* XFS with project quota support in Kernel 2.6.28(.9)
@ 2011-11-14 14:36 Maarten Vergouwe
2011-11-19 18:48 ` Christoph Hellwig
0 siblings, 1 reply; 2+ messages in thread
From: Maarten Vergouwe @ 2011-11-14 14:36 UTC (permalink / raw)
To: xfs
[-- Attachment #1.1: Type: text/plain, Size: 3065 bytes --]
Hi
I have problems in writes that hang for more than 7 seconds because of the
the kthreadd daemon keeping a mutex locked for too long.
Is there someone who might be able to help me with that?
I saw that a lot of changes were done on the XFS quota support in later
kernels, but the patch list is too long between my kernel and the changed
quota system.
process 411 ([khubd], kthreadd)
pid = 411, tgid = 411, parent = 2, real parent = 2
baduaddr = 0x00000000, badvaddr = 0x00000000, state = 2, flags =
0x80208040
policy = 0, prio = 115, static prio = 115, normal prio = 115, rt prio = 0
pc = vmlinux_flash.elf/context_switch+0x3ac (kernel/sched.c:2668)
schedule (kernel/sched.c:4496)
kernel stack
vmlinux_flash.elf/schedule_timeout+0x24 (kernel/timer.c:1264)
vmlinux_flash.elf/do_wait_for_common+0xd8 (kernel/sched.c:4715)
wait_for_common (kernel/sched.c:4732)
vmlinux_flash.elf/xfs_qm_sync+0x110 (fs/xfs/quota/xfs_qm.c:1076)
vmlinux_flash.elf/xfs_sync+0x1c (fs/xfs/xfs_vfsops.c:255)
vmlinux_flash.elf/xfs_fs_sync_super+0x3c
(fs/xfs/linux-2.6/xfs_super.c:1202)
vmlinux_flash.elf/fsync_super+0xc (fs/super.c:270)
process 1082 *(this one takes too long and my watchdog doesn't like that)*
pid = 1082, tgid = 230, parent = 228, real parent = 228
baduaddr = 0x00000000, badvaddr = 0x00000000, state = 2, flags =
0x00422140
policy = 0, prio = 120, static prio = 120, normal prio = 120, rt prio = 0
pc = vmlinux_flash.elf/context_switch+0x3ac (kernel/sched.c:2668)
schedule (kernel/sched.c:4496)
kernel stack
vmlinux_flash.elf/__mutex_lock_common+0x108 (kernel/mutex.c:182)
__mutex_lock_slowpath (kernel/mutex.c:321)
vmlinux_flash.elf/xfs_trans_dqresv+0x50
(fs/xfs/quota/xfs_trans_dquot.c:628)
vmlinux_flash.elf/xfs_trans_reserve_quota_bydquots+0xc0
(fs/xfs/quota/xfs_trans_dquot.c:792)
vmlinux_flash.elf/xfs_trans_reserve_quota_nblks+0x40
(fs/xfs/quota/xfs_trans_dquot.c:847)
vmlinux_flash.elf/xfs_iomap_write_direct+0x36c (fs/xfs/xfs_iomap.c:486)
vmlinux_flash.elf/xfs_iomap+0x250 (fs/xfs/xfs_iomap.c:247)
vmlinux_flash.elf/__xfs_get_blocks+0xb8
(fs/xfs/linux-2.6/xfs_aops.c:1345)
vmlinux_flash.elf/xfs_get_blocks_direct+0x14
(fs/xfs/linux-2.6/xfs_aops.c:1426)
vmlinux_flash.elf/get_more_blocks+0x54c (fs/direct-io.c:535)
do_direct_IO (fs/direct-io.c:816)
direct_io_worker (fs/direct-io.c:998)
__blockdev_direct_IO (fs/direct-io.c:1209)
vmlinux_flash.elf/blockdev_direct_IO_no_locking+0xf0
(include/linux/fs.h:2010)
xfs_vm_direct_IO (fs/xfs/linux-2.6/xfs_aops.c:1506)
vmlinux_flash.elf/generic_file_direct_write+0x148 (mm/filemap.c:2102)
vmlinux_flash.elf/xfs_write+0x460 (fs/xfs/linux-2.6/xfs_lrw.c:745)
vmlinux_flash.elf/do_sync_write+0xc8 (fs/read_write.c:305)
vmlinux_flash.elf/sys_write+0x50 (fs/read_write.c:386)
vmlinux_flash.elf/stack_done+0xdc (arch/mips/kernel/scall32-o32.S:59)
Best regards
Maarten Vergouwe
TASS Belgium
[-- Attachment #1.2: Type: text/html, Size: 3693 bytes --]
[-- Attachment #2: Type: text/plain, Size: 121 bytes --]
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: XFS with project quota support in Kernel 2.6.28(.9)
2011-11-14 14:36 XFS with project quota support in Kernel 2.6.28(.9) Maarten Vergouwe
@ 2011-11-19 18:48 ` Christoph Hellwig
0 siblings, 0 replies; 2+ messages in thread
From: Christoph Hellwig @ 2011-11-19 18:48 UTC (permalink / raw)
To: Maarten Vergouwe; +Cc: xfs
On Mon, Nov 14, 2011 at 03:36:51PM +0100, Maarten Vergouwe wrote:
> Hi
>
> I have problems in writes that hang for more than 7 seconds because of the
> the kthreadd daemon keeping a mutex locked for too long.
> Is there someone who might be able to help me with that?
> I saw that a lot of changes were done on the XFS quota support in later
> kernels, but the patch list is too long between my kernel and the changed
> quota system.
Did you manage to check if you can reproduce the issue with a recent
kernel?
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2011-11-19 18:48 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-11-14 14:36 XFS with project quota support in Kernel 2.6.28(.9) Maarten Vergouwe
2011-11-19 18:48 ` Christoph Hellwig
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox