Oleg (and all),
Made it 8 days before the 2.2.16 kernel crashed. This is not the NULL
pointer exception this time. I will recompile with Oleg's patch (below) and
hope this gives some more meaningful diagnostic hint. Here's the Oops:
Unable to handle kernel paging request at virtual address 66a93fb4
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<c0178cd0>]
EFLAGS: 00010002
eax: c7e80000 ebx: 66a93f94 ecx: 00000001 edx: 66a93f94
esi: 00000001 edi: 00000000 ebp: 0000cacb esp: c0217efc
ds: 0018 es: 0018 ss: 0018
Stack: 0000000f 0c513b6c 66a93f94 00000292 00000000 c0179109 c020c510 c020c4ec
c01790d8 c0217f44 c0217f48 c011126d 00000000 00000001 c0254594 00000000
00000001 c0216000 00000000 c0217f60 c01179f5 00000000 c0216000 06513b57
Call Trace: [<c0179109>] [<c01790d8>] [<c011126d>] [<c01179f5>] [<c010a311>] [<c0109fe0>] [<c01078a9>]
[<c0106000>] [<c01078cc>] [<c01090fc>] [<c0106000>] [<c010607b>] [<c0106000>] [<c0100175>]
Code: 8b 53 20 89 54 24 10 83 7b 30 00 0f 85 fb 00 00 00 8a 43 77
>>EIP: c0178cd0 <ip_cmsg_recv+79e0/cc3c>
Trace: c0179109 <ip_cmsg_recv+7e19/cc3c>
Trace: c01790d8 <ip_cmsg_recv+7de8/cc3c>
Trace: c011126d <sleep_on_timeout+821/1314>
Trace: c01179f5 <get_fast_time+6bd/78c>
Trace: c010a311 <enable_irq+8d/128>
Trace: c0109fe0 <dump_thread+224c/2488>
Trace: c01078a9 <enable_hlt+69/154>
Trace: c0106000 <get_options+0/a5c>
Code: c0178cd0 <ip_cmsg_recv+79e0/cc3c> 00000000 <_EIP>: <===
Code: c0178cd0 <ip_cmsg_recv+79e0/cc3c> 0: 8b 53 20 mov 0x20(%ebx),%edx <===
Code: c0178cd3 <ip_cmsg_recv+79e3/cc3c> 3: 89 54 24 10 mov %edx,0x10(%esp,1)
Code: c0178cd7 <ip_cmsg_recv+79e7/cc3c> 7: 83 7b 30 00 cmpl $0x0,0x30(%ebx)
Code: c0178cdb <ip_cmsg_recv+79eb/cc3c> b: 0f 85 fb 00 00 00 jne c0178ddc <ip_cmsg_recv+7aec/cc3c>
Code: c0178ce1 <ip_cmsg_recv+79f1/cc3c> 11: 8a 43 77 mov 0x77(%ebx),%al
Aiee, killing interrupt handler
Kernel panic: Attempted to kill the idle task!
In swapper task - not syncing
On Tue, Jun 13, 2000 at 01:22:09PM +0300, Oleg Drokin wrote:
>
> After next crash use the only patch from below, it doe not remove oops,
> but it cause it just when problem first occurs and easiers debug.
> May be the oops will not be fatal, then.
> If you receive anything after that patch - mail it to me.
>
> --- net/ipv4/tcp_ipv4.c.orig Tue Jun 13 11:24:25 2000
> +++ net/ipv4/tcp_ipv4.c Tue Jun 13 11:44:09 2000
> @@ -340,6 +340,10 @@
>
> static void tcp_v4_hash(struct sock *sk)
> {
> + if (sk->pprev != NULL) {
> + printk("sk->pprev = 0x%x in tcp_v4_hash, forcing oops\n",sk->pprev);
> + *((int*) 0)=0;
> + }
> if (sk->state != TCP_CLOSE) {
> SOCKHASH_LOCK();
> __tcp_v4_hash(sk);
>
> Bye,
> Oleg
> --
> System Administrator
> Tank Hill Ukraine
\/\/ I-I I T
Blauvelt
whit@transpect.com
-
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/
This archive was generated by hypermail 2b29 : Fri Jun 23 2000 - 21:00:22 EST