* [PATCH] ext4: Fix lock order during truncate.
@ 2008-07-10 8:56 Aneesh Kumar K.V
2008-07-11 23:00 ` Mingming Cao
0 siblings, 1 reply; 2+ messages in thread
From: Aneesh Kumar K.V @ 2008-07-10 8:56 UTC (permalink / raw)
To: cmm, tytso, sandeen, jack; +Cc: linux-ext4, Aneesh Kumar K.V
This can be merged to
ext4-fix-lock-inversion-in-ext4_ext_truncate.patch
=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.26-rc8 #32
-------------------------------------------------------
umount/4921 is trying to acquire lock:
(&ei->i_data_sem){----}, at: [<c01dd2c1>] ext4_get_blocks_wrap+0x29/0x120
but task is already holding lock:
(&type->s_lock_key#7){--..}, at: [<c0169211>] lock_super+0x1b/0x1d
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&type->s_lock_key#7){--..}:
[<c013c820>] __lock_acquire+0x97f/0xb32
[<c013ca3f>] lock_acquire+0x6c/0x89
[<c0442dc4>] mutex_lock_nested+0xc9/0x248
[<c0169211>] lock_super+0x1b/0x1d
[<c01e2f71>] ext4_orphan_del+0x21/0x162
[<c01ecaec>] ext4_ext_truncate+0x114/0x167
[<c01dde2d>] ext4_truncate+0x84/0x40d
[<c01582cb>] vmtruncate+0x103/0x129
[<c01799ac>] inode_setattr+0x67/0x139
[<c01df73c>] ext4_setattr+0x27b/0x2fa
[<c0179bab>] notify_change+0x12d/0x2a6
[<c016719b>] do_truncate+0x62/0x7b
[<c01672d9>] do_sys_ftruncate+0x125/0x142
[<c0167312>] sys_ftruncate64+0x1c/0x24
[<c0103949>] sysenter_past_esp+0x6a/0xb1
[<ffffffff>] 0xffffffff
-> #0 (&ei->i_data_sem){----}:
[<c013c747>] __lock_acquire+0x8a6/0xb32
[<c013ca3f>] lock_acquire+0x6c/0x89
[<c044330a>] down_read+0x2b/0x65
[<c01dd2c1>] ext4_get_blocks_wrap+0x29/0x120
[<c01dd634>] ext4_get_block+0xa2/0xd7
[<c01843f2>] generic_block_bmap+0x43/0x50
[<c01de6de>] ext4_bmap+0x8d/0x96
[<c01783d5>] bmap+0x26/0x2b
[<c02070ee>] jbd2_journal_bmap+0x20/0x75
[<c02072ae>] jbd2_journal_next_log_block+0x61/0x67
[<c0207b44>] jbd2_journal_get_descriptor_buffer+0x13/0x90
[<c0202f73>] journal_submit_commit_record+0x2f/0x19a
[<c0203bb7>] jbd2_journal_commit_transaction+0xabf/0x1038
[<c02069f9>] jbd2_journal_destroy+0xc8/0x1c9
[<c01e5835>] ext4_put_super+0x35/0x1ab
[<c0169a1a>] generic_shutdown_super+0x52/0xd3
[<c0169aaa>] kill_block_super+0xf/0x20
[<c0169b63>] deactivate_super+0x57/0x6a
[<c017b6f3>] mntput_no_expire+0xba/0xdd
[<c017bb95>] sys_umount+0x259/0x29b
[<c0103949>] sysenter_past_esp+0x6a/0xb1
[<ffffffff>] 0xffffffff
other info that might help us debug this:
2 locks held by umount/4921:
#0: (&type->s_umount_key#17){----}, at: [<c0169b5e>] deactivate_super+0x52/0x6a
#1: (&type->s_lock_key#7){--..}, at: [<c0169211>] lock_super+0x1b/0x1d
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
fs/ext4/extents.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 48b8764..7bbc095 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2971,6 +2971,7 @@ void ext4_ext_truncate(struct inode *inode)
handle->h_sync = 1;
out_stop:
+ up_write(&EXT4_I(inode)->i_data_sem);
/*
* If this was a simple ftruncate() and the file will remain alive,
* then we need to clear up the orphan record which we created above.
@@ -2981,7 +2982,6 @@ void ext4_ext_truncate(struct inode *inode)
if (inode->i_nlink)
ext4_orphan_del(handle, inode);
- up_write(&EXT4_I(inode)->i_data_sem);
inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
ext4_mark_inode_dirty(handle, inode);
ext4_journal_stop(handle);
--
1.5.6.2.255.gbed62.dirty
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH] ext4: Fix lock order during truncate.
2008-07-10 8:56 [PATCH] ext4: Fix lock order during truncate Aneesh Kumar K.V
@ 2008-07-11 23:00 ` Mingming Cao
0 siblings, 0 replies; 2+ messages in thread
From: Mingming Cao @ 2008-07-11 23:00 UTC (permalink / raw)
To: Aneesh Kumar K.V; +Cc: tytso, sandeen, jack, linux-ext4
在 2008-07-10四的 14:26 +0530，Aneesh Kumar K.V写道：
> This can be merged to
> ext4-fix-lock-inversion-in-ext4_ext_truncate.patch
>
>
Looks good to me.
Folded this fix to ext4-fix-lock-inversion-in-ext4_ext_truncate.patch
in patch queue
Mingming
>
> =======================================================
> [ INFO: possible circular locking dependency detected ]
> 2.6.26-rc8 #32
> -------------------------------------------------------
> umount/4921 is trying to acquire lock:
> (&ei->i_data_sem){----}, at: [<c01dd2c1>] ext4_get_blocks_wrap+0x29/0x120
>
> but task is already holding lock:
> (&type->s_lock_key#7){--..}, at: [<c0169211>] lock_super+0x1b/0x1d
>
> which lock already depends on the new lock.
>
>
> the existing dependency chain (in reverse order) is:
>
> -> #1 (&type->s_lock_key#7){--..}:
> [<c013c820>] __lock_acquire+0x97f/0xb32
> [<c013ca3f>] lock_acquire+0x6c/0x89
> [<c0442dc4>] mutex_lock_nested+0xc9/0x248
> [<c0169211>] lock_super+0x1b/0x1d
> [<c01e2f71>] ext4_orphan_del+0x21/0x162
> [<c01ecaec>] ext4_ext_truncate+0x114/0x167
> [<c01dde2d>] ext4_truncate+0x84/0x40d
> [<c01582cb>] vmtruncate+0x103/0x129
> [<c01799ac>] inode_setattr+0x67/0x139
> [<c01df73c>] ext4_setattr+0x27b/0x2fa
> [<c0179bab>] notify_change+0x12d/0x2a6
> [<c016719b>] do_truncate+0x62/0x7b
> [<c01672d9>] do_sys_ftruncate+0x125/0x142
> [<c0167312>] sys_ftruncate64+0x1c/0x24
> [<c0103949>] sysenter_past_esp+0x6a/0xb1
> [<ffffffff>] 0xffffffff
>
> -> #0 (&ei->i_data_sem){----}:
> [<c013c747>] __lock_acquire+0x8a6/0xb32
> [<c013ca3f>] lock_acquire+0x6c/0x89
> [<c044330a>] down_read+0x2b/0x65
> [<c01dd2c1>] ext4_get_blocks_wrap+0x29/0x120
> [<c01dd634>] ext4_get_block+0xa2/0xd7
> [<c01843f2>] generic_block_bmap+0x43/0x50
> [<c01de6de>] ext4_bmap+0x8d/0x96
> [<c01783d5>] bmap+0x26/0x2b
> [<c02070ee>] jbd2_journal_bmap+0x20/0x75
> [<c02072ae>] jbd2_journal_next_log_block+0x61/0x67
> [<c0207b44>] jbd2_journal_get_descriptor_buffer+0x13/0x90
> [<c0202f73>] journal_submit_commit_record+0x2f/0x19a
> [<c0203bb7>] jbd2_journal_commit_transaction+0xabf/0x1038
> [<c02069f9>] jbd2_journal_destroy+0xc8/0x1c9
> [<c01e5835>] ext4_put_super+0x35/0x1ab
> [<c0169a1a>] generic_shutdown_super+0x52/0xd3
> [<c0169aaa>] kill_block_super+0xf/0x20
> [<c0169b63>] deactivate_super+0x57/0x6a
> [<c017b6f3>] mntput_no_expire+0xba/0xdd
> [<c017bb95>] sys_umount+0x259/0x29b
> [<c0103949>] sysenter_past_esp+0x6a/0xb1
> [<ffffffff>] 0xffffffff
>
> other info that might help us debug this:
>
> 2 locks held by umount/4921:
> #0: (&type->s_umount_key#17){----}, at: [<c0169b5e>] deactivate_super+0x52/0x6a
> #1: (&type->s_lock_key#7){--..}, at: [<c0169211>] lock_super+0x1b/0x1d
>
> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
> ---
> fs/ext4/extents.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
> index 48b8764..7bbc095 100644
> --- a/fs/ext4/extents.c
> +++ b/fs/ext4/extents.c
> @@ -2971,6 +2971,7 @@ void ext4_ext_truncate(struct inode *inode)
> handle->h_sync = 1;
>
> out_stop:
> + up_write(&EXT4_I(inode)->i_data_sem);
> /*
> * If this was a simple ftruncate() and the file will remain alive,
> * then we need to clear up the orphan record which we created above.
> @@ -2981,7 +2982,6 @@ void ext4_ext_truncate(struct inode *inode)
> if (inode->i_nlink)
> ext4_orphan_del(handle, inode);
>
> - up_write(&EXT4_I(inode)->i_data_sem);
> inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
> ext4_mark_inode_dirty(handle, inode);
> ext4_journal_stop(handle);
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-07-11 23:00 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-07-10 8:56 [PATCH] ext4: Fix lock order during truncate Aneesh Kumar K.V
2008-07-11 23:00 ` Mingming Cao
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox