Hi all,
I'm currently investigating a regression that has showed up with my last
git pull yesterday. Bisecting the commits showed bf726e "semaphore: fix"
to be the culprit, reverting it fixed the issue.
Symptoms: During heavy filesystem usage (e.g. a kernel compile) I get
several compiler processes in uninterruptible sleep, blocking all i/o on
the filesystem. System is an Intel Core 2 Quad running a 64bit kernel and
userspace. Filesystem is xfs on top of lvm. See below for the output of
sysrq-w.
Sven
task PC stack pid father
gcc D ffff81000105c100 0 13923 13913
ffff8101ed51be18 0000000000000082 ffff8101ed51beb0 00000010802bf7b8
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff80585ad8>] __mutex_lock_slowpath+0xd8/0x270
[<ffffffff802bfa81>] ? do_path_lookup+0x91/0x220
[<ffffffff803d86ba>] ? strncpy_from_user+0x3a/0x50
[<ffffffff80585c79>] mutex_lock+0x9/0x10
[<ffffffff802c008a>] do_unlinkat+0xaa/0x1d0
[<ffffffff80254cf9>] ? up_read+0x9/0x10
[<ffffffff80589a5c>] ? do_page_fault+0x3ec/0x850
[<ffffffff802c01c1>] sys_unlink+0x11/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
gcc D ffff810001070100 0 13924 13920
ffff8101f6971e18 0000000000000082 ffff8101f6971eb0 00000010802bf7b8
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff80585ad8>] __mutex_lock_slowpath+0xd8/0x270
[<ffffffff802bfa81>] ? do_path_lookup+0x91/0x220
[<ffffffff803d86ba>] ? strncpy_from_user+0x3a/0x50
[<ffffffff80585c79>] mutex_lock+0x9/0x10
[<ffffffff802c008a>] do_unlinkat+0xaa/0x1d0
[<ffffffff80254cf9>] ? up_read+0x9/0x10
[<ffffffff80589a5c>] ? do_page_fault+0x3ec/0x850
[<ffffffff802c01c1>] sys_unlink+0x11/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
gcc D ffff81000107a100 0 13932 13931
ffff8101ed447e18 0000000000000086 ffff8101ed447eb0 00000010802bf7b8
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff80585ad8>] __mutex_lock_slowpath+0xd8/0x270
[<ffffffff802bfa81>] ? do_path_lookup+0x91/0x220
[<ffffffff803d86ba>] ? strncpy_from_user+0x3a/0x50
[<ffffffff80585c79>] mutex_lock+0x9/0x10
[<ffffffff802c008a>] do_unlinkat+0xaa/0x1d0
[<ffffffff80254cf9>] ? up_read+0x9/0x10
[<ffffffff80589a5c>] ? do_page_fault+0x3ec/0x850
[<ffffffff802c01c1>] sys_unlink+0x11/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
sync D ffff81000105c100 0 13996 9142
ffff8101ed481d48 0000000000000086 0000000000000000 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff80586d79>] __down_read+0x79/0xb7
[<ffffffff805860a9>] down_read+0x9/0x10
[<ffffffff80371254>] xfs_ilock+0x44/0x80
[<ffffffff8038bfb8>] xfs_sync_inodes+0x3e8/0x6b0
[<ffffffff8038c3de>] xfs_syncsub+0x15e/0x2d0
[<ffffffff805871f9>] ? _spin_unlock+0x9/0x40
[<ffffffff8038c594>] xfs_sync+0x44/0x60
[<ffffffff8039cd07>] xfs_fs_sync_super+0x37/0x100
[<ffffffff80587309>] ? _spin_unlock_irq+0x9/0x40
[<ffffffff80586da5>] ? __down_read+0xa5/0xb7
[<ffffffff803d5946>] ? __up_read+0x46/0xb0
[<ffffffff802b5c7f>] sync_filesystems+0xdf/0x130
[<ffffffff802da0e9>] do_sync+0x39/0x80
[<ffffffff802da13e>] sys_sync+0xe/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
gcc D ffff810001066100 0 14008 14007
ffff8101f69b5898 0000000000000086 ffff8101fb9e4de0 0000000100000000
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff802554f6>] ? down_trylock+0x36/0x50
[<ffffffff805851c5>] schedule_timeout+0x95/0xd0
[<ffffffff80396d38>] ? xfs_buf_get_flags+0xa8/0x170
[<ffffffff805861ca>] __down+0x5a/0x90
[<ffffffff80255605>] down+0x45/0x50
[<ffffffff8037d152>] xlog_state_get_iclog_space+0xa2/0x200
[<ffffffff8037d3cc>] xlog_write+0x11c/0x530
[<ffffffff8037d990>] xfs_log_write+0x40/0x70
[<ffffffff803897d8>] _xfs_trans_commit+0x2e8/0x3d0
[<ffffffff802c9f8f>] ? igrab+0x2f/0x50
[<ffffffff8038fe56>] xfs_create+0x426/0x480
[<ffffffff8039b00b>] xfs_vn_mknod+0x15b/0x2a0
[<ffffffff8039b16b>] xfs_vn_create+0xb/0x10
[<ffffffff802bd688>] vfs_create+0xe8/0x180
[<ffffffff802c114f>] do_filp_open+0x69f/0x9c0
[<ffffffff805871f9>] ? _spin_unlock+0x9/0x40
[<ffffffff802b1e65>] ? get_unused_fd_flags+0x105/0x130
[<ffffffff802b1eec>] do_sys_open+0x5c/0xf0
[<ffffffff802b1fab>] sys_open+0x1b/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
gcc D ffff81000105c100 0 14011 14010
ffff8101ed51fe18 0000000000000082 0000000000000000 00000010802b1704
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff80585ad8>] __mutex_lock_slowpath+0xd8/0x270
[<ffffffff802bfa81>] ? do_path_lookup+0x91/0x220
[<ffffffff803d86ba>] ? strncpy_from_user+0x3a/0x50
[<ffffffff80585c79>] mutex_lock+0x9/0x10
[<ffffffff802c008a>] do_unlinkat+0xaa/0x1d0
[<ffffffff80254cf9>] ? up_read+0x9/0x10
[<ffffffff80589a5c>] ? do_page_fault+0x3ec/0x850
[<ffffffff802c01c1>] sys_unlink+0x11/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
gcc D ffff81000107a100 0 14012 13987
ffff8101ed409e18 0000000000000082 0000000000000000 00000010802b1704
ffffffff807d7100 ffffffff807d7100 ffffffff807d7100 ffffffff807d7100
ffffffff807d7100 ffffffff807d7100 ffffffff807d3ca0 ffffffff807d7100
Call Trace:
[<ffffffff80585ad8>] __mutex_lock_slowpath+0xd8/0x270
[<ffffffff802bfa81>] ? do_path_lookup+0x91/0x220
[<ffffffff803d86ba>] ? strncpy_from_user+0x3a/0x50
[<ffffffff80585c79>] mutex_lock+0x9/0x10
[<ffffffff802c008a>] do_unlinkat+0xaa/0x1d0
[<ffffffff80254cf9>] ? up_read+0x9/0x10
[<ffffffff80589a5c>] ? do_page_fault+0x3ec/0x850
[<ffffffff802c01c1>] sys_unlink+0x11/0x20
[<ffffffff8020c4db>] system_call_after_swapgs+0x7b/0x80
--
