I got this in 2.6.9-rc1-mm3 and never managed to reproduce it again, i
decided to send it in anyway. Plus i swear i've run into it before
(lock breaking?) ...
Unable to handle kernel paging request at virtual address 6b6b6b93
printing eip:
c0262d0b
*pde = 00000000
Oops: 0000 [#1]
PREEMPT SMP DEBUG_PAGEALLOC
Modules linked in:
CPU: 0
EIP: 0060:[<c0262d0b>] Not tainted VLI
EFLAGS: 00010202 (2.6.9-rc1-mm3)
EIP is at __journal_clean_checkpoint_list+0x14b/0x1b0
eax: f7518888 ebx: 6b6b6b6b ecx: 00000000 edx: f6d68000
esi: f6d68000 edi: e3fc4cd0 ebp: f6d69da0 esp: f6d69d78
ds: 007b es: 007b ss: 0068
Process kjournald (pid: 798, threadinfo=f6d68000 task=f6d2ba90)
Stack: e3fc4cd0 f6d68000 e3fc4cd0 000000ff f7518914 f6da5888 f7518888 00000000
f6d68000 00000001 f6d69f4c c025fb31 f6e51df8 f6da5888 00000003 00000d36
f6d69dc0 f6e51edc f6e51e8c f6d69dc0 00000000 f6da58e0 00000001 f6da794c
Call Trace:
[<c010852f>] show_stack+0x7f/0xa0
[<c01086df>] show_registers+0x15f/0x1d0
[<c0108933>] die+0x123/0x220
[<c011ef05>] do_page_fault+0x255/0x5f2
[<c01080a9>] error_code+0x2d/0x38
[<c025fb31>] journal_commit_transaction+0x3a1/0x1d60
[<c0264240>] kjournald+0x120/0x3d0
[<c0105395>] kernel_thread_helper+0x5/0x10
Code: b6 82 e4 00 00 00 84 c0 7f 4c 8b 45 08 c6 80 e4 00 00 00 01 8b 55 dc 8b 42 08 ff 4a 14 a8 08 75 2e 8b 45 f0 8b 58 28 85 db 74 09 <8b> 43 28 8b 55 f0 89 42 28 8b 45 08 8b 40 40 85 c0 89 45 f0 74
(gdb) list *__journal_clean_checkpoint_list+0x14b
0xc0262d0b is in __journal_clean_checkpoint_list (fs/jbd/checkpoint.c:509).
504 * transaction's buffer list and the checkpoint list to
505 * try to avoid quadratic behaviour.
506 */
507 jh = transaction->t_checkpoint_list;
508 if (jh)
509 transaction->t_checkpoint_list = jh->b_cpnext;
510
511 transaction = journal->j_checkpoint_transactions;