> Last night I've got an oops on a production machine, while it
> was making a backup (tarring a directory with about 200 MB of files).
> The machine is running 2.2.10 for about 4 days. I never had problems
> on this machine with 2.2.9 or older kernels (2.1.120 >).
Is this the only time it's happened or does it happen often with 2.2.10?
> Unable to handle kernel NULL pointer dereference at virtual address 00000001
> current->tss.cr3 = 07466000, %cr3 = 07466000
> *pde = 00000000
> Oops: 0000
> CPU: 0
> EIP: 0010:[find_buffer+40/76]
> EFLAGS: 00010202
> eax: 00000001 ebx: 001616a2 ecx: 00000802 edx: 00000001
> esi: 00000400 edi: 00000014 ebp: c7240aa0 esp: c582be48
> ds: 0018 es: 0018 ss: 0018
> Process tar (pid: 1500, process nr: 24, stackpage=c582b000)
> Stack: c012436b 00000802 001616a2 00000400 c0124fc4 00000802 001616a2 00000400
> 00000000 c582bef4 00000014 c6a7c110 00000100 00000400 c7183000 00000802
> 00000002 c7240f80 00000000 c7240f80 c7240b00 001616a3 c0138934 c6ecd180
> Call Trace: [get_hash_table+23/36] [brw_page+332/872] [ext2_bmap+360/584] [generic_readpage+129/144] [try_to_read_ahead+266/288] [do_generic_file_read+758/1508] [generic_file_read+99/124]
> [file_read_actor+0/80] [sys_read+174/196] [system_call+52/56] [startup_32+43/286]
> Code: 8b 00 39 5a 04 75 11 39 72 08 75 0c 66 39 4a 0c 75 06 89 d0
It looks like one of the hash table entries is 0x1, or else one of the
buffer_heads is 0x1, instead of 0x0. It could be a bit error caused by
faulty memory, I suppose.
Tim.
*/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/