Skip to contentSkip to conversation

2.6.21 reiserfs -- cicular locking?

DISCUSSION David Brownell 6 participants 10 messages Apr 26, 2007 – Apr 27, 2007

lore
Share & tools
lore
#0 lore raw

This might be a Heisenberg, but I figure it's worth posting in case anyone else sees similar oddness. Never seen it before or since. It's as if a gremlin got annoyed with me for switching a filesystem from reiser to ext3. :)

- Dave

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-git #6
-------------------------------------------------------
vi/4556 is trying to acquire lock:
(&REISERFS_SB(s)->xattr_dir_sem){..--}, at: [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<ffffffff8026c4fa>] chown_common+0x93/0xb3

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #1 (&inode->i_mutex){--..}:

       [<ffffffff802405f4>] __lock_acquire+0x9f7/0xbaa
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff80240822>] lock_acquire+0x7b/0x9f
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff8023dee0>] save_trace+0x40/0x9e
       [<ffffffff8041e1db>] __mutex_lock_slowpath+0xd8/0x281
       [<ffffffff8041ff18>] _spin_unlock_irq+0x24/0x4a
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff802ca2a8>] open_xa_dir+0x1c/0xf8
       [<ffffffff8041f34b>] __down_read+0x34/0x9d
       [<ffffffff802cb50a>] reiserfs_delete_xattrs+0x64/0x185
       [<ffffffff802fd620>] _atomic_dec_and_lock+0x14/0x34
       [<ffffffff802b252e>] reiserfs_delete_inode+0x38/0xae
       [<ffffffff8027f6bc>] generic_delete_inode+0x64/0xf5
       [<ffffffff802b24f6>] reiserfs_delete_inode+0x0/0xae
       [<ffffffff8027f6d2>] generic_delete_inode+0x7a/0xf5
       [<ffffffff80276b95>] do_unlinkat+0xd9/0x14f
       [<ffffffff8023f75f>] trace_hardirqs_on+0x123/0x14d
       [<ffffffff8041f4d1>] trace_hardirqs_on_thunk+0x35/0x37
       [<ffffffff8020954e>] system_call+0x7e/0x83
       [<ffffffffffffffff>] 0xffffffffffffffff

-> #0 (&REISERFS_SB(s)->xattr_dir_sem){..--}:

       [<ffffffff8023ee6f>] print_circular_bug_header+0xcc/0xd3
       [<ffffffff802404f0>] __lock_acquire+0x8f3/0xbaa
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff80240822>] lock_acquire+0x7b/0x9f
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff8023bb03>] down_read+0x32/0x3b
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff8022e285>] __capable+0x9/0x1d
       [<ffffffff802b2828>] reiserfs_setattr+0x11e/0x1ec
       [<ffffffff8022bd6b>] current_fs_time+0x35/0x3a
       [<ffffffff802803a1>] notify_change+0x122/0x231
       [<ffffffff8026c505>] chown_common+0x9e/0xb3
       [<ffffffff8026e7a6>] fget+0x88/0xa7
       [<ffffffff8026c54a>] sys_fchown+0x30/0x47
       [<ffffffff8020954e>] system_call+0x7e/0x83
       [<ffffffffffffffff>] 0xffffffffffffffff

other info that might help us debug this:

1 lock held by vi/4556:
#0: (&inode->i_mutex){--..}, at: [<ffffffff8026c4fa>] chown_common+0x93/0xb3

stack backtrace:

Call Trace:
 [<ffffffff8023eacc>] print_circular_bug_tail+0x69/0x72
 [<ffffffff8023ee6f>] print_circular_bug_header+0xcc/0xd3
 [<ffffffff802404f0>] __lock_acquire+0x8f3/0xbaa
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff80240822>] lock_acquire+0x7b/0x9f
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff8023bb03>] down_read+0x32/0x3b
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff8022e285>] __capable+0x9/0x1d
 [<ffffffff802b2828>] reiserfs_setattr+0x11e/0x1ec
 [<ffffffff8022bd6b>] current_fs_time+0x35/0x3a
 [<ffffffff802803a1>] notify_change+0x122/0x231
 [<ffffffff8026c505>] chown_common+0x9e/0xb3
 [<ffffffff8026e7a6>] fget+0x88/0xa7
 [<ffffffff8026c54a>] sys_fchown+0x30/0x47
 [<ffffffff8020954e>] system_call+0x7e/0x83
#1 lore raw
Quoted 80 lines · On Thu, 26 Apr 2007 16:40:14 -0700 David Brownell <[email protected]> wrote:

This might be a Heisenberg, but I figure it's worth posting in case anyone else sees similar oddness. Never seen it before or since. It's as if a gremlin got annoyed with me for switching a filesystem from reiser to ext3. :)

- Dave

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-git #6
-------------------------------------------------------
vi/4556 is trying to acquire lock:
(&REISERFS_SB(s)->xattr_dir_sem){..--}, at: [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<ffffffff8026c4fa>] chown_common+0x93/0xb3

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #1 (&inode->i_mutex){--..}:

       [<ffffffff802405f4>] __lock_acquire+0x9f7/0xbaa
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff80240822>] lock_acquire+0x7b/0x9f
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff8023dee0>] save_trace+0x40/0x9e
       [<ffffffff8041e1db>] __mutex_lock_slowpath+0xd8/0x281
       [<ffffffff8041ff18>] _spin_unlock_irq+0x24/0x4a
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff802ca2a8>] open_xa_dir+0x1c/0xf8
       [<ffffffff8041f34b>] __down_read+0x34/0x9d
       [<ffffffff802cb50a>] reiserfs_delete_xattrs+0x64/0x185
       [<ffffffff802fd620>] _atomic_dec_and_lock+0x14/0x34
       [<ffffffff802b252e>] reiserfs_delete_inode+0x38/0xae
       [<ffffffff8027f6bc>] generic_delete_inode+0x64/0xf5
       [<ffffffff802b24f6>] reiserfs_delete_inode+0x0/0xae
       [<ffffffff8027f6d2>] generic_delete_inode+0x7a/0xf5
       [<ffffffff80276b95>] do_unlinkat+0xd9/0x14f
       [<ffffffff8023f75f>] trace_hardirqs_on+0x123/0x14d
       [<ffffffff8041f4d1>] trace_hardirqs_on_thunk+0x35/0x37
       [<ffffffff8020954e>] system_call+0x7e/0x83
       [<ffffffffffffffff>] 0xffffffffffffffff

-> #0 (&REISERFS_SB(s)->xattr_dir_sem){..--}:

       [<ffffffff8023ee6f>] print_circular_bug_header+0xcc/0xd3
       [<ffffffff802404f0>] __lock_acquire+0x8f3/0xbaa
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff80240822>] lock_acquire+0x7b/0x9f
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff8023bb03>] down_read+0x32/0x3b
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff8022e285>] __capable+0x9/0x1d
       [<ffffffff802b2828>] reiserfs_setattr+0x11e/0x1ec
       [<ffffffff8022bd6b>] current_fs_time+0x35/0x3a
       [<ffffffff802803a1>] notify_change+0x122/0x231
       [<ffffffff8026c505>] chown_common+0x9e/0xb3
       [<ffffffff8026e7a6>] fget+0x88/0xa7
       [<ffffffff8026c54a>] sys_fchown+0x30/0x47
       [<ffffffff8020954e>] system_call+0x7e/0x83
       [<ffffffffffffffff>] 0xffffffffffffffff

other info that might help us debug this:

1 lock held by vi/4556:
#0: (&inode->i_mutex){--..}, at: [<ffffffff8026c4fa>] chown_common+0x93/0xb3

stack backtrace:

Call Trace:
 [<ffffffff8023eacc>] print_circular_bug_tail+0x69/0x72
 [<ffffffff8023ee6f>] print_circular_bug_header+0xcc/0xd3
 [<ffffffff802404f0>] __lock_acquire+0x8f3/0xbaa
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff80240822>] lock_acquire+0x7b/0x9f
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff8023bb03>] down_read+0x32/0x3b
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
Quoted 9 lines
 [<ffffffff8022e285>] __capable+0x9/0x1d
 [<ffffffff802b2828>] reiserfs_setattr+0x11e/0x1ec
 [<ffffffff8022bd6b>] current_fs_time+0x35/0x3a
 [<ffffffff802803a1>] notify_change+0x122/0x231
 [<ffffffff8026c505>] chown_common+0x9e/0xb3
 [<ffffffff8026e7a6>] fget+0x88/0xa7
 [<ffffffff8026c54a>] sys_fchown+0x30/0x47
 [<ffffffff8020954e>] system_call+0x7e/0x83

cc added. This was also reported againt -rc7-mm1 (or 2)

#2 lore raw

At Thu, 26 Apr 2007 22:44:08 -0700,

Quoted 93 lines · Andrew Morton
Quoted 80 lines · On Thu, 26 Apr 2007 16:40:14 -0700 David Brownell <[email protected]> wrote:

This might be a Heisenberg, but I figure it's worth posting in case anyone else sees similar oddness. Never seen it before or since. It's as if a gremlin got annoyed with me for switching a filesystem from reiser to ext3. :)

- Dave

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-git #6
-------------------------------------------------------
vi/4556 is trying to acquire lock:
(&REISERFS_SB(s)->xattr_dir_sem){..--}, at: [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<ffffffff8026c4fa>] chown_common+0x93/0xb3

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #1 (&inode->i_mutex){--..}:

       [<ffffffff802405f4>] __lock_acquire+0x9f7/0xbaa
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff80240822>] lock_acquire+0x7b/0x9f
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff8023dee0>] save_trace+0x40/0x9e
       [<ffffffff8041e1db>] __mutex_lock_slowpath+0xd8/0x281
       [<ffffffff8041ff18>] _spin_unlock_irq+0x24/0x4a
       [<ffffffff802ca1ac>] get_xa_root+0x49/0x107
       [<ffffffff802ca2a8>] open_xa_dir+0x1c/0xf8
       [<ffffffff8041f34b>] __down_read+0x34/0x9d
       [<ffffffff802cb50a>] reiserfs_delete_xattrs+0x64/0x185
       [<ffffffff802fd620>] _atomic_dec_and_lock+0x14/0x34
       [<ffffffff802b252e>] reiserfs_delete_inode+0x38/0xae
       [<ffffffff8027f6bc>] generic_delete_inode+0x64/0xf5
       [<ffffffff802b24f6>] reiserfs_delete_inode+0x0/0xae
       [<ffffffff8027f6d2>] generic_delete_inode+0x7a/0xf5
       [<ffffffff80276b95>] do_unlinkat+0xd9/0x14f
       [<ffffffff8023f75f>] trace_hardirqs_on+0x123/0x14d
       [<ffffffff8041f4d1>] trace_hardirqs_on_thunk+0x35/0x37
       [<ffffffff8020954e>] system_call+0x7e/0x83
       [<ffffffffffffffff>] 0xffffffffffffffff

-> #0 (&REISERFS_SB(s)->xattr_dir_sem){..--}:

       [<ffffffff8023ee6f>] print_circular_bug_header+0xcc/0xd3
       [<ffffffff802404f0>] __lock_acquire+0x8f3/0xbaa
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff80240822>] lock_acquire+0x7b/0x9f
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff8023bb03>] down_read+0x32/0x3b
       [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
       [<ffffffff8022e285>] __capable+0x9/0x1d
       [<ffffffff802b2828>] reiserfs_setattr+0x11e/0x1ec
       [<ffffffff8022bd6b>] current_fs_time+0x35/0x3a
       [<ffffffff802803a1>] notify_change+0x122/0x231
       [<ffffffff8026c505>] chown_common+0x9e/0xb3
       [<ffffffff8026e7a6>] fget+0x88/0xa7
       [<ffffffff8026c54a>] sys_fchown+0x30/0x47
       [<ffffffff8020954e>] system_call+0x7e/0x83
       [<ffffffffffffffff>] 0xffffffffffffffff

other info that might help us debug this:

1 lock held by vi/4556:
#0: (&inode->i_mutex){--..}, at: [<ffffffff8026c4fa>] chown_common+0x93/0xb3

stack backtrace:

Call Trace:
 [<ffffffff8023eacc>] print_circular_bug_tail+0x69/0x72
 [<ffffffff8023ee6f>] print_circular_bug_header+0xcc/0xd3
 [<ffffffff802404f0>] __lock_acquire+0x8f3/0xbaa
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff80240822>] lock_acquire+0x7b/0x9f
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
 [<ffffffff8023bb03>] down_read+0x32/0x3b
 [<ffffffff802caf4b>] reiserfs_chown_xattrs+0x5b/0x128
Quoted 9 lines
 [<ffffffff8022e285>] __capable+0x9/0x1d
 [<ffffffff802b2828>] reiserfs_setattr+0x11e/0x1ec
 [<ffffffff8022bd6b>] current_fs_time+0x35/0x3a
 [<ffffffff802803a1>] notify_change+0x122/0x231
 [<ffffffff8026c505>] chown_common+0x9e/0xb3
 [<ffffffff8026e7a6>] fget+0x88/0xa7
 [<ffffffff8026c54a>] sys_fchown+0x30/0x47
 [<ffffffff8020954e>] system_call+0x7e/0x83
Quoted 1 line

cc added. This was also reported againt -rc7-mm1 (or 2)

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 (&inode->i_mutex){--..}:

       [<c0134d2e>] __lock_acquire+0xa27/0xbbb
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<c0134f29>] lock_acquire+0x67/0x81
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<c028b69d>] __mutex_lock_slowpath+0xe3/0x241
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e08a2f83>] open_xa_dir+0x16/0xd9 [reiserfs]
       [<e08a3feb>] reiserfs_delete_xattrs+0x4b/0x15b [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<c012e0d6>] down_read+0x3d/0x4e
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e08a3ff7>] reiserfs_delete_xattrs+0x57/0x15b [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e088b171>] reiserfs_delete_inode+0x35/0xa1 [reiserfs]
       [<c01c0fce>] _atomic_dec_and_lock+0x2a/0x48
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<c017507f>] generic_delete_inode+0x75/0xdd
       [<c01747c4>] iput+0x60/0x62
       [<e0892878>] finish_unfinished+0x2ee/0x350 [reiserfs]
       [<c016bc00>] lookup_one_len+0x21/0x59
       [<e08a3d45>] reiserfs_xattr_init+0x8f/0x1f6 [reiserfs]
       [<e0893cd4>] reiserfs_fill_super+0x95e/0xab6 [reiserfs]
       [<c0129520>] rcu_barrier+0x5a/0x6a
       [<c0105cd0>] dump_trace+0x89/0x93
       [<c010a075>] save_stack_trace+0x1c/0x37
       [<c01327a5>] save_trace+0x40/0x92
       [<c0165c02>] sget+0x1f/0x33b
       [<c0134e2e>] __lock_acquire+0xb27/0xbbb
       [<c0165c02>] sget+0x1f/0x33b
       [<c01c483e>] vsnprintf+0x450/0x48c
       [<c01c494c>] snprintf+0x1f/0x22
       [<c0194687>] disk_name+0x7e/0x88
       [<c016677f>] get_sb_bdev+0xe6/0x130
       [<e0891ccc>] get_super_block+0x20/0x25 [reiserfs]
       [<e0893376>] reiserfs_fill_super+0x0/0xab6 [reiserfs]
       [<c016633b>] vfs_kern_mount+0x83/0xf6
       [<c01663f0>] do_kern_mount+0x2d/0x3e
       [<c0177ee4>] do_mount+0x612/0x685
       [<c01543e9>] __handle_mm_fault+0x50c/0x902
       [<c01543b2>] __handle_mm_fault+0x4d5/0x902
       [<c028c912>] _spin_unlock+0x14/0x1c
       [<c01547a8>] __handle_mm_fault+0x8cb/0x902
       [<c014d104>] get_page_from_freelist+0x1fd/0x31d
       [<c0133eeb>] trace_hardirqs_on+0x126/0x150
       [<c012e14d>] up_read+0x14/0x27
       [<c014d28c>] __alloc_pages+0x68/0x2aa
       [<c0176899>] copy_mount_options+0x26/0x109
       [<c0177fce>] sys_mount+0x77/0xae
       [<c0104d74>] syscall_call+0x7/0xb
       [<ffffffff>] 0xffffffff

-> #1 (&REISERFS_SB(s)->xattr_dir_sem){..--}:

       [<c0134d2e>] __lock_acquire+0xa27/0xbbb
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<c0134f29>] lock_acquire+0x67/0x81
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<c012e0d6>] down_read+0x3d/0x4e
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<c0133eeb>] trace_hardirqs_on+0x126/0x150
       [<e08a3b98>] reiserfs_listxattr+0x0/0x11e [reiserfs]
       [<c017ab4f>] vfs_listxattr+0x46/0x7c
       [<c017abc9>] listxattr+0x44/0x87
       [<c017ac73>] sys_llistxattr+0x33/0x44
       [<c012e14d>] up_read+0x14/0x27
       [<c0104dbc>] restore_nocheck+0x12/0x15
       [<c0104d74>] syscall_call+0x7/0xb
       [<ffffffff>] 0xffffffff

-> #0 (&REISERFS_I(inode)->xattr_sem){..--}:

       [<c012e6d2>] print_stack_trace+0x4e/0x5c
       [<c0134c1a>] __lock_acquire+0x913/0xbbb
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<c017570b>] new_inode+0x24/0x8a
       [<c0134f29>] lock_acquire+0x67/0x81
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<c012e0d6>] down_read+0x3d/0x4e
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<e08873b5>] reiserfs_create+0x3d/0x1bc [reiserfs]
       [<e08a2e37>] reiserfs_permission+0x0/0x21 [reiserfs]
       [<c016abf2>] permission+0xc8/0xdb
       [<c016b11f>] vfs_create+0x9c/0x106
       [<c016d871>] open_namei+0x177/0x5a2
       [<c016348d>] do_filp_open+0x25/0x39
       [<c028c912>] _spin_unlock+0x14/0x1c
       [<c016325b>] get_unused_fd+0xb3/0xbd
       [<c01634e3>] do_sys_open+0x42/0xbe
       [<c0163598>] sys_open+0x1c/0x1e
       [<c0104d74>] syscall_call+0x7/0xb
       [<ffffffff>] 0xffffffff

other info that might help us debug this:

1 lock held by mktemp/1459:
#0: (&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

stack backtrace:
 [<c013343e>] print_circular_bug_tail+0x5f/0x67
 [<c0134c1a>] __lock_acquire+0x913/0xbbb
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<c017570b>] new_inode+0x24/0x8a
 [<c0134f29>] lock_acquire+0x67/0x81
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<c012e0d6>] down_read+0x3d/0x4e
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<e08873b5>] reiserfs_create+0x3d/0x1bc [reiserfs]
 [<e08a2e37>] reiserfs_permission+0x0/0x21 [reiserfs]
 [<c016abf2>] permission+0xc8/0xdb
 [<c016b11f>] vfs_create+0x9c/0x106
 [<c016d871>] open_namei+0x177/0x5a2
 [<c016348d>] do_filp_open+0x25/0x39
 [<c028c912>] _spin_unlock+0x14/0x1c
 [<c016325b>] get_unused_fd+0xb3/0xbd
 [<c01634e3>] do_sys_open+0x42/0xbe
 [<c0163598>] sys_open+0x1c/0x1e
 [<c0104d74>] syscall_call+0x7/0xb
 =======================

Takashi

#3 lore raw

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 142 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 (&inode->i_mutex){--..}:

       [<c0134d2e>] __lock_acquire+0xa27/0xbbb
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<c0134f29>] lock_acquire+0x67/0x81
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<c028b69d>] __mutex_lock_slowpath+0xe3/0x241
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<e08a2e9a>] get_xa_root+0x42/0xfc [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e08a2f83>] open_xa_dir+0x16/0xd9 [reiserfs]
       [<e08a3feb>] reiserfs_delete_xattrs+0x4b/0x15b [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<c012e0d6>] down_read+0x3d/0x4e
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e08a3ff7>] reiserfs_delete_xattrs+0x57/0x15b [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<e088b171>] reiserfs_delete_inode+0x35/0xa1 [reiserfs]
       [<c01c0fce>] _atomic_dec_and_lock+0x2a/0x48
       [<e088b13c>] reiserfs_delete_inode+0x0/0xa1 [reiserfs]
       [<c017507f>] generic_delete_inode+0x75/0xdd
       [<c01747c4>] iput+0x60/0x62
       [<e0892878>] finish_unfinished+0x2ee/0x350 [reiserfs]
       [<c016bc00>] lookup_one_len+0x21/0x59
       [<e08a3d45>] reiserfs_xattr_init+0x8f/0x1f6 [reiserfs]
       [<e0893cd4>] reiserfs_fill_super+0x95e/0xab6 [reiserfs]
       [<c0129520>] rcu_barrier+0x5a/0x6a
       [<c0105cd0>] dump_trace+0x89/0x93
       [<c010a075>] save_stack_trace+0x1c/0x37
       [<c01327a5>] save_trace+0x40/0x92
       [<c0165c02>] sget+0x1f/0x33b
       [<c0134e2e>] __lock_acquire+0xb27/0xbbb
       [<c0165c02>] sget+0x1f/0x33b
       [<c01c483e>] vsnprintf+0x450/0x48c
       [<c01c494c>] snprintf+0x1f/0x22
       [<c0194687>] disk_name+0x7e/0x88
       [<c016677f>] get_sb_bdev+0xe6/0x130
       [<e0891ccc>] get_super_block+0x20/0x25 [reiserfs]
       [<e0893376>] reiserfs_fill_super+0x0/0xab6 [reiserfs]
       [<c016633b>] vfs_kern_mount+0x83/0xf6
       [<c01663f0>] do_kern_mount+0x2d/0x3e
       [<c0177ee4>] do_mount+0x612/0x685
       [<c01543e9>] __handle_mm_fault+0x50c/0x902
       [<c01543b2>] __handle_mm_fault+0x4d5/0x902
       [<c028c912>] _spin_unlock+0x14/0x1c
       [<c01547a8>] __handle_mm_fault+0x8cb/0x902
       [<c014d104>] get_page_from_freelist+0x1fd/0x31d
       [<c0133eeb>] trace_hardirqs_on+0x126/0x150
       [<c012e14d>] up_read+0x14/0x27
       [<c014d28c>] __alloc_pages+0x68/0x2aa
       [<c0176899>] copy_mount_options+0x26/0x109
       [<c0177fce>] sys_mount+0x77/0xae
       [<c0104d74>] syscall_call+0x7/0xb
       [<ffffffff>] 0xffffffff

-> #1 (&REISERFS_SB(s)->xattr_dir_sem){..--}:

       [<c0134d2e>] __lock_acquire+0xa27/0xbbb
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<c0134f29>] lock_acquire+0x67/0x81
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<c012e0d6>] down_read+0x3d/0x4e
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<e08a3bf2>] reiserfs_listxattr+0x5a/0x11e [reiserfs]
       [<c0133eeb>] trace_hardirqs_on+0x126/0x150
       [<e08a3b98>] reiserfs_listxattr+0x0/0x11e [reiserfs]
       [<c017ab4f>] vfs_listxattr+0x46/0x7c
       [<c017abc9>] listxattr+0x44/0x87
       [<c017ac73>] sys_llistxattr+0x33/0x44
       [<c012e14d>] up_read+0x14/0x27
       [<c0104dbc>] restore_nocheck+0x12/0x15
       [<c0104d74>] syscall_call+0x7/0xb
       [<ffffffff>] 0xffffffff

-> #0 (&REISERFS_I(inode)->xattr_sem){..--}:

       [<c012e6d2>] print_stack_trace+0x4e/0x5c
       [<c0134c1a>] __lock_acquire+0x913/0xbbb
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<c017570b>] new_inode+0x24/0x8a
       [<c0134f29>] lock_acquire+0x67/0x81
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<c012e0d6>] down_read+0x3d/0x4e
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
       [<e08873b5>] reiserfs_create+0x3d/0x1bc [reiserfs]
       [<e08a2e37>] reiserfs_permission+0x0/0x21 [reiserfs]
       [<c016abf2>] permission+0xc8/0xdb
       [<c016b11f>] vfs_create+0x9c/0x106
       [<c016d871>] open_namei+0x177/0x5a2
       [<c016348d>] do_filp_open+0x25/0x39
       [<c028c912>] _spin_unlock+0x14/0x1c
       [<c016325b>] get_unused_fd+0xb3/0xbd
       [<c01634e3>] do_sys_open+0x42/0xbe
       [<c0163598>] sys_open+0x1c/0x1e
       [<c0104d74>] syscall_call+0x7/0xb
       [<ffffffff>] 0xffffffff

other info that might help us debug this:

1 lock held by mktemp/1459:
#0: (&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

stack backtrace:
 [<c013343e>] print_circular_bug_tail+0x5f/0x67
 [<c0134c1a>] __lock_acquire+0x913/0xbbb
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<c017570b>] new_inode+0x24/0x8a
 [<c0134f29>] lock_acquire+0x67/0x81
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<c012e0d6>] down_read+0x3d/0x4e
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]
 [<e08873b5>] reiserfs_create+0x3d/0x1bc [reiserfs]
 [<e08a2e37>] reiserfs_permission+0x0/0x21 [reiserfs]
 [<c016abf2>] permission+0xc8/0xdb
 [<c016b11f>] vfs_create+0x9c/0x106
Quoted 8 lines
 [<c016d871>] open_namei+0x177/0x5a2
 [<c016348d>] do_filp_open+0x25/0x39
 [<c028c912>] _spin_unlock+0x14/0x1c
 [<c016325b>] get_unused_fd+0xb3/0xbd
 [<c01634e3>] do_sys_open+0x42/0xbe
 [<c0163598>] sys_open+0x1c/0x1e
 [<c0104d74>] syscall_call+0x7/0xb
 =======================

The message disappears when I revert the patch:

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700

reiserfs: fix xattr root locking/refcount bug

So, likely a newly introduced bug after rc7...

Takashi

#4 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 26 lines · Takashi Iwai

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 18 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

Quoted 4 lines

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

The message disappears when I revert the patch:

Quoted 8 lines

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700

reiserfs: fix xattr root locking/refcount bug

So, likely a newly introduced bug after rc7...

I got a message with a trace similar to this from Vladimir before I submitted that patch. I'm not sure how to annotate this, since the xattr_sem can never be taken in the manner described. Internal inodes are protected by I_PRIVATE.

- -Jeff

- --
Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org

iD8DBQFGMdnNLPWxlyuTD7IRApM+AJwKynnSbQfGKByzvDFs5d0OqO82ggCfVzoP MNzGHMUQdmn2Xg31QWhB3mQ=
=A37l
-----END PGP SIGNATURE-----

#5 lore raw

At Fri, 27 Apr 2007 07:09:01 -0400,

Quoted 38 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 33 lines · Takashi Iwai

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 18 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

Quoted 4 lines

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

The message disappears when I revert the patch:

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700

reiserfs: fix xattr root locking/refcount bug

Quoted 6 lines
Quoted 1 line

So, likely a newly introduced bug after rc7...

I got a message with a trace similar to this from Vladimir before I submitted that patch. I'm not sure how to annotate this, since the xattr_sem can never be taken in the manner described. Internal inodes are protected by I_PRIVATE.

Hm, then maybe my case was just a coincidence.

FWIW, I can reproduce the deadlock warning at each time I boot non-patched 2.6.21, and after reverting the patch, it disappeared.

Takashi

#6 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 45 lines · Takashi Iwai

At Fri, 27 Apr 2007 07:09:01 -0400,

Quoted 34 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 30 lines · Takashi Iwai

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 18 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

Quoted 4 lines

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

The message disappears when I revert the patch:

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700
Quoted 8 lines
Quoted 4 lines

reiserfs: fix xattr root locking/refcount bug

So, likely a newly introduced bug after rc7...

I got a message with a trace similar to this from Vladimir before I submitted that patch. I'm not sure how to annotate this, since the xattr_sem can never be taken in the manner described. Internal inodes are protected by I_PRIVATE.

Quoted 4 lines

Hm, then maybe my case was just a coincidence.

FWIW, I can reproduce the deadlock warning at each time I boot non-patched 2.6.21, and after reverting the patch, it disappeared.

Ok, so I took another look at the report Vladimir sent me. The trace he ran into was in the delete inode path, but was still a race between the xattr_sem and the inode sem. Since we're locking the xattr root on the xattr read path now, this condition arises more freqently, but it's really the same one he reported.

I'm using the default openSUSE config, which doesn't enable mutex debugging. I'll rebuild with it, and hopefully come up with a way to kill the warning.

- -Jeff

- --
Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org

iD8DBQFGMhP8LPWxlyuTD7IRAiInAJwO7Tpga98WvcPICFMSyvb0vSAvrQCdEbal XScT+lxIu181A6KNzKhNuYw=
=K/gK
-----END PGP SIGNATURE-----

#7 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 56 lines · Jeff Mahoney
Quoted 40 lines · Takashi Iwai
Quoted 40 lines

At Fri, 27 Apr 2007 07:09:01 -0400,

Quoted 38 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 34 lines · Takashi Iwai

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 18 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

Quoted 4 lines

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

The message disappears when I revert the patch:

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700

reiserfs: fix xattr root locking/refcount bug

So, likely a newly introduced bug after rc7...

Quoted 8 lines
Quoted 8 lines
Quoted 4 lines

I got a message with a trace similar to this from Vladimir before I submitted that patch. I'm not sure how to annotate this, since the xattr_sem can never be taken in the manner described. Internal inodes are protected by I_PRIVATE.

Hm, then maybe my case was just a coincidence.

FWIW, I can reproduce the deadlock warning at each time I boot non-patched 2.6.21, and after reverting the patch, it disappeared.

Ok, so I took another look at the report Vladimir sent me. The trace he ran into was in the delete inode path, but was still a race between the xattr_sem and the inode sem. Since we're locking the xattr root on the xattr read path now, this condition arises more freqently, but it's really the same one he reported.

Quoted 3 lines

I'm using the default openSUSE config, which doesn't enable mutex debugging. I'll rebuild with it, and hopefully come up with a way to kill the warning.

I still didn't get the warning, but can you try this and let me know if it fixes it?

- -Jeff

- --- a/fs/reiserfs/xattr.c 2007-04-22 10:53:10.000000000 -0400
+++ b/fs/reiserfs/xattr.c 2007-04-27 12:09:02.000000000 -0400
@@ -68,7 +68,7 @@
if (!privroot)
return ERR_PTR(-ENODATA);

- - mutex_lock(&privroot->d_inode->i_mutex);
+ mutex_lock_nested(&privroot->d_inode->i_mutex, I_MUTEX_XATTR);
if (REISERFS_SB(sb)->xattr_root) {
xaroot = dget(REISERFS_SB(sb)->xattr_root); goto out;

- --
Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org

iD8DBQFGMiL1LPWxlyuTD7IRAuwoAKCob0DLVBcmeOApOr9vA5cVFRwBewCfTU/U jwIjK+Uy4RfXKsag5Uamzew=
=svba
-----END PGP SIGNATURE-----

#8 lore raw

At Fri, 27 Apr 2007 12:21:09 -0400,

Quoted 61 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 56 lines · Jeff Mahoney
Quoted 40 lines · Takashi Iwai
Quoted 40 lines

At Fri, 27 Apr 2007 07:09:01 -0400,

Quoted 38 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 34 lines · Takashi Iwai

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 18 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

Quoted 4 lines

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

The message disappears when I revert the patch:

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700

reiserfs: fix xattr root locking/refcount bug

So, likely a newly introduced bug after rc7...

Quoted 8 lines
Quoted 8 lines
Quoted 4 lines

I got a message with a trace similar to this from Vladimir before I submitted that patch. I'm not sure how to annotate this, since the xattr_sem can never be taken in the manner described. Internal inodes are protected by I_PRIVATE.

Hm, then maybe my case was just a coincidence.

FWIW, I can reproduce the deadlock warning at each time I boot non-patched 2.6.21, and after reverting the patch, it disappeared.

Ok, so I took another look at the report Vladimir sent me. The trace he ran into was in the delete inode path, but was still a race between the xattr_sem and the inode sem. Since we're locking the xattr root on the xattr read path now, this condition arises more freqently, but it's really the same one he reported.

Quoted 6 lines
Quoted 3 lines

I'm using the default openSUSE config, which doesn't enable mutex debugging. I'll rebuild with it, and hopefully come up with a way to kill the warning.

I still didn't get the warning, but can you try this and let me know if it fixes it?

Yes, it seems to fix the problem.

Thanks,

Takashi

Quoted 27 lines

- -Jeff

- --- a/fs/reiserfs/xattr.c 2007-04-22 10:53:10.000000000 -0400
+++ b/fs/reiserfs/xattr.c 2007-04-27 12:09:02.000000000 -0400

(quoted diff) +1 -2
@@ -68,7 +68,7 @@
 	if (!privroot)
 		return ERR_PTR(-ENODATA);
 
- -	mutex_lock(&privroot->d_inode->i_mutex);
+	mutex_lock_nested(&privroot->d_inode->i_mutex, I_MUTEX_XATTR);
 	if (REISERFS_SB(sb)->xattr_root) {
 		xaroot = dget(REISERFS_SB(sb)->xattr_root);
 		goto out;
 
- -- 

Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org

iD8DBQFGMiL1LPWxlyuTD7IRAuwoAKCob0DLVBcmeOApOr9vA5cVFRwBewCfTU/U jwIjK+Uy4RfXKsag5Uamzew=
=svba
-----END PGP SIGNATURE-----

#9 lore raw
In reply to Jeff Mahoney
Quoted 60 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 56 lines · Jeff Mahoney
Quoted 40 lines · Takashi Iwai
Quoted 40 lines

At Fri, 27 Apr 2007 07:09:01 -0400,

Quoted 38 lines · Jeff Mahoney

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 34 lines · Takashi Iwai

At Fri, 27 Apr 2007 12:09:03 +0200,

Quoted 18 lines · I wrote:

I got a similar bug right now at the fresh boot of 2.6.21.

ReiserFS: sda2: found reiserfs format "3.6" with standard journal
ReiserFS: sda2: using ordered data mode
ReiserFS: sda2: journal params: device sda2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: sda2: checking transaction log (sda2)
ReiserFS: sda2: Using r5 hash to sort names
ReiserFS: sda2: Removing [3613 1354701 0x0 SD]..done
ReiserFS: sda2: There were 1 uncompleted unlinks/truncates. Completed

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.21-work #1
-------------------------------------------------------
mktemp/1459 is trying to acquire lock:
(&REISERFS_I(inode)->xattr_sem){..--}, at: [<e08a5236>] reiserfs_cache_default_acl+0x2a/0x9c [reiserfs]

Quoted 4 lines

but task is already holding lock:
(&inode->i_mutex){--..}, at: [<c016d7dc>] open_namei+0xe2/0x5a2

which lock already depends on the new lock.

The message disappears when I revert the patch:

commit 9b7f375505f5611efb562065b57814b28a81abc3

Author: Jeff Mahoney <[email protected]>
Date:   Mon Apr 23 14:41:17 2007 -0700

reiserfs: fix xattr root locking/refcount bug

So, likely a newly introduced bug after rc7...

Quoted 8 lines
Quoted 8 lines
Quoted 4 lines

I got a message with a trace similar to this from Vladimir before I submitted that patch. I'm not sure how to annotate this, since the xattr_sem can never be taken in the manner described. Internal inodes are protected by I_PRIVATE.

Hm, then maybe my case was just a coincidence.

FWIW, I can reproduce the deadlock warning at each time I boot non-patched 2.6.21, and after reverting the patch, it disappeared.

Ok, so I took another look at the report Vladimir sent me. The trace he ran into was in the delete inode path, but was still a race between the xattr_sem and the inode sem. Since we're locking the xattr root on the xattr read path now, this condition arises more freqently, but it's really the same one he reported.

Quoted 6 lines
Quoted 3 lines

I'm using the default openSUSE config, which doesn't enable mutex debugging. I'll rebuild with it, and hopefully come up with a way to kill the warning.

I still didn't get the warning, but can you try this and let me know if it fixes it?

I also reported this in another thread. With this patch, I haven't seen the tracing anymore.

Tony