Re: [squid-users] Unable to handle kernel paging request at virtual address

On Thu, 31 Jan 2002, Jack wrote:

> Hello All,
>
> Suddenly my squid turns to defunct.
>
> When i check log/messages i got this error message repeatedly with different
> virtual address
> I am using resierfs for cache_dir.
>
> Jan 31 18:10:20 sample kernel: printing eip:
> Jan 31 18:10:20 sample kernel: c010ed74
> Jan 31 18:10:21 sample kernel: *pde = 00000000
> Jan 31 18:10:21 sample kernel: Oops: 0000
> Jan 31 18:10:21 sample kernel: CPU: 0
> Jan 31 18:10:21 sample kernel: EIP: 0010:[__wake_up+36/176] Not
> tainted
> Jan 31 18:10:21 sample kernel: EIP: 0010:[<c010ed74>] Not tainted
> Jan 31 18:10:21 sample kernel: EFLAGS: 00010086
> Jan 31 18:10:21 sample kernel: eax: c4ac0efc ebx: ba25a47b ecx: 00000001
> edx: 00000003
> Jan 31 18:10:21 sample kernel: esi: c4ac0e60 edi: 00000001 ebp: ce9f7e30
> esp: ce9f7e18
> Jan 31 18:10:21 sample kernel: ds: 0018 es: 0018 ss: 0018
> Jan 31 18:10:21 sample kernel: Process squid (pid: 985, stackpage=ce9f7000)
> Jan 31 18:10:21 sample kernel: Stack: 00000286 00000003 c4ac0efc ce9f7ea4
> c4ac0e60 c4ac0e60 da3f3800 c013df90
> Jan 31 18:10:21 sample kernel: cbb7b17c 00000008 542d5080 00000000
> ce9f7ea4 dff9a600 00012d5a da3f3800
> Jan 31 18:10:21 sample kernel: c013e1b2 da3f3800 00012d5a dff9a600
> 00000000 ce9f7ea4 00000008 00000104
> Jan 31 18:10:21 sample kernel: Call Trace: [get_new_inode+240/352]
> [iget4+194/208] [<e08a15e1>] [<e089de96>] [d_alloc+25/336]
> Jan 31 18:10:21 sample kernel: Call Trace: [<c013df90>] [<c013e1b2>]
> [<e08a15e1>] [<e089de96>] [<c013c699>]
> Jan 31 18:10:21 sample kernel: [lookup_hash+106/144] [sys_unlink+111/272]
> [system_call+51/56]
> Jan 31 18:10:21 sample kernel: [<c01355ca>] [<c013653f>] [<c0106ceb>]
> Jan 31 18:10:21 sample kernel: Code: 8b 13 0f 18 02 eb 6b 90 8d 74 26 00 8b
> 4b fc 8b 01 85 45 ec
> Jan 31 18:10:22 sample kernel: <1>Unable to handle kernel paging request at
> virtual address e10615f4
>
> What will be the problem?
> if reboot machine and restart squid gets 300 req i get same problem.
> I am using rehat linux-7.1 kernel 2.4.13 squid-2.5.PRE2

I don't know what the problem is, but it certainly is not a problem with
Squid. My first guesses would be a hardware problem or that you need to
update reiserfs or the linux kernel.

I had some messages like that when I played with XFS on linux. They went
away when I followed the instructions and compiled the kernel with
the older "EGCS" instead of the current GCC.

Duane W.
Received on Fri Feb 01 2002 - 09:18:47 MST