby [Tue, 1 Mar 2011 16:41:10 +0000 (17:41 +0100)]
[CPUFREQ] fix BUG on cpufreq policy init failure

cpufreq_register_driver sets cpufreq_driver to a structure owned (and
placed) in the caller's memory. If cpufreq policy fails in its ->init
function, sysdev_driver_register returns nonzero in
cpufreq_register_driver. Now, cpufreq_register_driver returns an error
without setting cpufreq_driver back to NULL.

Usually cpufreq policy modules are unloaded because they propagate the
error to the module init function and return that.

So a later access to any member of cpufreq_driver causes bugs like:
BUG: unable to handle kernel paging request at ffffffffa00270a0
IP: [<ffffffff8145eca3>] cpufreq_cpu_get+0x53/0xe0
PGD 1805067 PUD 1809063 PMD 1c3f90067 PTE 0
Oops: 0000 [#1] SMP
last sysfs file: /sys/devices/virtual/net/tun0/statistics/collisions
CPU 0
Modules linked in: ...
Pid: 5677, comm: thunderbird-bin Tainted: G        W   2.6.38-rc4-mm1_64+ #1389 To be filled by O.E.M./To Be Filled By O.E.M.
RIP: 0010:[<ffffffff8145eca3>]  [<ffffffff8145eca3>] cpufreq_cpu_get+0x53/0xe0
RSP: 0018:ffff8801aec37d98  EFLAGS: 00010086
RAX: 0000000000000202 RBX: 0000000000000000 RCX: 0000000000000001
RDX: ffffffffa00270a0 RSI: 0000000000001000 RDI: ffffffff8199ece8
...
Call Trace:
 [<ffffffff8145f490>] cpufreq_quick_get+0x10/0x30
 [<ffffffff8103f12b>] show_cpuinfo+0x2ab/0x300
 [<ffffffff81136292>] seq_read+0xf2/0x3f0
 [<ffffffff8126c5d3>] ? __strncpy_from_user+0x33/0x60
 [<ffffffff8116850d>] proc_reg_read+0x6d/0xa0
 [<ffffffff81116e53>] vfs_read+0xc3/0x180
 [<ffffffff81116f5c>] sys_read+0x4c/0x90
 [<ffffffff81030dbb>] system_call_fastpath+0x16/0x1b
...

It's all cause by weird fail path handling in cpufreq_register_driver.
To fix that, shuffle the code to do proper handling with gotos.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: Dave Jones <davej@redhat.com>
13 years ago[CPUFREQ] Fix another notifier leak in powernow-k8.
Dave Jones [Wed, 24 Nov 2010 02:29:31 +0000 (21:29 -0500)]
[CPUFREQ] Fix another notifier leak in powernow-k8.

Do the notifier registration later, so we don't have to worry
about freeing it if we fail the msr allocation.

Signed-off-by: Dave Jones <davej@redhat.com>
13 years ago[CPUFREQ] Missing "unregister_cpu_notifier" in powernow-k8.c
Neil Brown [Wed, 24 Nov 2010 00:28:01 +0000 (11:28 +1100)]
[CPUFREQ] Missing "unregister_cpu_notifier" in powernow-k8.c

It appears that when powernow-k8 finds that

    No compatible ACPI _PSS objects found.

 and suggests

    Try again with latest BIOS.

 it fails the module load, but does not unregister the cpu_notifier that was
 registered in powernowk8_init

 This ends up leaving freed memory on the cpu notifier list for some other
 poor module (e.g. md/raid5) to come along and trip over.

 The following might be a partial fix, but I suspect there is probably other
 clean-up that is needed.

 ( https://bugzilla.novell.com/show_bug.cgi?id=655215 has full dmesg traces).

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Neil Brown <neilb@suse.de>
13 years agodrm/i915: fix memory corruption with GM965 and >4GB RAM
Jan Niehusmann [Tue, 1 Mar 2011 22:24:16 +0000 (23:24 +0100)]
drm/i915: fix memory corruption with GM965 and >4GB RAM

On a Thinkpad x61s, I noticed some memory corruption when
plugging/unplugging the external VGA connection. The symptoms are that
4 bytes at the beginning of a page get overwritten by zeroes.
The address of the corruption varies when rebooting the machine, but
stays constant while it's running (so it's possible to repeatedly write
some data and then corrupt it again by plugging the cable).

Further investigation revealed that the corrupted address is
(dev_priv->status_page_dmah->busaddr & 0xffffffff), ie. the beginning of
the hardware status page of the i965 graphics card, cut to 32 bits.

So it seems that for some memory access, the hardware uses only 32 bit
addressing. If the hardware status page is located >4GB, this
corrupts unrelated memory.

Signed-off-by: Jan Niehusmann <jan@gondor.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org
13 years agoLinux 2.6.38-rc7
Linus Torvalds [Tue, 1 Mar 2011 21:55:12 +0000 (13:55 -0800)]
Linux 2.6.38-rc7

13 years agoRevert "TPM: Long default timeout fix"
Linus Torvalds [Tue, 1 Mar 2011 21:23:27 +0000 (13:23 -0800)]
Revert "TPM: Long default timeout fix"

This reverts commit c4ff4b829ef9e6353c0b133b7adb564a68054979.

Ted Ts'o reports:

 "TPM is working for me so I can log into employer's network in 2.6.37.
  It broke when I tried 2.6.38-rc6, with the following relevant lines
  from my dmesg:

  [   11.081627] tpm_tis 00:0b: 1.2 TPM (device-id 0x0, rev-id 78)
  [   25.734114] tpm_tis 00:0b: Operation Timed out
  [   78.040949] tpm_tis 00:0b: Operation Timed out

  This caused me to get suspicious, especially since the _other_ TPM
  commit in 2.6.38 had already been reverted, so I tried reverting
  commit c4ff4b829e: "TPM: Long default timeout fix".  With this commit
  reverted, my TPM on my Lenovo T410 is once again working."

Requested-and-tested-by: Theodore Ts'o <tytso@mit.edu>
Acked-by: Rajiv Andrade <srajiv@linux.vnet.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoblock: fix kernel-doc format for blkdev_issue_zeroout
Ben Hutchings [Tue, 1 Mar 2011 18:45:24 +0000 (13:45 -0500)]
block: fix kernel-doc format for blkdev_issue_zeroout

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoblk-throttle: Do not use kblockd workqueue for throtl work
Vivek Goyal [Tue, 1 Mar 2011 18:40:54 +0000 (13:40 -0500)]
blk-throttle: Do not use kblockd workqueue for throtl work

o Dominik Klein reported a system hang issue while doing some blkio
  throttling testing.

  https://lkml.org/lkml/2011/2/24/173

o Some tracing revealed that CFQ was not dispatching any more jobs as
  queue unplug was not happening. And queue unplug was not happening
  because unplug work was not being called as there was one throttling
  work on same cpu which as not finished yet. And throttling work had not
  finished as it was tyring to dispatch a bio to CFQ but all the request
  descriptors were consume to it was put to sleep.

o So basically it is a cyclic dependecny between CFQ unplug work and
  throtl dispatch work. Tejun suggested that use separate workqueue for
  such cases.

o This patch uses a separate workqueue for throttle related work and
  does not rely on kblockd workqueue anymore.

Cc: stable@kernel.org
Reported-by: Dominik Klein <dk@in-telegence.net>
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
13 years agoMerge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groec...
Linus Torvalds [Tue, 1 Mar 2011 02:09:02 +0000 (18:09 -0800)]
Merge branch 'hwmon-for-linus' of git://git./linux/kernel/git/groeck/staging

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/staging:
  hwmon: (adt7411) add MODULE_DEVICE_TABLE
  hwmon: (ad7414) add MODULE_DEVICE_TABLE

13 years agofs/block_dev.c: fix new kernel-doc warning
Randy Dunlap [Sat, 26 Feb 2011 18:54:00 +0000 (10:54 -0800)]
fs/block_dev.c: fix new kernel-doc warning

Fix new kernel-doc warning in fs/block_dev.c:

Warning(fs/block_dev.c:937): No description found for parameter 'kill_dirty'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoACPI: Fix build for CONFIG_NET unset
Rafael J. Wysocki [Tue, 1 Mar 2011 00:12:19 +0000 (01:12 +0100)]
ACPI: Fix build for CONFIG_NET unset

Several ACPI drivers fail to build if CONFIG_NET is unset, because
they refer to things depending on CONFIG_THERMAL that in turn depends
on CONFIG_NET.  However, CONFIG_THERMAL doesn't really need to depend
on CONFIG_NET, because the only part of it requiring CONFIG_NET is
the netlink interface in thermal_sys.c.

Put the netlink interface in thermal_sys.c under #ifdef CONFIG_NET
and remove the dependency of CONFIG_THERMAL on CONFIG_NET from
drivers/thermal/Kconfig.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Len Brown <lenb@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Luming Yu <luming.yu@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoMerge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied...
Linus Torvalds [Tue, 1 Mar 2011 01:58:09 +0000 (17:58 -0800)]
Merge branch 'drm-fixes' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
  drm: fix unsigned vs signed comparison issue in modeset ctl ioctl.
  drm/nv50-nvc0: make sure vma is definitely unmapped when destroying bo

13 years agoMerge branch 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Tue, 1 Mar 2011 01:57:30 +0000 (17:57 -0800)]
Merge branch 'omap-fixes-for-linus' of git://git./linux/kernel/git/tmlind/linux-omap-2.6

* 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6:
  omap4: prcm: Fix the CPUx clockdomain offsets
  OMAP2+: clocksource: fix crash on boot when !CONFIG_OMAP_32K_TIMER
  OMAP2/3: clock: fix fint calculation for DPLL_FREQSEL
  OMAP2+: mailbox: fix lookups for multiple mailboxes
  OMAP2420: mailbox: fix IVA vs DSP IRQ numbering
  mach-omap2: smartreflex: world-writable debugfs voltage files
  mach-omap2: pm: world-writable debugfs timer files
  mach-omap2: mux: world-writable debugfs files

13 years agoMerge branches 'perf-fixes-for-linus', 'x86-fixes-for-linus' and 'timers-fixes-for...
Linus Torvalds [Tue, 1 Mar 2011 01:55:08 +0000 (17:55 -0800)]
Merge branches 'perf-fixes-for-linus', 'x86-fixes-for-linus' and 'timers-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  perf timechart: Fix max number of cpus
  perf timechart: Fix black idle boxes in the title
  perf hists: Print number of samples, not the period sum

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86: Use u32 instead of long to set reset vector back to 0

* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  clockevents: Prevent oneshot mode when broadcast device is periodic

13 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi...
Linus Torvalds [Tue, 1 Mar 2011 01:53:04 +0000 (17:53 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mszeredi/fuse

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
  fuse: fix truncate after open
  fuse: fix hang of single threaded fuseblk filesystem

13 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec...
Linus Torvalds [Tue, 1 Mar 2011 01:52:47 +0000 (17:52 -0800)]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jlbec/ocfs2

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2:
  ocfs2: Check heartbeat mode for kernel stacks only
  Ocfs2/refcounttree: Fix a bug for refcounttree to writeback clusters in a right number.
  ocfs2: Fix estimate of necessary credits for mkdir

13 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Tue, 1 Mar 2011 01:47:09 +0000 (17:47 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  eukrea-tlv320: fix platform_name
  ASoC: correct pxa AC97 DAI names
  ALSA: hda - Add support for new IDT 92HD98 and 92HD99 codecs
  ALSA: HDA: Add ideapad quirk for two Dell machines
  ALSA: HDA: Add a new Conexant codec 506e (20590)
  ALSA: usb-audio: fix oops due to cleanup race when disconnecting
  ASoC: Hook wm_hubs micbiases up to CLK_SYS
  ASoC: Correct definition of WM8903_VMID_RES_5K
  ASoC: Fix WM8958 default microphone detection argument ordering
  ALSA: HDA: Fix mic initialization in VIA auto parser
  ALSA: fix one memory leak in sound jack

13 years agomm: <asm-generic/pgtable.h> must include <linux/mm_types.h>
Ben Hutchings [Sun, 27 Feb 2011 05:41:35 +0000 (05:41 +0000)]
mm: <asm-generic/pgtable.h> must include <linux/mm_types.h>

Commit e2cda3226481 ("thp: add pmd mangling generic functions") replaced
some macros in <asm-generic/pgtable.h> with inline functions.

If the functions are to be defined (not all architectures need them)
then struct vm_area_struct must be defined first.  So include
<linux/mm_types.h>.

Fixes a build failure seen in Debian:

    CC [M]  drivers/media/dvb/mantis/mantis_pci.o
  In file included from arch/arm/include/asm/pgtable.h:460,
                   from drivers/media/dvb/mantis/mantis_pci.c:25:
  include/asm-generic/pgtable.h: In function 'ptep_test_and_clear_young':
  include/asm-generic/pgtable.h:29: error: dereferencing pointer to incomplete type

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agox86: Use u32 instead of long to set reset vector back to 0
Don Zickus [Tue, 8 Feb 2011 04:25:00 +0000 (23:25 -0500)]
x86: Use u32 instead of long to set reset vector back to 0

A customer of ours, complained that when setting the reset
vector back to 0, it trashed other data and hung their box.
They noticed when only 4 bytes were set to 0 instead of 8,
everything worked correctly.

Mathew pointed out:

 |
 | We're supposed to be resetting trampoline_phys_low and
 | trampoline_phys_high here, which are two 16-bit values.
 | Writing 64 bits is definitely going to overwrite space
 | that we're not supposed to be touching.
 |

So limit the area modified to u32.

Signed-off-by: Don Zickus <dzickus@redhat.com>
Acked-by: Matthew Garrett <mjg@redhat.com>
Cc: <stable@kernel.org>
LKML-Reference: <1297139100-424-1-git-send-email-dzickus@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
13 years agoperf timechart: Fix max number of cpus
Thomas Renninger [Sun, 27 Feb 2011 21:36:46 +0000 (22:36 +0100)]
perf timechart: Fix max number of cpus

Currently numcpus is determined in pid_put_sample which is only
called on sched_switch/sched_wakeup sample processing.

On a machine with a lot cpus I often saw the last cpu missing.

Check for (max) numcpus on every event happening and in the
beginning. -> fixes the issue for me.

Signed-off-by: Thomas Renninger <trenn@suse.de>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: lenb@kernel.org
LKML-Reference: <1298842606-55712-6-git-send-email-trenn@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
13 years agoperf timechart: Fix black idle boxes in the title
Thomas Renninger [Sun, 27 Feb 2011 21:36:45 +0000 (22:36 +0100)]
perf timechart: Fix black idle boxes in the title

This fix is needed for eye of gnome and firefox svg viewers.
Only Inkscape can handle the broken case.

Compare with the other svg_legenda_box declarations, looks
like a typo slipped in at this place.

Signed-off-by: Thomas Renninger <trenn@suse.de>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: lenb@kernel.org
LKML-Reference: <1298842606-55712-5-git-send-email-trenn@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
13 years agoMerge remote branch 'nouveau/drm-nouveau-fixes' of /ssd/git/drm-nouveau-next into...
Dave Airlie [Mon, 28 Feb 2011 05:35:16 +0000 (15:35 +1000)]
Merge remote branch 'nouveau/drm-nouveau-fixes' of /ssd/git/drm-nouveau-next into drm-fixes

* 'nouveau/drm-nouveau-fixes' of /ssd/git/drm-nouveau-next:
  drm/nv50-nvc0: make sure vma is definitely unmapped when destroying bo

13 years agodrm: fix unsigned vs signed comparison issue in modeset ctl ioctl.
Dave Airlie [Wed, 23 Feb 2011 22:35:06 +0000 (08:35 +1000)]
drm: fix unsigned vs signed comparison issue in modeset ctl ioctl.

This fixes CVE-2011-1013.

Reported-by: Matthiew Herrb (OpenBSD X.org team)
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agodrm/nv50-nvc0: make sure vma is definitely unmapped when destroying bo
Ben Skeggs [Mon, 28 Feb 2011 04:22:12 +0000 (14:22 +1000)]
drm/nv50-nvc0: make sure vma is definitely unmapped when destroying bo

Somehow fixes a misrendering + hang at GDM startup on my NVA8...

My first guess would have been stale TLB entries laying around that a new
bo then accidentally inherits.  That doesn't make a great deal of sense
however, as when we mapped the pages for the new bo the TLBs would've
gotten flushed anyway.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
13 years agohwmon: (adt7411) add MODULE_DEVICE_TABLE
axel lin [Thu, 24 Feb 2011 02:22:01 +0000 (02:22 +0000)]
hwmon: (adt7411) add MODULE_DEVICE_TABLE

The device table is required to load modules based on modaliases.

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
13 years agohwmon: (ad7414) add MODULE_DEVICE_TABLE
axel lin [Thu, 24 Feb 2011 02:20:37 +0000 (02:20 +0000)]
hwmon: (ad7414) add MODULE_DEVICE_TABLE

The device table is required to load modules based on modaliases.

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
13 years agoMerge branch 'fix/asoc' into for-linus
Takashi Iwai [Sat, 26 Feb 2011 10:27:47 +0000 (11:27 +0100)]
Merge branch 'fix/asoc' into for-linus

13 years agoclockevents: Prevent oneshot mode when broadcast device is periodic
Thomas Gleixner [Fri, 25 Feb 2011 21:34:23 +0000 (22:34 +0100)]
clockevents: Prevent oneshot mode when broadcast device is periodic

When the per cpu timer is marked CLOCK_EVT_FEAT_C3STOP, then we only
can switch into oneshot mode, when the backup broadcast device
supports oneshot mode as well. Otherwise we would try to switch the
broadcast device into an unsupported mode unconditionally. This went
unnoticed so far as the current available broadcast devices support
oneshot mode. Seth unearthed this problem while debugging and working
around an hpet related BIOS wreckage.

Add the necessary check to tick_is_oneshot_available().

Reported-and-tested-by: Seth Forshee <seth.forshee@canonical.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
LKML-Reference: <alpine.LFD.2.00.1102252231200.2701@localhost6.localdomain6>
Cc: stable@kernel.org # .21 ->
13 years agoMerge branch 'pm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspe...
Linus Torvalds [Fri, 25 Feb 2011 23:15:17 +0000 (15:15 -0800)]
Merge branch 'pm-fixes' of git://git./linux/kernel/git/rafael/suspend-2.6

* 'pm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspend-2.6:
  PM: Make ACPI wakeup from S5 work again when CONFIG_PM_SLEEP is unset

13 years agorapidio: fix sysfs config attribute to access 16MB of maint space
Alexandre Bounine [Fri, 25 Feb 2011 22:44:31 +0000 (14:44 -0800)]
rapidio: fix sysfs config attribute to access 16MB of maint space

Fixes sysfs config attribute to allow access to entire 16MB maintenance
space of RapidIO devices.

Signed-off-by: Alexandre Bounine <alexandre.bounine@idt.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: Matt Porter <mporter@kernel.crashing.org>
Cc: Li Yang <leoli@freescale.com>
Cc: Thomas Moll <thomas.moll@sysgo.com>
Cc: Micha Nelissen <micha@neli.hopto.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agopps: initialize ts_real properly
Alexander Gordeev [Fri, 25 Feb 2011 22:44:30 +0000 (14:44 -0800)]
pps: initialize ts_real properly

Initialize ts_real.flags to fix compiler warning about possible
uninitialized use of this field.

Signed-off-by: Alexander Gordeev <lasaine@lvk.cs.msu.su>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Rodolfo Giometti <giometti@linux.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agomemcg: more mem_cgroup_uncharge() batching
Hugh Dickins [Fri, 25 Feb 2011 22:44:29 +0000 (14:44 -0800)]
memcg: more mem_cgroup_uncharge() batching

It seems odd that truncate_inode_pages_range(), called not only when
truncating but also when evicting inodes, has mem_cgroup_uncharge_start
and _end() batching in its second loop to clear up a few leftovers, but
not in its first loop that does almost all the work: add them there too.

Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agothp: fix interleaving for transparent hugepages
Andi Kleen [Fri, 25 Feb 2011 22:44:28 +0000 (14:44 -0800)]
thp: fix interleaving for transparent hugepages

The THP code didn't pass the correct interleaving shift to the memory
policy code.  Fix this here by adjusting for the order.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Acked-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoaio: fix race between io_destroy() and io_submit()
Jan Kara [Fri, 25 Feb 2011 22:44:27 +0000 (14:44 -0800)]
aio: fix race between io_destroy() and io_submit()

A race can occur when io_submit() races with io_destroy():

 CPU1 CPU2
io_submit()
  do_io_submit()
    ...
    ctx = lookup_ioctx(ctx_id);
io_destroy()
    Now do_io_submit() holds the last reference to ctx.
    ...
    queue new AIO
    put_ioctx(ctx) - frees ctx with active AIOs

We solve this issue by checking whether ctx is being destroyed in AIO
submission path after adding new AIO to ctx.  Then we are guaranteed that
either io_destroy() waits for new AIO or we see that ctx is being
destroyed and bail out.

Cc: Nick Piggin <npiggin@kernel.dk>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoaio: fix rcu ioctx lookup
Nick Piggin [Fri, 25 Feb 2011 22:44:26 +0000 (14:44 -0800)]
aio: fix rcu ioctx lookup

aio-dio-invalidate-failure GPFs in aio_put_req from io_submit.

lookup_ioctx doesn't implement the rcu lookup pattern properly.
rcu_read_lock does not prevent refcount going to zero, so we might take
a refcount on a zero count ioctx.

Fix the bug by atomically testing for zero refcount before incrementing.

[jack@suse.cz: added comment into the code]
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agomm: fix dubious code in __count_immobile_pages()
Namhyung Kim [Fri, 25 Feb 2011 22:44:25 +0000 (14:44 -0800)]
mm: fix dubious code in __count_immobile_pages()

When pfn_valid_within() failed 'iter' was incremented twice.

Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agodrivers/rtc/rtc-ds3232.c: fix time range difference between linux and RTC chip
Lei Xu [Fri, 25 Feb 2011 22:44:23 +0000 (14:44 -0800)]
drivers/rtc/rtc-ds3232.c: fix time range difference between linux and RTC chip

In linux rtc_time struct, tm_mon range is 0~11, tm_wday range is 0~6,
while in RTC HW REG, month range is 1~12, day of the week range is 1~7,
this patch adjusts difference of them.

The efect of this bug was that most of month will be operated on as the
next month by the hardware (When in Jan it maybe even worse).  For
example, if in May, software wrote 4 to the hardware, which handled it as
April.  Then the logic would be different between software and hardware,
which would cause weird things to happen.

Signed-off-by: Lei Xu <B33228@freescale.com>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Jack Lan <jack.lan@freescale.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoldm: corrupted partition table can cause kernel oops
Timo Warns [Fri, 25 Feb 2011 22:44:21 +0000 (14:44 -0800)]
ldm: corrupted partition table can cause kernel oops

The kernel automatically evaluates partition tables of storage devices.
The code for evaluating LDM partitions (in fs/partitions/ldm.c) contains
a bug that causes a kernel oops on certain corrupted LDM partitions.  A
kernel subsystem seems to crash, because, after the oops, the kernel no
longer recognizes newly connected storage devices.

The patch changes ldm_parse_vmdb() to Validate the value of vblk_size.

Signed-off-by: Timo Warns <warns@pre-sense.de>
Cc: Eugene Teo <eugeneteo@kernel.sg>
Acked-by: Richard Russon <ldm@flatcap.org>
Cc: Harvey Harrison <harvey.harrison@gmail.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agomm: vmscan: stop reclaim/compaction earlier due to insufficient progress if !__GFP_REPEAT
Mel Gorman [Fri, 25 Feb 2011 22:44:20 +0000 (14:44 -0800)]
mm: vmscan: stop reclaim/compaction earlier due to insufficient progress if !__GFP_REPEAT

should_continue_reclaim() for reclaim/compaction allows scanning to
continue even if pages are not being reclaimed until the full list is
scanned.  In terms of allocation success, this makes sense but potentially
it introduces unwanted latency for high-order allocations such as
transparent hugepages and network jumbo frames that would prefer to fail
the allocation attempt and fallback to order-0 pages.  Worse, there is a
potential that the full LRU scan will clear all the young bits, distort
page aging information and potentially push pages into swap that would
have otherwise remained resident.

This patch will stop reclaim/compaction if no pages were reclaimed in the
last SWAP_CLUSTER_MAX pages that were considered.  For allocations such as
hugetlbfs that use __GFP_REPEAT and have fewer fallback options, the full
LRU list may still be scanned.

Order-0 allocation should not be affected because RECLAIM_MODE_COMPACTION
is not set so the following avoids the gfp_mask being examined:

        if (!(sc->reclaim_mode & RECLAIM_MODE_COMPACTION))
                return false;

A tool was developed based on ftrace that tracked the latency of
high-order allocations while transparent hugepage support was enabled and
three benchmarks were run.  The "fix-infinite" figures are 2.6.38-rc4 with
Johannes's patch "vmscan: fix zone shrinking exit when scan work is done"
applied.

  STREAM Highorder Allocation Latency Statistics
                 fix-infinite     break-early
  1 :: Count            10298           10229
  1 :: Min             0.4560          0.4640
  1 :: Mean            1.0589          1.0183
  1 :: Max            14.5990         11.7510
  1 :: Stddev          0.5208          0.4719
  2 :: Count                2               1
  2 :: Min             1.8610          3.7240
  2 :: Mean            3.4325          3.7240
  2 :: Max             5.0040          3.7240
  2 :: Stddev          1.5715          0.0000
  9 :: Count           111696          111694
  9 :: Min             0.5230          0.4110
  9 :: Mean           10.5831         10.5718
  9 :: Max            38.4480         43.2900
  9 :: Stddev          1.1147          1.1325

Mean time for order-1 allocations is reduced.  order-2 looks increased but
with so few allocations, it's not particularly significant.  THP mean
allocation latency is also reduced.  That said, allocation time varies so
significantly that the reductions are within noise.

Max allocation time is reduced by a significant amount for low-order
allocations but reduced for THP allocations which presumably are now
breaking before reclaim has done enough work.

  SysBench Highorder Allocation Latency Statistics
                 fix-infinite     break-early
  1 :: Count            15745           15677
  1 :: Min             0.4250          0.4550
  1 :: Mean            1.1023          1.0810
  1 :: Max            14.4590         10.8220
  1 :: Stddev          0.5117          0.5100
  2 :: Count                1               1
  2 :: Min             3.0040          2.1530
  2 :: Mean            3.0040          2.1530
  2 :: Max             3.0040          2.1530
  2 :: Stddev          0.0000          0.0000
  9 :: Count             2017            1931
  9 :: Min             0.4980          0.7480
  9 :: Mean           10.4717         10.3840
  9 :: Max            24.9460         26.2500
  9 :: Stddev          1.1726          1.1966

Again, mean time for order-1 allocations is reduced while order-2
allocations are too few to draw conclusions from.  The mean time for THP
allocations is also slightly reduced albeit the reductions are within
varianes.

Once again, our maximum allocation time is significantly reduced for
low-order allocations and slightly increased for THP allocations.

  Anon stream mmap reference Highorder Allocation Latency Statistics
  1 :: Count             1376            1790
  1 :: Min             0.4940          0.5010
  1 :: Mean            1.0289          0.9732
  1 :: Max             6.2670          4.2540
  1 :: Stddev          0.4142          0.2785
  2 :: Count                1               -
  2 :: Min             1.9060               -
  2 :: Mean            1.9060               -
  2 :: Max             1.9060               -
  2 :: Stddev          0.0000               -
  9 :: Count            11266           11257
  9 :: Min             0.4990          0.4940
  9 :: Mean        27250.4669      24256.1919
  9 :: Max      11439211.0000    6008885.0000
  9 :: Stddev     226427.4624     186298.1430

This benchmark creates one thread per CPU which references an amount of
anonymous memory 1.5 times the size of physical RAM.  This pounds swap
quite heavily and is intended to exercise THP a bit.

Mean allocation time for order-1 is reduced as before.  It's also reduced
for THP allocations but the variations here are pretty massive due to
swap.  As before, maximum allocation times are significantly reduced.

Overall, the patch reduces the mean and maximum allocation latencies for
the smaller high-order allocations.  This was with Slab configured so it
would be expected to be more significant with Slub which uses these size
allocations more aggressively.

The mean allocation times for THP allocations are also slightly reduced.
The maximum latency was slightly increased as predicted by the comments
due to reclaim/compaction breaking early.  However, workloads care more
about the latency of lower-order allocations than THP so it's an
acceptable trade-off.

Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>