Public bug reported:
We've observed issues with XFS on ARM hardware on both 12.04 and 12.10.
Details from ubuntu-bug are attached from a 12.10 installation. The
specific load in this case is setting up and running a single node
install of Swift and then exercising it with swift-bench for a prolonged
period. After a few hours you get the following:
[361819.562519] SGI XFS with ACLs, security attributes, large block/inode
numbers, no debug enabled
[361819.566196] XFS (loop0): Mounting Filesystem
[361819.617186] XFS (loop0): Ending clean mount
[372224.630644] alloc_vmap_area: 18 callbacks suppressed
[372224.630656] vmap allocation for size 2097152 failed: use vmalloc=<size> to
increase size.
[372224.630687] vmap allocation for size 2097152 failed: use vmalloc=<size> to
increase size.
[372224.630716] vmap allocation for size 2097152 failed: use vmalloc=<size> to
increase size.
[372224.630730] XFS (loop0): xfs_buf_get: failed to map pages
[372224.630730]
[372224.630745] XFS (loop0): Internal error xfs_trans_cancel at line 1466 of
file /build/buildd/linux-3.5.0/fs/xfs/xfs_trans.c. Caller 0xbf0465e0
[372224.630745]
[372224.645280] [<c00164cc>] (unwind_backtrace+0x0/0x104) from [<c04ed184>]
(dump_stack+0x20/0x24)
[372224.645475] [<c04ed184>] (dump_stack+0x20/0x24) from [<bf03391c>]
(xfs_error_report+0x60/0x6c [xfs])
[372224.645800] [<bf03391c>] (xfs_error_report+0x60/0x6c [xfs]) from
[<bf0863f8>] (xfs_trans_cancel+0xfc/0x11c [xfs])
[372224.646117] [<bf0863f8>] (xfs_trans_cancel+0xfc/0x11c [xfs]) from
[<bf0465e0>] (xfs_create+0x228/0x558 [xfs])
[372224.646387] [<bf0465e0>] (xfs_create+0x228/0x558 [xfs]) from [<bf03d7cc>]
(xfs_vn_mknod+0x9c/0x180 [xfs])
[372224.646648] [<bf03d7cc>] (xfs_vn_mknod+0x9c/0x180 [xfs]) from [<bf03d8f0>]
(xfs_vn_create+0x1c/0x20 [xfs])
[372224.646791] [<bf03d8f0>] (xfs_vn_create+0x1c/0x20 [xfs]) from [<c01359c0>]
(vfs_create+0xb4/0x120)
[372224.646805] [<c01359c0>] (vfs_create+0xb4/0x120) from [<c0137c28>]
(do_last+0x860/0x9bc)
[372224.646818] [<c0137c28>] (do_last+0x860/0x9bc) from [<c0137fc8>]
(path_openat+0xcc/0x428)
[372224.646829] [<c0137fc8>] (path_openat+0xcc/0x428) from [<c0138444>]
(do_filp_open+0x3c/0x90)
[372224.646844] [<c0138444>] (do_filp_open+0x3c/0x90) from [<c0128240>]
(do_sys_open+0xfc/0x1d0)
[372224.646855] [<c0128240>] (do_sys_open+0xfc/0x1d0) from [<c0128340>]
(sys_open+0x2c/0x30)
[372224.646868] [<c0128340>] (sys_open+0x2c/0x30) from [<c000e320>]
(ret_fast_syscall+0x0/0x30)
[372224.646881] XFS (loop0): xfs_do_force_shutdown(0x8) called from line 1467
of file /build/buildd/linux-3.5.0/fs/xfs/xfs_trans.c. Return address =
0xbf08640c
[372224.647010] XFS (loop0): Corruption of in-memory data detected. Shutting
down filesystem
[372224.655289] XFS (loop0): Please umount the filesystem and rectify the
problem(s)
[372227.273367] XFS (loop0): xfs_log_force: error 5 returned.
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
** Attachment added: "Output of 'ubuntu-bug linux'"
https://bugs.launchpad.net/bugs/1176977/+attachment/3667050/+files/apport.linux-image-3.5.0-27-highbank.u_74vs.apport
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1176977
Title:
XFS instability on armhf under load
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1176977/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
