From: Jan Kara <jack@suse.cz>
To: Valdis.Kletnieks@vt.edu
Cc: Andrew Morton <akpm@linux-foundation.org>,
Theodore Ts'o <tytso@mit.edu>,
Andreas Dilger <adilger.kernel@dilger.ca>,
Jan Kara <jack@suse.cz>,
linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org
Subject: Re: mmotm 2010-11-09 - lockdep splat in ext4/quota code
Date: Sat, 13 Nov 2010 00:33:43 +0100 [thread overview]
Message-ID: <20101112233343.GA12153@quack.suse.cz> (raw)
In-Reply-To: <19688.1289503541@localhost>
On Thu 11-11-10 14:25:41, Valdis.Kletnieks@vt.edu wrote:
> On Tue, 09 Nov 2010 15:31:15 PST, akpm@linux-foundation.org said:
> > The mm-of-the-moment snapshot 2010-11-09-15-31 has been uploaded to
> >
> > http://userweb.kernel.org/~akpm/mmotm/
>
> Still seeing this just after quotas get turned on ext4. First seen in
> 2.6.36-mmotm1022, but problem has apparently been there since
> 2.6.36-rc5-mmotm0922 (oldest kernel I have lying around).
Thanks for the report. In fact, it is a false positive caused by the fact
that EXT4(inode)->i_data_sem for quota file inode would need a special
locking class (or subclass) but does not have one. I'll have a look into
implementing that sometime next week hopefully.
Honza
> The last few relevant commands from rc.sysinit:
>
> /sbin/quotaon -aug
> dd if=/dev/urandom of=/var/lib/random-seed count=1 bs=512 <- this triggers the trace
>
> [ 48.967833] =======================================================
> [ 48.968162] [ INFO: possible circular locking dependency detected ]
> [ 48.968411] 2.6.37-rc1-mmotm1109 #1
> [ 48.968553] -------------------------------------------------------
> [ 48.968592] dd/3253 is trying to acquire lock:
> [ 48.968592] (&s->s_dquot.dqio_mutex){+.+...}, at: [<ffffffff81141b0a>] dquot_commit+0x26/0xda
> [ 48.968592]
> [ 48.968592] but task is already holding lock:
> [ 48.968592] (&s->s_dquot.dqptr_sem){++++..}, at: [<ffffffff81143e70>] __dquot_free_space+0x66/0x16b
> [ 48.968592]
> [ 48.968592] which lock already depends on the new lock.
> [ 48.968592]
> [ 48.968592]
> [ 48.968592] the existing dependency chain (in reverse order) is:
> [ 48.968592]
> [ 48.968592] -> #2 (&s->s_dquot.dqptr_sem){++++..}:
> [ 48.968592] [<ffffffff8106b441>] lock_acquire+0x100/0x126
> [ 48.968592] [<ffffffff81565615>] down_read+0x42/0x51
> [ 48.968592] [<ffffffff81143e70>] __dquot_free_space+0x66/0x16b
> [ 48.968592] [<ffffffff81199a70>] dquot_free_block+0x19/0x2a
> [ 48.968592] [<ffffffff8119fc4f>] ext4_free_blocks+0x75b/0x810
> [ 48.968592] [<ffffffff81196f0f>] ext4_ext_truncate+0x3af/0x853
> [ 48.968592] [<ffffffff8117a02e>] ext4_truncate+0x98/0x51a
> [ 48.968592] [<ffffffff810c46a4>] vmtruncate+0x3f/0x4b
> [ 48.968592] [<ffffffff8117af55>] ext4_setattr+0x305/0x3a7
> [ 48.968592] [<ffffffff8110c6d9>] notify_change+0x1a1/0x2a1
> [ 48.968592] [<ffffffff810f49fa>] do_truncate+0x67/0x84
> [ 48.968592] [<ffffffff811029f5>] do_last+0x4ce/0x5b2
> [ 48.968592] [<ffffffff81102d21>] do_filp_open+0x248/0x64a
> [ 48.968592] [<ffffffff810f5481>] do_sys_open+0x60/0xfb
> [ 48.968592] [<ffffffff810f5537>] sys_open+0x1b/0x1d
> [ 48.968592] [<ffffffff810027eb>] system_call_fastpath+0x16/0x1b
> [ 48.968592]
> [ 48.968592] -> #1 (&ei->i_data_sem){++++..}:
> [ 48.968592] [<ffffffff8106b441>] lock_acquire+0x100/0x126
> [ 48.968592] [<ffffffff81565615>] down_read+0x42/0x51
> [ 48.968592] [<ffffffff81179383>] ext4_map_blocks+0x3e/0x20e
> [ 48.968592] [<ffffffff81179e0b>] ext4_getblk+0x71/0x19b
> [ 48.968592] [<ffffffff81179f46>] ext4_bread+0x11/0x61
> [ 48.968592] [<ffffffff8118b297>] ext4_quota_read+0x86/0xf0
> [ 48.968592] [<ffffffff81145131>] v2_read_header+0x1e/0x4f
> [ 48.968592] [<ffffffff811455de>] v2_read_file_info+0x20/0x2ae
> [ 48.968592] [<ffffffff81144423>] vfs_load_quota_inode+0x292/0x448
> [ 48.968592] [<ffffffff8114480e>] dquot_quota_on_path+0x4c/0x55
> [ 48.968592] [<ffffffff8118bc97>] ext4_quota_on+0x123/0x175
> [ 48.968592] [<ffffffff81146cf2>] do_quotactl+0x1d1/0x465
> [ 48.968592] [<ffffffff81147083>] sys_quotactl+0xfd/0x11a
> [ 48.968592] [<ffffffff810027eb>] system_call_fastpath+0x16/0x1b
> [ 48.968592]
> [ 48.968592] -> #0 (&s->s_dquot.dqio_mutex){+.+...}:
> [ 48.968592] [<ffffffff8106ac2d>] __lock_acquire+0xa4c/0xd4e
> [ 48.968592] [<ffffffff8106b441>] lock_acquire+0x100/0x126
> [ 48.968592] [<ffffffff81564bbf>] __mutex_lock_common+0x5d/0x56c
> [ 48.968592] [<ffffffff81565174>] mutex_lock_nested+0x34/0x39
> [ 48.968592] [<ffffffff81141b0a>] dquot_commit+0x26/0xda
> [ 48.968592] [<ffffffff8118d622>] ext4_write_dquot+0x65/0x87
> [ 48.968592] [<ffffffff8118d682>] ext4_mark_dquot_dirty+0x3e/0x49
> [ 48.968592] [<ffffffff81141790>] mark_all_dquot_dirty+0x22/0x48
> [ 48.968592] [<ffffffff81143f43>] __dquot_free_space+0x139/0x16b
> [ 48.968592] [<ffffffff81199a70>] dquot_free_block+0x19/0x2a
> [ 48.968592] [<ffffffff8119fc4f>] ext4_free_blocks+0x75b/0x810
> [ 48.968592] [<ffffffff81196f0f>] ext4_ext_truncate+0x3af/0x853
> [ 48.968592] [<ffffffff8117a02e>] ext4_truncate+0x98/0x51a
> [ 48.968592] [<ffffffff810c46a4>] vmtruncate+0x3f/0x4b
> [ 48.968592] [<ffffffff8117af55>] ext4_setattr+0x305/0x3a7
> [ 48.968592] [<ffffffff8110c6d9>] notify_change+0x1a1/0x2a1
> [ 48.968592] [<ffffffff810f49fa>] do_truncate+0x67/0x84
> [ 48.968592] [<ffffffff811029f5>] do_last+0x4ce/0x5b2
> [ 48.968592] [<ffffffff81102d21>] do_filp_open+0x248/0x64a
> [ 48.968592] [<ffffffff810f5481>] do_sys_open+0x60/0xfb
> [ 48.968592] [<ffffffff810f5537>] sys_open+0x1b/0x1d
> [ 48.968592] [<ffffffff810027eb>] system_call_fastpath+0x16/0x1b
> [ 48.968592]
> [ 48.968592] other info that might help us debug this:
> [ 48.968592]
> [ 49.458584] 5 locks held by dd/3253:
> [ 49.458584] #0: (&sb->s_type->i_mutex_key#13){+.+.+.}, at: [<ffffffff810f49ee>] do_truncate+0x5b/0x84
> [ 49.458584] #1: (&sb->s_type->i_alloc_sem_key#5){+.+...}, at: [<ffffffff8110c6c1>] notify_change+0x189/0x2a1
> [ 49.458584] #2: (jbd2_handle){+.+...}, at: [<ffffffff811ad5b4>] start_this_handle+0x57a/0x605
> [ 49.458584] #3: (&ei->i_data_sem){++++..}, at: [<ffffffff81196bf1>] ext4_ext_truncate+0x91/0x853
> [ 49.458584] #4: (&s->s_dquot.dqptr_sem){++++..}, at: [<ffffffff81143e70>] __dquot_free_space+0x66/0x16b
> [ 49.458584]
> [ 49.458584] stack backtrace:
> [ 49.458584] Pid: 3253, comm: dd Not tainted 2.6.37-rc1-mmotm1109 #1
> [ 49.458584] Call Trace:
> [ 49.458584] [<ffffffff81069591>] print_circular_bug+0xa8/0xb7
> [ 49.458584] [<ffffffff8106ac2d>] __lock_acquire+0xa4c/0xd4e
> [ 49.458584] [<ffffffff81141b0a>] ? dquot_commit+0x26/0xda
> [ 49.458584] [<ffffffff8106b441>] lock_acquire+0x100/0x126
> [ 49.458584] [<ffffffff81141b0a>] ? dquot_commit+0x26/0xda
> [ 49.458584] [<ffffffff81564bbf>] __mutex_lock_common+0x5d/0x56c
> [ 49.458584] [<ffffffff81141b0a>] ? dquot_commit+0x26/0xda
> [ 49.458584] [<ffffffff81141ef2>] ? dquot_mark_dquot_dirty+0x27/0x7a
> [ 49.458584] [<ffffffff81141b0a>] ? dquot_commit+0x26/0xda
> [ 49.458584] [<ffffffff81565174>] mutex_lock_nested+0x34/0x39
> [ 49.458584] [<ffffffff81141b0a>] dquot_commit+0x26/0xda
> [ 49.458584] [<ffffffff8118d622>] ext4_write_dquot+0x65/0x87
> [ 49.458584] [<ffffffff8118d682>] ext4_mark_dquot_dirty+0x3e/0x49
> [ 49.458584] [<ffffffff81141790>] mark_all_dquot_dirty+0x22/0x48
> [ 49.458584] [<ffffffff81143f43>] __dquot_free_space+0x139/0x16b
> [ 49.458584] [<ffffffff81199a70>] dquot_free_block+0x19/0x2a
> [ 49.458584] [<ffffffff8119fc4f>] ext4_free_blocks+0x75b/0x810
> [ 49.458584] [<ffffffff811ada1d>] ? jbd2_journal_extend+0x5d/0xb8
> [ 49.458584] [<ffffffff81196f0f>] ext4_ext_truncate+0x3af/0x853
> [ 49.458584] [<ffffffff810a5759>] ? trace_preempt_on+0x15/0x28
> [ 49.458584] [<ffffffff810d63f6>] ? unmap_mapping_range+0x279/0x288
> [ 49.458584] [<ffffffff8117a02e>] ext4_truncate+0x98/0x51a
> [ 49.458584] [<ffffffff810d63f6>] ? unmap_mapping_range+0x279/0x288
> [ 49.458584] [<ffffffff8106b861>] ? trace_hardirqs_on_caller+0x117/0x13b
> [ 49.458584] [<ffffffff8106b892>] ? trace_hardirqs_on+0xd/0xf
> [ 49.458584] [<ffffffff810c46a4>] vmtruncate+0x3f/0x4b
> [ 49.458584] [<ffffffff8117af55>] ext4_setattr+0x305/0x3a7
> [ 49.458584] [<ffffffff8110c6d9>] notify_change+0x1a1/0x2a1
> [ 49.458584] [<ffffffff810f49fa>] do_truncate+0x67/0x84
> [ 49.458584] [<ffffffff810ff92d>] ? get_write_access+0x49/0x50
> [ 49.458584] [<ffffffff81566794>] ? _raw_spin_unlock+0x30/0x69
> [ 49.458584] [<ffffffff8156a040>] ? sub_preempt_count+0x35/0x48
> [ 49.458584] [<ffffffff811e0833>] ? security_path_truncate+0x48/0x4d
> [ 49.458584] [<ffffffff811029f5>] do_last+0x4ce/0x5b2
> [ 49.458584] [<ffffffff81102d21>] do_filp_open+0x248/0x64a
> [ 49.458584] [<ffffffff810a5759>] ? trace_preempt_on+0x15/0x28
> [ 49.458584] [<ffffffff8110d8ce>] ? alloc_fd+0x17c/0x18e
> [ 49.458584] [<ffffffff81566794>] ? _raw_spin_unlock+0x30/0x69
> [ 49.458584] [<ffffffff8156a040>] ? sub_preempt_count+0x35/0x48
> [ 49.458584] [<ffffffff8110d8ce>] ? alloc_fd+0x17c/0x18e
> [ 49.458584] [<ffffffff810f5481>] do_sys_open+0x60/0xfb
> [ 49.458584] [<ffffffff81565fde>] ? trace_hardirqs_on_thunk+0x3a/0x3f
> [ 49.458584] [<ffffffff810f5537>] sys_open+0x1b/0x1d
> [ 49.458584] [<ffffffff810027eb>] system_call_fastpath+0x16/0x1b
>
>
--
Jan Kara <jack@suse.cz>
SUSE Labs, CR
next prev parent reply other threads:[~2010-11-12 23:33 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-11-09 23:31 mmotm 2010-11-09-15-31 uploaded akpm
2010-11-10 1:01 ` mmotm 2010-11-09-15-31 uploaded (sound/soc/soc-core.c) Randy Dunlap
2010-11-11 19:25 ` mmotm 2010-11-09 - lockdep splat in ext4/quota code Valdis.Kletnieks
2010-11-12 23:33 ` Jan Kara [this message]
2010-11-25 5:35 ` Valdis.Kletnieks
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20101112233343.GA12153@quack.suse.cz \
--to=jack@suse.cz \
--cc=Valdis.Kletnieks@vt.edu \
--cc=adilger.kernel@dilger.ca \
--cc=akpm@linux-foundation.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=tytso@mit.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).