XFS oops in vanilla 2.6.24
Hi there,
I get the following with 2.6.24:
Ending clean XFS mount for filesystem: dm-0
BUG: unable to handle kernel paging request at virtual address f2134000 printing eip: c021a13a *pde = 010b5067 *pte = 32134000 Oops: 0000 [#1] PREEMPT DEBUG_PAGEALLOC Modules linked in: radeon drm rfcomm l2cap sym53c8xx scsi_transport_spi snd_via82xx 8139too snd_mpu401_uart snd_ens1371 snd_rawmidi snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd_page_alloc via_agp agpgart
Pid: 3889, comm: bash Not tainted (2.6.24 #3)
EIP: 0060:[<c021a13a>] EFLAGS: 00010282 CPU: 0
EIP is at xfs_file_readdir+0xfa/0x18c
EAX: 00000000 EBX: 000002f5 ECX: 00000020 EDX: 00000000
ESI: 00000000 EDI: f2133ff8 EBP: f227ff68 ESP: f227ff10
DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
Process bash (pid: 3889, ti=f227e000 task=f7205a80 task.ti=f227e000)
Stack: 000002f5 00000000 2c000137 00000000 00000000 c0165358 f227ff94 f221c810
f2d85e48 00000000 00000000 00000000 000002f5 00000000 f2133000 00001000
00000ff8 000002f9 00000000 c0421c80 f221c810 f1cdbe48 f227ff88 c0165543
Call Trace:
[<c0104da3>] show_trace_log_lvl+0x1a/0x2f
[<c0104e53>] show_stack_log_lvl+0x9b/0xa3
[<c0104efb>] show_registers+0xa0/0x1e2
[<c010514c>] die+0x10f/0x1dd
[<c0113555>] do_page_fault+0x43a/0x519
[<c040fe52>] error_code+0x6a/0x70
[<c0165543>] vfs_readdir+0x5d/0x89
[<c01655cd>] sys_getdents64+0x5e/0xa0
[<c0103f06>] syscall_call+0x7/0xb
=======================
Code: 89 74 24 04 81 e3 ff ff ff 7f 89 1c 24 ff 55 bc 85 c0 0f 85 82 00 00 00 8b 4f 10 31 d2 83 c1 1f 83 e1 f8 29 4d d0 19 55 d4 01 cf <8b> 57 08 8b 4f 0c 89 55 d8 89 4d dc 83 7d d4 00 7f a1 7c 06 83
EIP: [<c021a13a>] xfs_file_readdir+0xfa/0x18c SS:ESP 0068:f227ff10
---[ end trace e518e1370efb695e ]---Sven
Signature · 4 lines
"Software is like sex; it's better when it's free" (Linus Torvalds) /me is giggls@ircnet, http://sven.gegg.us/ on the Web
Lets CC the XFS maintainer..
Quoted 37 lines · Sven Geggus
Hi there,
I get the following with 2.6.24:
Ending clean XFS mount for filesystem: dm-0
BUG: unable to handle kernel paging request at virtual address f2134000 printing eip: c021a13a *pde = 010b5067 *pte = 32134000 Oops: 0000 [#1] PREEMPT DEBUG_PAGEALLOC Modules linked in: radeon drm rfcomm l2cap sym53c8xx scsi_transport_spi snd_via82xx 8139too snd_mpu401_uart snd_ens1371 snd_rawmidi snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd_page_alloc via_agp agpgartPid: 3889, comm: bash Not tainted (2.6.24 #3)
EIP: 0060:[<c021a13a>] EFLAGS: 00010282 CPU: 0
EIP is atxfs_file_readdir+0xfa/0x18c
EAX: 00000000 EBX:000002f5ECX: 00000020 EDX: 00000000
ESI: 00000000 EDI:f2133ff8EBP:f227ff68ESP:f227ff10
DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
Process bash (pid: 3889, ti=f227e000 task=f7205a80 task.ti=f227e000)Stack: 000002f5 00000000 2c000137 00000000 00000000 c0165358 f227ff94 f221c810 f2d85e48 00000000 00000000 00000000 000002f5 00000000 f2133000 00001000 00000ff8 000002f9 00000000 c0421c80 f221c810 f1cdbe48 f227ff88 c0165543 Call Trace: [<c0104da3>] show_trace_log_lvl+0x1a/0x2f [<c0104e53>] show_stack_log_lvl+0x9b/0xa3 [<c0104efb>] show_registers+0xa0/0x1e2 [<c010514c>] die+0x10f/0x1dd [<c0113555>] do_page_fault+0x43a/0x519 [<c040fe52>] error_code+0x6a/0x70 [<c0165543>] vfs_readdir+0x5d/0x89 [<c01655cd>] sys_getdents64+0x5e/0xa0 [<c0103f06>] syscall_call+0x7/0xb ======================= Code: 89 74 24 04 81 e3 ff ff ff 7f 89 1c 24 ff 55 bc 85 c0 0f 85 82 00 00 00 8b 4f 10 31 d2 83 c1 1f 83 e1 f8 29 4d d0 19 55 d4 01 cf <8b> 57 08 8b 4f 0c 89 55 d8 89 4d dc 83 7d d4 00 7f a1 7c 06 83 EIP: [<c021a13a>] xfs_file_readdir+0xfa/0x18c SS:ESP 0068:f227ff10 ---[ end trace e518e1370efb695e ]---Sven
Quoted 1 line · Peter Zijlstra
Lets CC the XFS maintainer..
Adding the xfs list and hch.
It might be a couple of days before I get to this - I've got a week of backlog to catch up on after LCA....
Quoted 7 lines · Sven Geggus
Quoted 6 lines
Hi there,
I get the following with 2.6.24:
Ending clean XFS mount for filesystem: dm-0
BUG: unable to handle kernel paging request at virtual address f2134000
How long after mount does this happen? Does it happen when listing a specific directory? i.e. do you have a reproducable test case for it?
Cheers,
Dave.
Quoted 31 lines
Quoted 31 lines
printing eip:
c021a13a*pde =010b5067*pte = 32134000Oops: 0000 [#1] PREEMPT DEBUG_PAGEALLOC Modules linked in: radeon drm rfcomm l2cap sym53c8xx scsi_transport_spi snd_via82xx 8139too snd_mpu401_uart snd_ens1371 snd_rawmidi snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd_page_alloc via_agp agpgartPid: 3889, comm: bash Not tainted (2.6.24 #3)
EIP: 0060:[<c021a13a>] EFLAGS: 00010282 CPU: 0
EIP is atxfs_file_readdir+0xfa/0x18c
EAX: 00000000 EBX:000002f5ECX: 00000020 EDX: 00000000
ESI: 00000000 EDI:f2133ff8EBP:f227ff68ESP:f227ff10
DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
Process bash (pid: 3889, ti=f227e000 task=f7205a80 task.ti=f227e000)Stack: 000002f5 00000000 2c000137 00000000 00000000 c0165358 f227ff94 f221c810 f2d85e48 00000000 00000000 00000000 000002f5 00000000 f2133000 00001000 00000ff8 000002f9 00000000 c0421c80 f221c810 f1cdbe48 f227ff88 c0165543 Call Trace: [<c0104da3>] show_trace_log_lvl+0x1a/0x2f [<c0104e53>] show_stack_log_lvl+0x9b/0xa3 [<c0104efb>] show_registers+0xa0/0x1e2 [<c010514c>] die+0x10f/0x1dd [<c0113555>] do_page_fault+0x43a/0x519 [<c040fe52>] error_code+0x6a/0x70 [<c0165543>] vfs_readdir+0x5d/0x89 [<c01655cd>] sys_getdents64+0x5e/0xa0 [<c0103f06>] syscall_call+0x7/0xb ======================= Code: 89 74 24 04 81 e3 ff ff ff 7f 89 1c 24 ff 55 bc 85 c0 0f 85 82 00 00 00 8b 4f 10 31 d2 83 c1 1f 83 e1 f8 29 4d d0 19 55 d4 01 cf <8b> 57 08 8b 4f 0c 89 55 d8 89 4d dc 83 7d d4 00 7f a1 7c 06 83 EIP: [<c021a13a>] xfs_file_readdir+0xfa/0x18c SS:ESP 0068:f227ff10 ---[ end trace e518e1370efb695e ]---Sven
Signature · 3 lines
Dave Chinner Principal Engineer SGI Australian Software Group
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 9 lines · Sven Geggus
David Chinner schrieb am Sonntag, den 03. Februar um 23:13 Uhr:
Quoted 2 lines
How long after mount does this happen? Does it happen when listing a specific directory? i.e. do you have a reproducable test case for it?
Well, it happens very soon after mounting the device. Typically when doing ls or something trivial like this. Unfortunately this is not really reproducable but it does happen reproducable on one specific device here sooner or later when booting into 2.6.24.
Ok. Can you use xfs_metadump to grab an image of the filesystem that you are using to reproduce this and put it up somewhere on the web so I can try to reproduce it locally?
Quoted 3 lines
I run xfs-repair just to make shure that the filesystem is not corrupted. Works fine when booting 2.6.23.8 which I have been using before.
Good - that saves me from having to ask. ;)
Cheers,
Dave.
Signature · 3 lines
Dave Chinner Principal Engineer SGI Australian Software Group