* xfs_efi_item slab leak
@ 2010-12-17 18:04 Nick Piggin
2010-12-18 0:02 ` Dave Chinner
0 siblings, 1 reply; 4+ messages in thread
From: Nick Piggin @ 2010-12-17 18:04 UTC (permalink / raw)
To: xfs
After running xfstests and then unmounting and rmmodding xfs, I get this bug
[ 3473.194517] =============================================================================
[ 3473.194697] BUG xfs_efi_item: Objects remaining on kmem_cache_close()
[ 3473.194793] -----------------------------------------------------------------------------
[ 3473.194794]
[ 3473.195056] INFO: Slab 0xffffea0006ad5e00 objects=35 used=2
fp=0xffff8801e8640e80 flags=0xc000000000004080
[ 3473.195201] Pid: 22011, comm: rmmod Not tainted 2.6.37-rc6+ #116
[ 3473.195201] Call Trace:
[ 3473.195201] [<ffffffff8110bcf1>] slab_err+0x91/0xa0
[ 3473.195201] [<ffffffff8110f088>] ? __slab_alloc+0x138/0x480
[ 3473.195201] [<ffffffff8108081d>] ? trace_hardirqs_off+0xd/0x10
[ 3473.195201] [<ffffffff811112a4>] ? kmem_cache_destroy+0x154/0x3e0
[ 3473.195201] [<ffffffff8108081d>] ? trace_hardirqs_off+0xd/0x10
[ 3473.195201] [<ffffffff811105a4>] ? __kmalloc+0x104/0x220
[ 3473.195201] [<ffffffff811112c7>] kmem_cache_destroy+0x177/0x3e0
[ 3473.195201] [<ffffffffa0086e47>] xfs_destroy_zones+0x37/0x100 [xfs]
[ 3473.195201] [<ffffffffa0088f88>] exit_xfs_fs+0x33/0x3d [xfs]
[ 3473.195201] [<ffffffff8109060a>] sys_delete_module+0x17a/0x270
[ 3473.195201] [<ffffffff8160340f>] ? lockdep_sys_exit_thunk+0x35/0x67
[ 3473.195201] [<ffffffff810030bb>] system_call_fastpath+0x16/0x1b
[ 3473.195201] INFO: Object 0xffff8801e8640910 @offset=2320
[ 3473.195201] INFO: Allocated in kmem_zone_alloc+0x8c/0xd0 [xfs]
age=2263917 cpu=15 pid=4514
[ 3473.195201] INFO: Freed in xfs_efi_item_free+0x19/0x30 [xfs]
age=2263918 cpu=15 pid=4514
[ 3473.195201] INFO: Object 0xffff8801e86427e0 @offset=10208
[ 3473.195201] INFO: Allocated in kmem_zone_alloc+0x8c/0xd0 [xfs]
age=2103034 cpu=15 pid=22329
[ 3473.195201] INFO: Freed in xfs_efi_item_free+0x19/0x30 [xfs]
age=2103034 cpu=15 pid=22329
[ 3473.197616] SLUB xfs_efi_item: kmem_cache_destroy called for cache
that still has objects.
[ 3473.197794] Pid: 22011, comm: rmmod Not tainted 2.6.37-rc6+ #116
[ 3473.197892] Call Trace:
[ 3473.197980] [<ffffffff811114a7>] kmem_cache_destroy+0x357/0x3e0
[ 3473.198120] [<ffffffffa0086e47>] xfs_destroy_zones+0x37/0x100 [xfs]
[ 3473.198240] [<ffffffffa0088f88>] exit_xfs_fs+0x33/0x3d [xfs]
[ 3473.198337] [<ffffffff8109060a>] sys_delete_module+0x17a/0x270
[ 3473.198435] [<ffffffff8160340f>] ? lockdep_sys_exit_thunk+0x35/0x67
[ 3473.198534] [<ffffffff810030bb>] system_call_fastpath+0x16/0x1b
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: xfs_efi_item slab leak
2010-12-17 18:04 xfs_efi_item slab leak Nick Piggin
@ 2010-12-18 0:02 ` Dave Chinner
2010-12-20 11:31 ` Christoph Hellwig
0 siblings, 1 reply; 4+ messages in thread
From: Dave Chinner @ 2010-12-18 0:02 UTC (permalink / raw)
To: Nick Piggin; +Cc: xfs
On Sat, Dec 18, 2010 at 05:04:40AM +1100, Nick Piggin wrote:
> After running xfstests and then unmounting and rmmodding xfs, I get this bug
>
> [ 3473.194517] =============================================================================
> [ 3473.194697] BUG xfs_efi_item: Objects remaining on kmem_cache_close()
> [ 3473.194793] -----------------------------------------------------------------------------
> [ 3473.194794]
> [ 3473.195056] INFO: Slab 0xffffea0006ad5e00 objects=35 used=2
> fp=0xffff8801e8640e80 flags=0xc000000000004080
> [ 3473.195201] Pid: 22011, comm: rmmod Not tainted 2.6.37-rc6+ #116
> [ 3473.195201] Call Trace:
> [ 3473.195201] [<ffffffff8110bcf1>] slab_err+0x91/0xa0
> [ 3473.195201] [<ffffffff8110f088>] ? __slab_alloc+0x138/0x480
> [ 3473.195201] [<ffffffff8108081d>] ? trace_hardirqs_off+0xd/0x10
> [ 3473.195201] [<ffffffff811112a4>] ? kmem_cache_destroy+0x154/0x3e0
> [ 3473.195201] [<ffffffff8108081d>] ? trace_hardirqs_off+0xd/0x10
> [ 3473.195201] [<ffffffff811105a4>] ? __kmalloc+0x104/0x220
> [ 3473.195201] [<ffffffff811112c7>] kmem_cache_destroy+0x177/0x3e0
> [ 3473.195201] [<ffffffffa0086e47>] xfs_destroy_zones+0x37/0x100 [xfs]
> [ 3473.195201] [<ffffffffa0088f88>] exit_xfs_fs+0x33/0x3d [xfs]
> [ 3473.195201] [<ffffffff8109060a>] sys_delete_module+0x17a/0x270
> [ 3473.195201] [<ffffffff8160340f>] ? lockdep_sys_exit_thunk+0x35/0x67
> [ 3473.195201] [<ffffffff810030bb>] system_call_fastpath+0x16/0x1b
> [ 3473.195201] INFO: Object 0xffff8801e8640910 @offset=2320
> [ 3473.195201] INFO: Allocated in kmem_zone_alloc+0x8c/0xd0 [xfs]
> age=2263917 cpu=15 pid=4514
> [ 3473.195201] INFO: Freed in xfs_efi_item_free+0x19/0x30 [xfs]
> age=2263918 cpu=15 pid=4514
> [ 3473.195201] INFO: Object 0xffff8801e86427e0 @offset=10208
> [ 3473.195201] INFO: Allocated in kmem_zone_alloc+0x8c/0xd0 [xfs]
> age=2103034 cpu=15 pid=22329
> [ 3473.195201] INFO: Freed in xfs_efi_item_free+0x19/0x30 [xfs]
> age=2103034 cpu=15 pid=22329
> [ 3473.197616] SLUB xfs_efi_item: kmem_cache_destroy called for cache
> that still has objects.
> [ 3473.197794] Pid: 22011, comm: rmmod Not tainted 2.6.37-rc6+ #116
> [ 3473.197892] Call Trace:
> [ 3473.197980] [<ffffffff811114a7>] kmem_cache_destroy+0x357/0x3e0
> [ 3473.198120] [<ffffffffa0086e47>] xfs_destroy_zones+0x37/0x100 [xfs]
> [ 3473.198240] [<ffffffffa0088f88>] exit_xfs_fs+0x33/0x3d [xfs]
> [ 3473.198337] [<ffffffff8109060a>] sys_delete_module+0x17a/0x270
> [ 3473.198435] [<ffffffff8160340f>] ? lockdep_sys_exit_thunk+0x35/0x67
> [ 3473.198534] [<ffffffff810030bb>] system_call_fastpath+0x16/0x1b
Ok, after unmounting everything I can see two items in remaining in
the xfs_efi_item cache after an xfstests run. Іt's probably to do
with a filesystem shutdown not cleaning them up correctly. I'll put
it on my list of stuff to track down.
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: xfs_efi_item slab leak
2010-12-18 0:02 ` Dave Chinner
@ 2010-12-20 11:31 ` Christoph Hellwig
2010-12-21 0:54 ` Dave Chinner
0 siblings, 1 reply; 4+ messages in thread
From: Christoph Hellwig @ 2010-12-20 11:31 UTC (permalink / raw)
To: Dave Chinner; +Cc: Nick Piggin, xfs
On Sat, Dec 18, 2010 at 11:02:45AM +1100, Dave Chinner wrote:
> Ok, after unmounting everything I can see two items in remaining in
> the xfs_efi_item cache after an xfstests run. ??t's probably to do
> with a filesystem shutdown not cleaning them up correctly. I'll put
> it on my list of stuff to track down.
Does it still happen with your working tree? The incorrect handling
of IOP_UNPING with remove == 1 looks like it could cause something
similar.
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: xfs_efi_item slab leak
2010-12-20 11:31 ` Christoph Hellwig
@ 2010-12-21 0:54 ` Dave Chinner
0 siblings, 0 replies; 4+ messages in thread
From: Dave Chinner @ 2010-12-21 0:54 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Nick Piggin, xfs
On Mon, Dec 20, 2010 at 06:31:39AM -0500, Christoph Hellwig wrote:
> On Sat, Dec 18, 2010 at 11:02:45AM +1100, Dave Chinner wrote:
> > Ok, after unmounting everything I can see two items in remaining in
> > the xfs_efi_item cache after an xfstests run. ??t's probably to do
> > with a filesystem shutdown not cleaning them up correctly. I'll put
> > it on my list of stuff to track down.
>
> Does it still happen with your working tree? The incorrect handling
> of IOP_UNPING with remove == 1 looks like it could cause something
> similar.
I haven't checked the current patch set - just on the kernel that I
was currently running - I'll do more analysis later this afternoon.
I agree that the IOP_UNPIN() handling is a likely culprit, though.
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2010-12-21 0:52 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-12-17 18:04 xfs_efi_item slab leak Nick Piggin
2010-12-18 0:02 ` Dave Chinner
2010-12-20 11:31 ` Christoph Hellwig
2010-12-21 0:54 ` Dave Chinner
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox