#254 Bad page state in process 'jfsCommit'

[Anonymous]

After I resized the logical volume of my /home filesystem and remounted the filesystem with the option resize, I've observed the following kernel crash:

ea_get: invalid extended attribute
Bad page state in process 'jfsCommit'
page:c108ce60 flags:0x4000080d mapping:c8160d54 mapcount:0 count:0
Trying to fix it up, but a reboot is needed
Backtrace:
[bad_page+106/176] bad_page+0x6a/0xb0
[free_hot_cold_page+355/368] free_hot_cold_page+0x163/0x170
[<d0c07050>] jfs_lazycommit+0x0/0x230 [jfs]
[<d0c0448e>] txUnlock+0x17e/0x280 [jfs]
[<d0c07050>] jfs_lazycommit+0x0/0x230 [jfs]
[<d0c07252>] jfs_lazycommit+0x202/0x230 [jfs]
[default_wake_function+0/16] default_wake_function+0x0/0x10
[kthread+186/240] kthread+0xba/0xf0
[kthread+0/240] kthread+0x0/0xf0
[kernel_thread_helper+7/16] kernel_thread_helper+0x7/0x10
=======================
------------[ cut here ]------------
kernel BUG at mm/filemap.c:537!
invalid opcode: 0000 [#1]
SMP
Modules linked in: udf binfmt_misc ipv6 ppdev ... tileblit font bitblit softcursor vesafb capability commoncap
CPU: 0
EIP: 0060:[unlock_page+14/48] Tainted: G B VLI
EFLAGS: 00010246 (2.6.20-15-generic #2)
EIP is at unlock_page+0xe/0x30
eax: 00000000 ebx: c108ce60 ecx: 00000292 edx: c108ce60
esi: c5e7a498 edi: 01cef3cc ebp: 00000292 esp: cada9f38
ds: 007b es: 007b ss: 0068
Process jfsCommit (pid: 4271, ti=cada8000 task=ce342050 task.ti=cada8000)
Stack: c5e7a498 d0c00d2a 4000080d d0c07050 00000256 00000212 00000000 d0c18240
c5e7a498 01cef3cc 00000292 d0c04405 d0baf328 d0baf324 00000000 00000000
00078f94 ca8ef108 00000089 d0baf2e0 00000003 ca8ef060 d0baf2e0 ca1865c0
Call Trace:
[<d0c00d2a>] put_metapage+0x6a/0x130 [jfs]
[<d0c07050>] jfs_lazycommit+0x0/0x230 [jfs]
[<d0c04405>] txUnlock+0xf5/0x280 [jfs]
[<d0c07050>] jfs_lazycommit+0x0/0x230 [jfs]
[<d0c07252>] jfs_lazycommit+0x202/0x230 [jfs]
[default_wake_function+0/16] default_wake_function+0x0/0x10
[kthread+186/240] kthread+0xba/0xf0
[kthread+0/240] kthread+0x0/0xf0
[kernel_thread_helper+7/16] kernel_thread_helper+0x7/0x10
=======================
Code: 50 08 85 d2 74 04 89 c8 ff d2 8d 76 00 e8 db 62 19 00 31 c0 c3 ba 60 91 3a c0 eb db 90 53 89 c3 90 0f ba 30 00 19 c0 85 c0 75 04 <0f> 0b eb fe 89 d8 e8 b7 f7 ff ff 89 da 31 c9 5b e9 ed 3b fe ff
EIP: [unlock_page+14/48] unlock_page+0xe/0x30 SS:ESP 0068:cada9f38
<0>Bad page state in process 'scribus'
page:c108ce60 flags:0x4000000c mapping:00000000 mapcount:0 count:1
Trying to fix it up, but a reboot is needed
Backtrace:
[bad_page+106/176] bad_page+0x6a/0xb0
[get_page_from_freelist+800/880] get_page_from_freelist+0x320/0x370
[__alloc_pages+205/832] __alloc_pages+0xcd/0x340
[read_swap_cache_async+129/176] read_swap_cache_async+0x81/0xb0
[swapin_readahead+86/112] swapin_readahead+0x56/0x70
[__handle_mm_fault+1979/2624] __handle_mm_fault+0x7bb/0xa40
[do_page_fault+296/1520] do_page_fault+0x128/0x5f0
[do_page_fault+0/1520] do_page_fault+0x0/0x5f0
[error_code+124/144] error_code+0x7c/0x90
=======================

I use the standard 2.6.20-15-generic from Ubuntu feisty.

[David Kleikamp]

Logged In: YES
user_id=422440
Originator: NO

This bug doesn't look familiar. I will investigate it.

[Nobody/Anonymous]

Logged In: NO

It happened once again. This time with ubuntu kernel 2.6.20-16-generic.
Here's the stack trace:
kernel BUG at mm/filemap.c:537!
invalid opcode: 0000 [#1]
SMP
Modules linked in: udf binfmt_misc ipv6 ... tileblit font bitblit softcursor vesafb capability commo
CPU: 0
EIP: 0060:[unlock_page+14/48] Tainted: P VLI
EFLAGS: 00010246 (2.6.20-16-generic #2)
EIP is at unlock_page+0xe/0x30
eax: 00000000 ebx: c11711e0 ecx: 00000202 edx: 00000000
esi: c07b4e28 edi: c9d1066c ebp: c11212e0 esp: c332bc40
ds: 007b es: 007b ss: 0068
Process beagled (pid: 5864, ti=c332a000 task=c7f03580 task.ti=c332a000)
Stack: c07b4e28 d0c1c63c 00000000 cac8ab84 c07b4e50 00001000 0000006c c9d105c8
001e4000 00000000 00000000 00000001 c7f03580 c0120b40 c07b4e54 c07b4e54
00000000 0000006c 00000000 00000000 d0c11a33 00000000 00000000 d0c1c130
Call Trace:
[<d0c1c63c>] __get_metapage+0x30c/0x420 [jfs]
[default_wake_function+0/16] default_wake_function+0x0/0x10
[<d0c11a33>] dbAlloc+0x383/0x5a0 [jfs]
[<d0c1c130>] metapage_readpage+0x0/0x1a0 [jfs]
[<d0c23953>] ea_write+0x133/0x470 [jfs]
[<d0c24438>] __jfs_setxattr+0x608/0x660 [jfs]
[<d0c1efc1>] txBegin+0x21/0x370 [jfs]
[<d0c246f8>] jfs_setxattr+0x88/0xe0 [jfs]
[<d0c24670>] jfs_setxattr+0x0/0xe0 [jfs]
[vfs_setxattr+193/576] vfs_setxattr+0xc1/0x240
[setxattr+164/272] setxattr+0xa4/0x110
[__link_path_walk+2527/3696] __link_path_walk+0x9df/0xe70
[mntput_no_expire+36/160] mntput_no_expire+0x24/0xa0
[link_path_walk+101/192] link_path_walk+0x65/0xc0
[enable_8259A_irq+13/80] enable_8259A_irq+0xd/0x50
[handle_level_irq+193/288] handle_level_irq+0xc1/0x120
[do_IRQ+69/128] do_IRQ+0x45/0x80
[do_path_lookup+131/448] do_path_lookup+0x83/0x1c0
[getname+167/208] getname+0xa7/0xd0
[__user_walk_fd+73/96] __user_walk_fd+0x49/0x60
[sys_lsetxattr+72/96] sys_lsetxattr+0x48/0x60
[enable_8259A_irq+13/80] enable_8259A_irq+0xd/0x50
[handle_level_irq+193/288] handle_level_irq+0xc1/0x120
[do_IRQ+69/128] do_IRQ+0x45/0x80
[sysenter_past_esp+105/169] sysenter_past_esp+0x69/0xa9
=======================
Code: 50 08 85 d2 74 04 89 c8 ff d2 8d 76 00 e8 db 63 19 00 31 c0 c3 ba 60 b1 3a c0 eb db 90 53 89 c3 90 0f ba 30 00 19 c0 85 c0 75 04 <0f> 0b eb fe 89 d8 e8 b7 f7 ff ff 89 da 31 c9 5b e9 ed 3b fe ff
EIP: [unlock_page+14/48] unlock_page+0xe/0x30 SS:ESP 0068:c332bc40

And here's the output of fsck -nd /home:

fsck 1.40-WIP (14-Nov-2006)
fsck.jfs version 1.1.11, 05-Jun-2006
processing started: 7/1/2007 12.18.33
The current device is: /dev/mapper/vg01-home [xchkdsk.c:1520]
Open(...READONLY...) returned rc = 0 [fsckpfs.c:3199]
Primary superblock is valid. [fsckmeta.c:1551]
The type of file system for the device is JFS. [xchkdsk.c:1537]
Block size in bytes: 4096 [xchkdsk.c:1850]
Filesystem size in blocks: 3932160 [xchkdsk.c:1857]
**Phase 1 - Check Blocks, Files/Directories, and Directory Entries [xchkdsk.c:1989]
**Phase 2 - Count links [xchkdsk.c:2080]
**Phase 3 - Duplicate Block Rescan and Directory Connectedness [xchkdsk.c:2113]
**Phase 4 - Report Problems [xchkdsk.c:2191]
**Phase 5 - Check Connectivity [xchkdsk.c:2223]
**Phase 6 - Perform Approved Corrections [xchkdsk.c:2252]
**Phase 7 - Verify File/Directory Allocation Maps [xchkdsk.c:2386]
Discrepancies detected in the Free Inode List for Allocation Group F23. (3) [fsckimap.c:2501]
Inode Allocation Group F23 is inconsistent. [fsckimap.c:2950]
The Free Inode List is inconsistent for Allocation Group F0. [fsckimap.c:498]
Errors detected in the Fileset File/Directory Allocation Map. (F) [fsckimap.c:3181]
**Phase 8 - Verify Disk Allocation Maps [xchkdsk.c:2421]
Filesystem Summary: [xchkdsk.c:1009]
Blocks in use for inodes: 14056 [xchkdsk.c:1014]
Inode count: 112448 [xchkdsk.c:1016]
File count: 87560 [xchkdsk.c:1018]
Directory count: 8299 [xchkdsk.c:1020]
Block count: 3932160 [xchkdsk.c:1022]
Free block count: 664775 [xchkdsk.c:1024]
15728640 kilobytes total disk space. [xchkdsk.c:1067]
40313 kilobytes in 8299 directories. [xchkdsk.c:1070]
12917412 kilobytes in 87560 user files. [xchkdsk.c:1073]
67376 kilobytes in extended attributes [xchkdsk.c:1075]
0 kilobytes in access control lists [xchkdsk.c:1077]
125065 kilobytes reserved for system use. [xchkdsk.c:1079]
2659100 kilobytes are available for use. [xchkdsk.c:1081]
File system checked READ ONLY. [xchkdsk.c:446]
Filesystem is dirty. [fsckmeta.c:142]
Filesystem is dirty but is marked clean. In its present state,
the results of accessing /dev/mapper/vg01-home (except by this utility) are undefined. [fsckmeta.c:148]
processing terminated: 7/1/2007 12:21:42 with return code: 0 exit code: 4. [xchkdsk.c:468]

[David Kleikamp]

Logged In: YES
user_id=422440
Originator: NO

Did the second crash also involve resizing the file system?