Hi All,
Have encountered a repeatable OOPS in 2.4.18 that does not occur in
2.4.19pre9 or pre10 (not sure about other preX patches). Since this
seems fixed in upcoming 2.4.19, this may be moot; but reported here for
completeness.
The oops is triggered by large amounts of disk buffer page management
(e.g. tar xzf linux-2.4.18.tar.gz) in combination with a RealTek
RTL8139C ethernet controller chip PCI card installed in the system.
Interestingly, driver support for the 8139 need not be compiled into the
kernel; the mere presence of the card is sufficient, suggesting perhaps
a shared IRQ or other resource overlap. More interestingly, the 8139
driver, when loaded, works even after OOPSes are triggered.
A representative OOPS in text is below, a few more GZIPped, and finally
dot.config.gz.
Hardware: Soltek DRV2, Athlon XP, IDE disk and CDROM on hda and hdb,
Radeon 8500LEE, SIIG USB2.0+IEEE1394 card via Hint PCI->PCI bridge.
Kris
Jun 22 19:28:44 madmax kernel: cpu: 0, clocks: 2686339, slice: 1343169
Jun 22 19:30:21 madmax kernel: invalid operand: 0000
Jun 22 19:30:21 madmax kernel: CPU: 0
Jun 22 19:30:21 madmax kernel: EIP: 0010:[rmqueue+316/432] Not tainted
Jun 22 19:30:21 madmax kernel: EIP: 0010:[<c012b89c>] Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
Jun 22 19:30:21 madmax kernel: EFLAGS: 00010087
Jun 22 19:30:21 madmax kernel: eax: c02c8e28 ebx: c02c8e40 ecx: 00000001 edx: c17fff80
Jun 22 19:30:21 madmax kernel: esi: c154f4c0 edi: 00000000 ebp: 00000001 esp: de3c5ef4
Jun 22 19:30:21 madmax kernel: ds: 0018 es: 0018 ss: 0018
Jun 22 19:30:21 madmax kernel: Process tar (pid: 151, stackpage=de3c5000)
Jun 22 19:30:21 madmax kernel: Stack: c02c8f9c 000001ff 0000000c 00000000 000143d3 00000282 00000000 c02c8e28
Jun 22 19:30:21 madmax kernel: c012bad3 000001d2 de3c5f7c 0000000c c18fd3dc c02c8e28 c02c8f98 000001d2
Jun 22 19:30:21 madmax kernel: c0127781 c012b926 00000000 c012779d de95c940 de95c940 ffffffea 00002800
Jun 22 19:30:21 madmax kernel: Call Trace: [__alloc_pages+51/368] [generic_file_write+993/1728] [_alloc_pages+22/32] [generic_file_write+1021/1728] [ext3_file_write+70/80]
Jun 22 19:30:23 madmax kernel: Call Trace: [<c012bad3>] [<c0127781>] [<c012b926>] [<c012779d>] [<c014f2f6>]
Jun 22 19:30:24 madmax kernel: [<c0131156>] [<c0106c5b>]
Jun 22 19:30:26 madmax kernel: Code: 0f 0b ff 74 24 14 9d c7 46 14 01 00 00 00 8b 44 24 1c 3b 46
>>EIP; c012b89c <rmqueue+13c/1b0> <=====
>>eax; c02c8e28 <contig_page_data+a8/320>
>>ebx; c02c8e40 <contig_page_data+c0/320>
>>edx; c17fff80 <_end+14a83d4/24544454>
>>esi; c154f4c0 <_end+11f7914/24544454>
>>esp; de3c5ef4 <_end+1e06e348/24544454>
Trace; c012bad3 <__alloc_pages+33/170>
Trace; c0127781 <generic_file_write+3e1/6c0>
Trace; c012b926 <_alloc_pages+16/20>
Trace; c012779d <generic_file_write+3fd/6c0>
Trace; c014f2f6 <ext3_file_write+46/50>
Trace; c0131156 <sys_write+96/f0>
Trace; c0106c5b <system_call+33/38>
Code; c012b89c <rmqueue+13c/1b0>
00000000 <_EIP>:
Code; c012b89c <rmqueue+13c/1b0> <=====
0: 0f 0b ud2a <=====
Code; c012b89e <rmqueue+13e/1b0>
2: ff 74 24 14 pushl 0x14(%esp,1)
Code; c012b8a2 <rmqueue+142/1b0>
6: 9d popf
Code; c012b8a3 <rmqueue+143/1b0>
7: c7 46 14 01 00 00 00 movl $0x1,0x14(%esi)
Code; c012b8aa <rmqueue+14a/1b0>
e: 8b 44 24 1c mov 0x1c(%esp,1),%eax
Code; c012b8ae <rmqueue+14e/1b0>
12: 3b 46 00 cmp 0x0(%esi),%eax
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Jun 23 2002 - 22:00:26 EST