* 2.6.9-rc3 does not like diablo news reader daemon
@ 2004-10-07 7:19 Marian Eichholz
2004-10-07 8:03 ` Jeff Garzik
2004-10-07 8:23 ` Andrew Morton
0 siblings, 2 replies; 3+ messages in thread
From: Marian Eichholz @ 2004-10-07 7:19 UTC (permalink / raw)
To: linux-kernel
Hi,
the kernel 2.6.9-rc3 tends to crash when executing our NNTP
reader daemon (dreaderd from diablo).
2.6.3-rc1 runs on the sibling machine like a charm.
Please apologise, if this posting is considered off topic.
You may CC me on this :)
Please see the kernel-log below for evidence. Hope it
helps anything.
The hardware is a usual Intel STL2, 2 Intel-P3 and so on.
Mit freundlichen Gruessen / Yours sincerely
Marian Eichholz
-------------------------------------------------
[...]
Oct 6 22:48:39 newsread5 kernel: c0132eb1
Oct 6 22:48:39 newsread5 kernel: *pde = 00000000
Oct 6 22:48:39 newsread5 kernel: Oops: 0002 [#1]
Oct 6 22:48:39 newsread5 kernel: SMP
Oct 6 22:48:39 newsread5 kernel: CPU: 0
Oct 6 22:48:39 newsread5 kernel: EIP: 0060:[<c0132eb1>] Not tainted VLI
Oct 6 22:48:39 newsread5 kernel: EFLAGS: 00010246 (2.6.9-rc3)
Oct 6 22:48:39 newsread5 kernel: EIP is at generic_file_aio_write_nolock+0x1d1/0x500
Oct 6 22:48:39 newsread5 kernel: eax: 00014c38 ebx: 00000000 ecx: 00000000 edx: 00000000
Oct 6 22:48:39 newsread5 kernel: esi: de8fe000 edi: 00014c38 ebp: 00000000 esp: de8ffe38
Oct 6 22:48:39 newsread5 kernel: ds: 007b es: 007b ss: 0068
Oct 6 22:48:39 newsread5 kernel: Process dreaderd (pid: 166, threadinfo=de8fe000 task=de86d000)
Oct 6 22:48:39 newsread5 kernel: Stack: 00014c38 00000040 00000000 de8ffe48 00000000 ca258200 de8ffee0 c02a8f05
Oct 6 22:48:39 newsread5 kernel: de8ffee0 00000001 de8ffe8c ffffffff df37f3ec 00000000 00000000 df37f3ec
Oct 6 22:48:39 newsread5 kernel: 00000048 df37f490 de961560 00000048 00014c38 00000000 de8ffeb4 de961560
Oct 6 22:48:39 newsread5 kernel: Call Trace:
Oct 6 22:48:39 newsread5 kernel: [<c02a8f05>] sock_aio_read+0xf5/0x110
Oct 6 22:48:39 newsread5 kernel: [<c013328a>] generic_file_write_nolock+0xaa/0xd0
Oct 6 22:48:39 newsread5 kernel: [<c0118460>] autoremove_wake_function+0x0/0x60
Oct 6 22:48:39 newsread5 kernel: [<c0133406>] generic_file_write+0x56/0xd0
Oct 6 22:48:39 newsread5 kernel: [<c01333b0>] generic_file_write+0x0/0xd0
Oct 6 22:48:39 newsread5 kernel: [<c014f738>] vfs_write+0xb8/0x130
Oct 6 22:48:39 newsread5 kernel: [<c014f881>] sys_write+0x51/0x80
Oct 6 22:48:39 newsread5 kernel: [<c010422f>] syscall_call+0x7/0xb
Oct 6 22:48:39 newsread5 kernel: Code: 00 00 39 c7 0f 83 d2 01 00 00 8b 44 24 28 8b 4c 24 4c 29 c8 39 44 24 48 76 10 0f 83 58 01 00 00 8b 5c 24 48 31 f6 8b 48 1c 8b 50 <18> 01 fb 11 ee 39 ce 0f 83 24 01 00 00 31 c9 00 00 8b 74 24 30
Oct 6 22:48:39 newsread5 kernel: <0>Fatal exception: panic in 5 seconds
Oct 6 23:01:09 newsread5 kernel: klogd 1.4-0, log source = /proc/kmsg started.
Oct 6 23:01:09 newsread5 kernel: Cannot open map file: /boot/System.map.
Oct 6 23:01:09 newsread5 kernel: No module symbols loaded - kernel modules not enabled.
Oct 6 23:01:09 newsread5 kernel: process `named' is using obsolete setsockopt SO_BSDCOMPAT
Oct 6 23:01:17 newsread5 kernel: eth0: no IPv6 routers present
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: 2.6.9-rc3 does not like diablo news reader daemon
2004-10-07 7:19 2.6.9-rc3 does not like diablo news reader daemon Marian Eichholz
@ 2004-10-07 8:03 ` Jeff Garzik
2004-10-07 8:23 ` Andrew Morton
1 sibling, 0 replies; 3+ messages in thread
From: Jeff Garzik @ 2004-10-07 8:03 UTC (permalink / raw)
To: marian.eichholz; +Cc: linux-kernel
Marian Eichholz wrote:
> Hi,
>
> the kernel 2.6.9-rc3 tends to crash when executing our NNTP
> reader daemon (dreaderd from diablo).
>
> 2.6.3-rc1 runs on the sibling machine like a charm.
> Oct 6 22:48:39 newsread5 kernel: EIP is at generic_file_aio_write_nolock+0x1d1/0x500
> Oct 6 22:48:39 newsread5 kernel: eax: 00014c38 ebx: 00000000 ecx: 00000000 edx: 00000000
> Oct 6 22:48:39 newsread5 kernel: esi: de8fe000 edi: 00014c38 ebp: 00000000 esp: de8ffe38
> Oct 6 22:48:39 newsread5 kernel: ds: 007b es: 007b ss: 0068
> Oct 6 22:48:39 newsread5 kernel: Process dreaderd (pid: 166, threadinfo=de8fe000 task=de86d000)
> Oct 6 22:48:39 newsread5 kernel: Stack: 00014c38 00000040 00000000 de8ffe48 00000000 ca258200 de8ffee0 c02a8f05
> Oct 6 22:48:39 newsread5 kernel: de8ffee0 00000001 de8ffe8c ffffffff df37f3ec 00000000 00000000 df37f3ec
> Oct 6 22:48:39 newsread5 kernel: 00000048 df37f490 de961560 00000048 00014c38 00000000 de8ffeb4 de961560
> Oct 6 22:48:39 newsread5 kernel: Call Trace:
> Oct 6 22:48:39 newsread5 kernel: [<c02a8f05>] sock_aio_read+0xf5/0x110
hmmmm, interesting. Looks like AIO is ailing.
As an ex-Usenet news admin, I really like DIABLO... maybe I'll pull the
latest copy and try to debug this.
Jeff
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: 2.6.9-rc3 does not like diablo news reader daemon
2004-10-07 7:19 2.6.9-rc3 does not like diablo news reader daemon Marian Eichholz
2004-10-07 8:03 ` Jeff Garzik
@ 2004-10-07 8:23 ` Andrew Morton
1 sibling, 0 replies; 3+ messages in thread
From: Andrew Morton @ 2004-10-07 8:23 UTC (permalink / raw)
To: marian.eichholz; +Cc: linux-kernel
Marian Eichholz <marian.eichholz@freenet-ag.de> wrote:
>
> the kernel 2.6.9-rc3 tends to crash when executing our NNTP
> reader daemon (dreaderd from diablo).
>
> 2.6.3-rc1 runs on the sibling machine like a charm.
Impressive.
> [...]
> Oct 6 22:48:39 newsread5 kernel: c0132eb1
> Oct 6 22:48:39 newsread5 kernel: *pde = 00000000
> Oct 6 22:48:39 newsread5 kernel: Oops: 0002 [#1]
> Oct 6 22:48:39 newsread5 kernel: SMP
> Oct 6 22:48:39 newsread5 kernel: CPU: 0
> Oct 6 22:48:39 newsread5 kernel: EIP: 0060:[<c0132eb1>] Not tainted VLI
> Oct 6 22:48:39 newsread5 kernel: EFLAGS: 00010246 (2.6.9-rc3)
> Oct 6 22:48:39 newsread5 kernel: EIP is at generic_file_aio_write_nolock+0x1d1/0x500
> Oct 6 22:48:39 newsread5 kernel: eax: 00014c38 ebx: 00000000 ecx: 00000000 edx: 00000000
> Oct 6 22:48:39 newsread5 kernel: esi: de8fe000 edi: 00014c38 ebp: 00000000 esp: de8ffe38
> Oct 6 22:48:39 newsread5 kernel: ds: 007b es: 007b ss: 0068
> Oct 6 22:48:39 newsread5 kernel: Process dreaderd (pid: 166, threadinfo=de8fe000 task=de86d000)
> Oct 6 22:48:39 newsread5 kernel: Stack: 00014c38 00000040 00000000 de8ffe48 00000000 ca258200 de8ffee0 c02a8f05
> Oct 6 22:48:39 newsread5 kernel: de8ffee0 00000001 de8ffe8c ffffffff df37f3ec 00000000 00000000 df37f3ec
> Oct 6 22:48:39 newsread5 kernel: 00000048 df37f490 de961560 00000048 00014c38 00000000 de8ffeb4 de961560
> Oct 6 22:48:39 newsread5 kernel: Call Trace:
> Oct 6 22:48:39 newsread5 kernel: [<c02a8f05>] sock_aio_read+0xf5/0x110
> Oct 6 22:48:39 newsread5 kernel: [<c013328a>] generic_file_write_nolock+0xaa/0xd0
> Oct 6 22:48:39 newsread5 kernel: [<c0118460>] autoremove_wake_function+0x0/0x60
> Oct 6 22:48:39 newsread5 kernel: [<c0133406>] generic_file_write+0x56/0xd0
> Oct 6 22:48:39 newsread5 kernel: [<c01333b0>] generic_file_write+0x0/0xd0
> Oct 6 22:48:39 newsread5 kernel: [<c014f738>] vfs_write+0xb8/0x130
> Oct 6 22:48:39 newsread5 kernel: [<c014f881>] sys_write+0x51/0x80
> Oct 6 22:48:39 newsread5 kernel: [<c010422f>] syscall_call+0x7/0xb
It got a null-pointer deref right in the heart of the write(2) syscall.
Everybody is using that all the time. So I'd say that either there's
something unusual about your setup or your hardware is failing.
- Are the oopses always the same? Please send some more.
- What types of filesytem are in use?
- Are you using any unusual mount options?
- Are you using any uncommon hardware?
- Does the application do anything unusual such as O_DIRECT I/O?
- Are any other machines running the same application and kernel? If so,
are they failing? If not, are you able to use a different machine with
2.6.9-rc3? That'll help us work out whether it's a hardware or software
failure.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2004-10-07 8:25 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-10-07 7:19 2.6.9-rc3 does not like diablo news reader daemon Marian Eichholz
2004-10-07 8:03 ` Jeff Garzik
2004-10-07 8:23 ` Andrew Morton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox