Background inode reclaim needs to run more frequently that the XFS
syncd work is run as 30s is too long between optimal reclaim runs.
Add a new periodic work item to the xfs syncd workqueue to run a
fast, non-blocking inode reclaim scan.
Background inode reclaim is kicked by the act of marking inodes for
reclaim. When an AG is first marked as having reclaimable inodes,
the background reclaim work is kicked. It will continue to run
periodically untill it detects that there are no more reclaimable
inodes. It will be kicked again when the first inode is queued for
reclaim.
To ensure shrinker based inode reclaim throttles to the inode
cleaning and reclaim rate but still reclaim inodes efficiently, make it kick the
background inode reclaim so that when we are low on memory we are
trying to reclaim inodes as efficiently as possible. This kick shoul
d not be necessary, but it will protect against failures to kick the
background reclaim when inodes are first dirtied.
To provide the rate throttling, make the shrinker pass do
synchronous inode reclaim so that it blocks on inodes under IO. This
means that the shrinker will reclaim inodes rather than just
skipping over them, but it does not adversely affect the rate of
reclaim because most dirty inodes are already under IO due to the
background reclaim work the shrinker kicked.
These two modifications solve one of the two OOM killer invocations
Chris Mason reported recently when running a stress testing script.
The particular workload trigger for the OOM killer invocation is
where there are more threads than CPUs all unlinking files in an
extremely memory constrained environment. Unlike other solutions,
this one does not have a performance impact on performance when
memory is not constrained or the number of concurrent threads
operating is <= to the number of CPUs.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
On of the problems with the current inode flush at ENOSPC is that we
queue a flush per ENOSPC event, regardless of how many are already
queued. Thi can result in hundreds of queued flushes, most of
which simply burn CPU scanned and do no real work. This simply slows
down allocation at ENOSPC.
We really only need one active flush at a time, and we can easily
implement that via the new xfs_syncd_wq. All we need to do is queue
a flush if one is not already active, then block waiting for the
currently active flush to complete. The result is that we only ever
have a single ENOSPC inode flush active at a time and this greatly
reduces the overhead of ENOSPC processing.
On my 2p test machine, this results in tests exercising ENOSPC
conditions running significantly faster - 042 halves execution time,
083 drops from 60s to 5s, etc - while not introducing test
regressions.
This allows us to remove the old xfssyncd threads and infrastructure
as they are no longer used.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
All of the work xfssyncd does is background functionality. There is
no need for a thread per filesystem to do this work - it can al be
managed by a global workqueue now they manage concurrency
effectively.
Introduce a new gglobal xfssyncd workqueue, and convert the periodic
work to use this new functionality. To do this, use a delayed work
construct to schedule the next running of the periodic sync work
for the filesystem. When the sync work is complete, queue a new
delayed work for the next running of the sync work.
For laptop mode, we wait on completion for the sync works, so ensure
that the sync work queuing interface can flush and wait for work to
complete to enable the work queue infrastructure to replace the
current sequence number and wakeup that is used.
Because the sync work does non-trivial amounts of work, mark the
new work queue as CPU intensive.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
When formatting an inode item, we have to allocate a separate buffer
to hold extents when there are delayed allocation extents on the
inode and it is in extent format. The allocation size is derived
from the in-core data fork representation, which accounts for
delayed allocation extents, while the on-disk representation does
not contain any delalloc extents.
As a result of this mismatch, the allocated buffer can be far larger
than needed to hold the real extent list which, due to the fact the
inode is in extent format, is limited to the size of the literal
area of the inode. However, we can have thousands of delalloc
extents, resulting in an allocation size orders of magnitude larger
than is needed to hold all the real extents.
Fix this by limiting the size of the buffer being allocated to the
size of the literal area of the inodes in the filesystem (i.e. the
maximum size an inode fork can grow to).
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
Commit 1018b5c016 ("Set rt->rt_iif more
sanely on output routes.") breaks rt_is_{output,input}_route.
This became the cause to return "IP_PKTINFO's ->ipi_ifindex == 0".
To fix it, this does:
1) Add "int rt_route_iif;" to struct rtable
2) For input routes, always set rt_route_iif to same value as rt_iif
3) For output routes, always set rt_route_iif to zero. Set rt_iif
as it is done currently.
4) Change rt_is_{output,input}_route() to test rt_route_iif
Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since 1c48a5c93d (dt: Eliminate of_platform_{,un}register_driver)
mpc8xxx_wdt no longer builds as it tries to refer to a 'match' variable
rather than ofdev->dev.of_match that it checks just before.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
When attempting an initial mount, we should only attempt other
authflavors if AUTH_UNIX receives a NFS4ERR_WRONGSEC error.
This allows other errors to be passed back to userspace programs.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* 'fbdev-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/fbdev-2.6:
efifb: Add override for 11" Macbook Air 3,1
efifb: Support overriding fields FW tells us with the DMI data.
fb: Reduce priority of resource conflict message
savagefb: Remove obsolete else clause in savage_setup_i2c_bus
savagefb: Set up I2C based on chip family instead of card id
savagefb: Replace magic register address with define
drivers/video/bfin-lq035q1-fb.c: introduce missing kfree
video: s3c-fb: fix checkpatch errors and warning
efifb: support AMD Radeon HD 6490
s3fb: fix Virge/GX2
fbcon: Remove unused 'display *p' variable from fb_flashcursor()
fbdev: sh_mobile_lcdcfb: fix module lock acquisition
fbdev: sh_mobile_lcdcfb: add blanking support
viafb: initialize margins correct
viafb: refresh rate bug collection
sh: mach-ap325rxa: move backlight control code
sh: mach-ecovec24: support for main lcd backlight
* 'rmobile-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6:
ARM: arch-shmobile: only run FSI init on respective boards
ARM: arch-shmobile: only run HDMI init on respective boards
ARM: mach-shmobile: Correctly check for CONFIG_MACH_MACKEREL
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86-32, fpu: Fix FPU exception handling on non-SSE systems
x86, hibernate: Initialize mmu_cr4_features during boot
x86-32, NUMA: Fix ACPI NUMA init broken by recent x86-64 change
x86: visws: Fixup irq overhaul fallout
* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: Clean up rebalance_domains() load-balance interval calculation
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86/mrst/vrtc: Fix boot crash in mrst_rtc_init()
rtc, x86/mrst/vrtc: Fix boot crash in rtc_read_alarm()
* 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
genirq: Fix cpumask leak in __setup_irq()
* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf probe: Fix listing incorrect line number with inline function
perf probe: Fix to find recursively inlined function
perf probe: Fix multiple --vars options behavior
perf probe: Fix to remove redundant close
perf probe: Fix to ensure function declared file
* 'for-linus' of git://git.infradead.org/ubifs-2.6:
UBI: do not select KALLSYMS_ALL
UBI: do not compare array with NULL
UBI: check if we are in RO mode in the erase routine
UBIFS: fix debugging failure in dbg_check_space_info
UBIFS: fix error path in dbg_debugfs_init_fs
UBIFS: unify error path dbg_debugfs_init_fs
UBIFS: do not select KALLSYMS_ALL
UBIFS: fix assertion warnings
UBIFS: fix oops on error path in read_pnode
UBIFS: do not read flash unnecessarily
There's a race condition in stop_queue() in some drivers -
if drv_data->queue is empty, but drv_data->busy is still set
(or opposite situation) stop_queue will return -EBUSY.
So fix loop condition to check that both drv_data->queue is empty
and drv_data->busy is not set.
This patch affects following drivers:
pxa2xx_spi
spi_bfin5xx
amba-pl022
dw_spi
Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Acked-by: Eric Miao <eric.y.miao@gmail.com>
Acked-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid:
HID: Add support for CH Pro Throttle
HID: hid-magicmouse: Increase evdev buffer size
HID: add FF support for Logitech G25/G27
HID: roccat: Add support for wireless variant of Pyra
HID: Fix typo Keyoutch -> Keytouch
HID: add support for Skycable 0x3f07 wireless presenter
Fix build failure issue for hv_mouse
When build 2.6.39-rc1 kernel, it will be blocked at build hv_mouse.
drivers/staging/hv/hv_mouse.c: In function ‘ReleaseInputDevice’:
drivers/staging/hv/hv_mouse.c:293: error: implicit declaration of function ‘udelay’
Signed-off-by: Youquan Song <youquan.song@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The normal mmap paths all avoid creating a mapping where the pgoff
inside the mapping could wrap around due to overflow. However, an
expanding mremap() can take such a non-wrapping mapping and make it
bigger and cause a wrapping condition.
Noticed by Robert Swiecki when running a system call fuzzer, where it
caused a BUG_ON() due to terminally confusing the vma_prio_tree code. A
vma dumping patch by Hugh then pinpointed the crazy wrapped case.
Reported-and-tested-by: Robert Swiecki <robert@swiecki.net>
Acked-by: Hugh Dickins <hughd@google.com>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The MCP7x hardware computes the audio infoframe channel count
automatically, but requires the audio driver to set the audio
infoframe checksum manually via the Nv_VERB_SET_Info_Frame_Checksum
control verb.
When audio starts playing, nvhdmi_8ch_7x_pcm_prepare sets the checksum
to (0x71 - chan - chanmask). For example, for 2ch audio, chan == 1
and chanmask == 0 so the checksum is set to 0x70. When audio playback
finishes and the device is closed, nvhdmi_8ch_7x_pcm_close resets the
channel formats, causing the channel count to revert to 8ch. Since
the checksum is not reset, the hardware starts generating audio
infoframes with invalid checksums. This causes some displays to blank
the video.
Fix this by updating the checksum and channel mask when the device is
closed and also when it is first initialized. In addition, make sure
that the channel mask is appropriate for an 8ch infoframe by setting
it to 0x13 (FL FR LFE FC RL RR RLC RRC).
Signed-off-by: Aaron Plattner <aplattner@nvidia.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Dave Chinner
OGAWA Hirofumi
Linus Torvalds
Peter Korsgaard
Bryan Schumaker
Vasily Khoruzhick
Grant Likely
Youquan Song
Takashi Iwai
David Henningsson
Aaron Plattner
Feng Tang