(my previous email got damaged, I think a KMail cvs-head bug)
Please CC me, because I'm not subscribed to linux kernel any more ...
This is my first real (log piped through ksymoops) kernel oops report,
so I hope it's helpful ...
It would be great if someone would set up an automated stress tester that
will test each kernel before release, so that kernel developers and bug
hunters can concentrate on the really important issues.
Kernel 2.5.10 is the first 2.5 vanilla kerneI that compiles on my machine
since 2.5.4 and this PC is a quite usual one :-((
Pentium2 266MHz,
Intel 440LX Chipset, Gigabyte mainboard,
Advansys SCSI OEM card (bundled with TEAC8x burner)
nVidia riva128 graphic card 4MB RAM,
8GB IDE HDD,
USB ACM modem, USB scanner, ...
Each time I try to boot kernel 2.5.10 I get this kernel oops
(see attachment). The kernel is a monolitic one without modules.
It took me > 1h to write down the oops log but here it is :-)
The ksymoops output:
[root@localhost ~]# ksymoops -v /usr/src/linux/vmlinux -m
/boot/System.map-2.5.10 -K -L -O oops
ksymoops 2.4.1 on i686 2.4.17. Options used
-v /usr/src/linux/vmlinux (specified)
-K (specified)
-L (specified)
-O (specified)
-m /boot/System.map-2.5.10 (specified)
c01b6076
Oops: 0002
CPU: 0
EFLAGS: 00010002
eax: 00000004 ebx: c13f3e94 ecx: c02a0b54 edx: 00000000
esi: c13f3dfc edi: c031b28c ebp: c13dbe48
ds: 0018 es:0018 ss: 0018
Stack: 00000001 00000018 c13f3dfc c031b28c 00000003 c01b9f41 c031b28c
00000001
00000000 c01c1d83 c031b28c 00000001 00000000 00000000 00000000
00000051
00000000 c031b28c c13c52c0 c13da000 c01b7e5d c031b28c c01c1cb0
c031b198
Call Trace: [<c01b9f41>] [<c01c1d83>] [<c01b7e5d>] [<c01c1cb0>] [<c010856a>]
[<c0108761>] [<c010728e>] [<c0110018>] [<c011139e>] [<c0111446>]
[<c01116a0>]
[<c0111d4d>] [<c0111e5d>] [<c01127ad>] [<c0105000>] [<c0150018>]
[<c0105000>]
[<c0105676>] [<c0112510>]
Code: c7 04 90 00 00 00 00 8b 53 0c 8b 87 34 02 00 00 0f b3 10 8b
Using defaults from ksymoops -t elf32-i386 -a i386
Trace; c01b9f41 <ide_end_request+11/20>
Trace; c01c1d83 <cdrom_pc_intr+d3/1e0>
Trace; c01b7e5d <ide_intr+ed/1a0>
Trace; c01c1cb0 <cdrom_pc_intr+0/1e0>
Trace; c010856a <handle_IRQ_event+3a/70>
Trace; c0108761 <do_IRQ+91/f0>
Trace; c010728e <common_interrupt+22/28>
Trace; c0110018 <mtrr_add_page+2c8/370>
Trace; c011139e <apm_bios_call+6e/80>
Trace; c0111446 <apm_get_event+26/70>
Trace; c01116a0 <apm_cpu_idle+130/140>
Trace; c0111d4d <check_events+16d/180>
Trace; c0111e5d <apm_mainloop+7d/b0>
Trace; c01127ad <apm+29d/2b0>
Trace; c0105000 <_stext+0/0>
Trace; c0150018 <elf_core_dump+1e8/a06>
Trace; c0105000 <_stext+0/0>
Trace; c0105676 <kernel_thread+26/30>
Trace; c0112510 <apm+0/2b0>
Code; 00000000 Before first symbol
00000000 <_EIP>:
Code; 00000000 Before first symbol
0: c7 04 90 00 00 00 00 movl $0x0,(%eax,%edx,4)
Code; 00000007 Before first symbol
7: 8b 53 0c mov 0xc(%ebx),%edx
Code; 0000000a Before first symbol
a: 8b 87 34 02 00 00 mov 0x234(%edi),%eax
Code; 00000010 Before first symbol
10: 0f b3 10 btr %edx,(%eax)
Code; 00000013 Before first symbol
13: 8b 00 mov (%eax),%eax
<0>Kernel panic: Aieee, killing interrupt handler!
[root@localhost ~]#
--
GMX - Die Kommunikationsplattform im Internet.
http://www.gmx.net
On Thu, 25 Apr 2002 [email protected] wrote:
| It would be great if someone would set up an automated stress tester that
| will test each kernel before release, so that kernel developers and bug
| hunters can concentrate on the really important issues.
OSDL's STP (Scalable Test Platform) often runs thru a set of
tests on released and pre-released kernels.
However, like you say, some of those 2.5.x kernels just won't
compile without some external patches...and if the external
patches are available, we'll be able to handle that soon.
Full coverage will still be difficult...some drivers etc.
don't compile due to kernel API changes.
Some drivers break when loading, like IDE in your case,
and we don't have many IDE drives here.
I'm sure that we could come up with other coverage
problems^W challenges, but lots of them can be overcome.
We haven't reached the point of being a go/no-go decider
(that I know of).
--
~Randy
> Pentium2 266MHz,
> Intel 440LX Chipset, Gigabyte mainboard,
> Advansys SCSI OEM card (bundled with TEAC8x burner)
> nVidia riva128 graphic card 4MB RAM,
> 8GB IDE HDD,
> USB ACM modem, USB scanner, ...
Same trace here. I've got P-III/733, no SCSI or other unusual
peripherals. Works like a charm with all sorts of 2.4.x.
Can add EIP points to __ide_end_requiest
> Trace; c01b9f41 <ide_end_request+11/20>
> Trace; c01c1d83 <cdrom_pc_intr+d3/1e0>
> Trace; c01b7e5d <ide_intr+ed/1a0>
> Trace; c01c1cb0 <cdrom_pc_intr+0/1e0>
> Trace; c010856a <handle_IRQ_event+3a/70>
> Trace; c0108761 <do_IRQ+91/f0>
> Trace; c010728e <common_interrupt+22/28>
> Trace; c0110018 <mtrr_add_page+2c8/370>
> Trace; c011139e <apm_bios_call+6e/80>
> Trace; c0111446 <apm_get_event+26/70>
> Trace; c01116a0 <apm_cpu_idle+130/140>
My trace is the same interrupt, but it starts at some unrelated
thread, so I do not think it's APM. Probably IDE.
-- Pete