119981 – Internal error xfs_bmapi_read after resume from hibernate

Bug 119981 - Internal error xfs_bmapi_read after resume from hibernate

Summary: Internal error xfs_bmapi_read after resume from hibernate

Status:	NEW

Alias:	None

Product:	File System
Classification:	Unclassified
Component:	XFS (show other bugs)
Hardware:	x86-64 Linux

Importance:	P1 normal
Assignee:	XFS Guru

URL:
Keywords:

Depends on:
Blocks:

Reported:	2016-06-10 16:49 UTC by Orion Poplawski
Modified:	2016-06-10 16:55 UTC (History)
CC List:	0 users

See Also:
Kernel Version:	4.6.0
Tree:	Mainline
Regression:	No

Attachments
Add an attachment (proposed patch, testcase, etc.)

Description Orion Poplawski 2016-06-10 16:49:06 UTC

We started seeing XFS trouble after a resume from hibernate:

Jun  5 21:09:57 adelie systemd: Stopping Hibernate.
Jun  5 21:09:57 adelie kernel: XFS (sda2): Internal error xfs_bmapi_read at line 4042 of file fs/
xfs/libxfs/xfs_bmap.c.  Caller xfs_free_eofblocks+0xc6/0x240 [xfs]
Jun  5 21:09:57 adelie kernel: CPU: 0 PID: 11186 Comm: ansible Tainted: G        W I     4.6.0-1.
el7.elrepo.x86_64 #1
Jun  5 21:09:57 adelie kernel: Hardware name: Hewlett-Packard HP EliteBook 8730w/30EC, BIOS 68PAD
 Ver. F.20 12/07/2011
Jun  5 21:09:57 adelie kernel: 0000000000000286 000000000a66f365 ffff8800692cbcf0 ffffffff81344ad
f
Jun  5 21:09:57 adelie kernel: ffff880231eebc00 0000000000000000 ffff8800692cbd08 ffffffffa0430fa
b
Jun  5 21:09:57 adelie kernel: ffffffffa042b3e6 ffff8800692cbdb8 ffffffffa0401f63 ffff8800692cbd6
8
Jun  5 21:09:57 adelie kernel: Call Trace:
Jun  5 21:09:57 adelie kernel: [<ffffffff81344adf>] dump_stack+0x63/0x84
Jun  5 21:09:57 adelie kernel: [<ffffffffa0430fab>] xfs_error_report+0x3b/0x40 [xfs]
Jun  5 21:09:57 adelie kernel: [<ffffffffa042b3e6>] ? xfs_free_eofblocks+0xc6/0x240 [xfs]
Jun  5 21:09:57 adelie kernel: [<ffffffffa0401f63>] xfs_bmapi_read+0x2b3/0x2f0 [xfs]
Jun  5 21:09:57 adelie kernel: [<ffffffff81197ec7>] ? __lru_cache_add+0x57/0x70
Jun  5 21:09:57 adelie kernel: [<ffffffff81714982>] ? down_read+0x12/0x30
Jun  5 21:09:57 adelie kernel: [<ffffffffa042b3e6>] xfs_free_eofblocks+0xc6/0x240 [xfs]
Jun  5 21:09:57 adelie kernel: [<ffffffffa0441bec>] xfs_release+0x8c/0x150 [xfs]
Jun  5 21:09:57 adelie kernel: [<ffffffffa0432035>] xfs_file_release+0x15/0x20 [xfs]
Jun  5 21:09:57 adelie kernel: [<ffffffff81211f57>] __fput+0xe7/0x220
Jun  5 21:09:57 adelie kernel: [<ffffffff812120ce>] ____fput+0xe/0x10
Jun  5 21:09:57 adelie kernel: [<ffffffff8109dd07>] task_work_run+0x77/0x90
Jun  5 21:09:57 adelie kernel: [<ffffffff8107a55c>] exit_to_usermode_loop+0x73/0xa2
Jun  5 21:09:57 adelie kernel: [<ffffffff81003ca5>] do_syscall_64+0xf5/0x110
Jun  5 21:09:57 adelie kernel: [<ffffffff817164e1>] entry_SYSCALL64_slow_path+0x25/0x25

and then on each subsequent resume from hibernate:

Jun  6 14:30:39 adelie systemd: Stopping Hibernate.
Jun  6 14:30:39 adelie kernel: XFS (sda2): Internal error xfs_bmapi_read at line 4042 of file fs/
xfs/libxfs/xfs_bmap.c.  Caller xfs_free_eofblocks+0xc6/0x240 [xfs]
Jun  6 14:30:39 adelie kernel: CPU: 1 PID: 11505 Comm: ansible Tainted: G        W I     4.6.0-1.
el7.elrepo.x86_64 #1
Jun  6 14:30:39 adelie kernel: Hardware name: Hewlett-Packard HP EliteBook 8730w/30EC, BIOS 68PAD
 Ver. F.20 12/07/2011
Jun  6 14:30:39 adelie kernel: 0000000000000286 00000000f3878830 ffff88018d257cf0 ffffffff81344ad
f
Jun  6 14:30:39 adelie kernel: ffff880231eebc00 0000000000000000 ffff88018d257d08 ffffffffa0430fa
b
Jun  6 14:30:39 adelie kernel: ffffffffa042b3e6 ffff88018d257db8 ffffffffa0401f63 ffff88018d257d6
8
Jun  6 14:30:39 adelie kernel: Call Trace:
Jun  6 14:30:39 adelie kernel: [<ffffffff81344adf>] dump_stack+0x63/0x84
Jun  6 14:30:39 adelie kernel: [<ffffffffa0430fab>] xfs_error_report+0x3b/0x40 [xfs]
Jun  6 14:30:39 adelie kernel: [<ffffffffa042b3e6>] ? xfs_free_eofblocks+0xc6/0x240 [xfs]
Jun  6 14:30:39 adelie kernel: [<ffffffffa0401f63>] xfs_bmapi_read+0x2b3/0x2f0 [xfs]
Jun  6 14:30:39 adelie kernel: [<ffffffff81197ec7>] ? __lru_cache_add+0x57/0x70
Jun  6 14:30:39 adelie kernel: [<ffffffff81714982>] ? down_read+0x12/0x30
Jun  6 14:30:39 adelie kernel: [<ffffffffa042b3e6>] xfs_free_eofblocks+0xc6/0x240 [xfs]
Jun  6 14:30:39 adelie kernel: [<ffffffff810b7554>] ? set_next_entity+0x494/0x7c0
Jun  6 14:30:39 adelie kernel: [<ffffffffa0441bec>] xfs_release+0x8c/0x150 [xfs]
Jun  6 14:30:39 adelie kernel: [<ffffffffa0432035>] xfs_file_release+0x15/0x20 [xfs]
Jun  6 14:30:39 adelie kernel: [<ffffffff81211f57>] __fput+0xe7/0x220
Jun  6 14:30:39 adelie kernel: [<ffffffff812120ce>] ____fput+0xe/0x10
Jun  6 14:30:39 adelie kernel: [<ffffffff8109dd07>] task_work_run+0x77/0x90
Jun  6 14:30:39 adelie kernel: [<ffffffff8107a55c>] exit_to_usermode_loop+0x73/0xa2
Jun  6 14:30:39 adelie kernel: [<ffffffff81003ca5>] do_syscall_64+0xf5/0x110
Jun  6 14:30:39 adelie kernel: [<ffffffff817164e1>] entry_SYSCALL64_slow_path+0x25/0x25

But not apparently on normal boots:

Jun  8 21:30:31 adelie systemd: Starting File System Check on /dev/disk/by-uuid/7c371789-3bd6-433
c-a85a-238ea53f8ce7...
Jun  8 21:30:31 adelie systemd-fsck: /sbin/fsck.xfs: XFS file system.
Jun  8 21:30:31 adelie systemd: Started File System Check on /dev/disk/by-uuid/7c371789-3bd6-433c
-a85a-238ea53f8ce7.
Jun  8 21:30:31 adelie kernel: SGI XFS with ACLs, security attributes, no debug enabled
Jun  8 21:30:31 adelie kernel: XFS (sda2): Mounting V4 Filesystem
Jun  8 21:30:34 adelie kernel: XFS (sda2): Starting recovery (logdev: internal)
Jun  8 21:30:34 adelie kernel: XFS (sda2): Ending recovery (logdev: internal)

Eventually system locked up completely.  mount would fail with:

mount: Structure needs cleaning

xfs_repair would fail with:

ERROR: The filesystem has valuable metadata changes in a log which needs to
be replayed. Mount the filesystem to replay the log, and unmount it before
re-running xfs_check. If you are unable to mount the filesystem, then use
the xfs_repair -L option to destroy the log and attempt a repair.
Note that destroying the log may cause corruption -- please attempt a mount
of the filesystem before doing this.

xfs_repair -L allowed the filesystem to be mounted.

Comment 1 Orion Poplawski 2016-06-10 16:55:38 UTC

Only indication of any disk issue was this during an earlier boot of 4.6.0-1.el7:

May 27 11:26:25 adelie kernel: ata1.00: exception Emask 0x0 SAct 0x8000 SErr 0x40000 action 0x6
May 27 11:26:25 adelie kernel: ata1.00: irq_stat 0x40000008
May 27 11:26:25 adelie kernel: ata1: SError: { CommWake }
May 27 11:26:25 adelie kernel: ata1.00: failed command: READ FPDMA QUEUED
May 27 11:26:25 adelie kernel: ata1.00: cmd 60/a0:78:b8:7e:71/00:00:04:00:00/40 tag 15 ncq 81920
in#012         res 41/84:a0:38:7f:71/00:00:04:00:00/00 Emask 0x410 (ATA bus error) <F>
May 27 11:26:25 adelie kernel: ata1.00: status: { DRDY ERR }
May 27 11:26:25 adelie kernel: ata1.00: error: { ICRC ABRT }
May 27 11:26:25 adelie kernel: ata1: hard resetting link
May 27 11:26:26 adelie kernel: ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
May 27 11:26:26 adelie kernel: ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
May 27 11:26:26 adelie kernel: ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
May 27 11:26:26 adelie kernel: ata1.00: configured for UDMA/133
May 27 11:26:26 adelie kernel: ata1: EH complete

Note You need to log in before you can comment on or make changes to this bug.