might_sleep in quota code path as of 2.6.6-rc2, may be fixed already

might_sleep in quota code path as of 2.6.6-rc2, may be fixed already

[Paul P Komkoff Jr]

Hello.

Have you seen that?

May 12 16:01:14 ns Debug: sleeping function called from invalid context at include/asm/semaphore.h:119
May 12 16:01:14 ns in_atomic():1, irqs_disabled():0
May 12 16:01:14 ns Call Trace:
May 12 16:01:14 ns [<c01193e8>] __might_sleep+0xa5/0xaf
May 12 16:01:14 ns [<c0176a3a>] dquot_release+0x32/0xb0
May 12 16:01:14 ns [<c0176fad>] dqput+0x126/0x181
May 12 16:01:14 ns [<c017747c>] remove_inode_dquot_ref+0xae/0xb8
May 12 16:01:14 ns [<c016568e>] remove_dquot_ref+0x6a/0x157
May 12 16:01:14 ns [<c0177506>] drop_dquot_ref+0x4b/0x6b
May 12 16:01:14 ns [<c0178470>] vfs_quota_off+0x85/0x116
May 12 16:01:14 ns [<c017a551>] do_quotactl+0x131/0x34a
May 12 16:01:14 ns [<c0154f65>] bdev_set+0x0/0xc
May 12 16:01:14 ns [<c0154f55>] bdev_test+0x0/0x10
May 12 16:01:14 ns [<c016221d>] dput+0x75/0x211
May 12 16:01:14 ns [<c0159dca>] path_release+0xd/0x2b
May 12 16:01:14 ns [<c0155b6d>] lookup_bdev+0x6c/0x7b
May 12 16:01:14 ns [<c017a823>] sys_quotactl+0xb9/0xd2
May 12 16:01:14 ns [<c0105bad>] sysenter_past_esp+0x52/0x71

Only once, in quotacheck -a after umount ; e2fsck -fD ; mount

-- 
Paul P 'Stingray' Komkoff Jr // http://stingr.net/key <- my pgp key
 This message represents the official view of the voices in my head

Re: might_sleep in quota code path as of 2.6.6-rc2, may be fixed already

[Jan Kara]

  Hello,

> Have you seen that?
  No. Thanks for the trace. Which kernel do you use?

								Honza

> 
> May 12 16:01:14 ns Debug: sleeping function called from invalid context at include/asm/semaphore.h:119
> May 12 16:01:14 ns in_atomic():1, irqs_disabled():0
> May 12 16:01:14 ns Call Trace:
> May 12 16:01:14 ns [<c01193e8>] __might_sleep+0xa5/0xaf
> May 12 16:01:14 ns [<c0176a3a>] dquot_release+0x32/0xb0
> May 12 16:01:14 ns [<c0176fad>] dqput+0x126/0x181
> May 12 16:01:14 ns [<c017747c>] remove_inode_dquot_ref+0xae/0xb8
> May 12 16:01:14 ns [<c016568e>] remove_dquot_ref+0x6a/0x157
> May 12 16:01:14 ns [<c0177506>] drop_dquot_ref+0x4b/0x6b
> May 12 16:01:14 ns [<c0178470>] vfs_quota_off+0x85/0x116
> May 12 16:01:14 ns [<c017a551>] do_quotactl+0x131/0x34a
> May 12 16:01:14 ns [<c0154f65>] bdev_set+0x0/0xc
> May 12 16:01:14 ns [<c0154f55>] bdev_test+0x0/0x10
> May 12 16:01:14 ns [<c016221d>] dput+0x75/0x211
> May 12 16:01:14 ns [<c0159dca>] path_release+0xd/0x2b
> May 12 16:01:14 ns [<c0155b6d>] lookup_bdev+0x6c/0x7b
> May 12 16:01:14 ns [<c017a823>] sys_quotactl+0xb9/0xd2
> May 12 16:01:14 ns [<c0105bad>] sysenter_past_esp+0x52/0x71
> 
> Only once, in quotacheck -a after umount ; e2fsck -fD ; mount
> 
> -- 
> Paul P 'Stingray' Komkoff Jr // http://stingr.net/key <- my pgp key
>  This message represents the official view of the voices in my head
-- 
Jan Kara <jack@suse.cz>
SuSE CR Labs

Re: might_sleep in quota code path as of 2.6.6-rc2, may be fixed already

[Paul P Komkoff Jr]

Replying to Jan Kara:
>   Hello,
> 
> > Have you seen that?
>   No. Thanks for the trace. Which kernel do you use?

that was in 2.6.6-rc2


-- 
Paul P 'Stingray' Komkoff Jr // http://stingr.net/key <- my pgp key
 This message represents the official view of the voices in my head

Re: might_sleep in quota code path as of 2.6.6-rc2, may be fixed already

[Jan Kara]

> Replying to Jan Kara:
> >   Hello,
> > 
> > > Have you seen that?
> >   No. Thanks for the trace. Which kernel do you use?
> 
> that was in 2.6.6-rc2
  OK. I fixed a bug which could cause your problem and the fix is in
2.6.6 so please try that kernel.

						Thanks for report
								Honza
-- 
Jan Kara <jack@suse.cz>
SuSE CR Labs