Re: Oops
Peter Enderborg (pme@ufh.se)
Sat, 24 May 1997 11:31:23 +0200
Zlatko Calusic wrote:
>
> Peter Enderborg <pme@ufh.se> writes:
>
> > Got an oops. (0)
> > EIP: 0c12a0c
> >
> >
> > 0c120a0c somewere in __wait_on_inode
> > 0c120f42 sync_inodes
> > 0c121eaa fsync_dev
> > 0c121ec7 sys_sync
> > 0c10942a system_call
> >
> > This is on a Pentium with ncr-810, ext2 filesystems and kernel 2.1.40
> > compiled with no SMP. (With the same config but with SMP i got
> > an hang without Oops at the same place.) It's hapend when
> > i boot the system and the system is checking the filesystems.
> > Any ideas ? I have not been able to use any kernels since 2.1.35.
> > I have noted that it have been an upgrade in the NCR driver
> > from 1.18d to 2.1b, is that the problem ?
>
> Looks like I got similar "results":
>
> Unable to handle kernel NULL pointer dereference at virtual address 00000004
> current->tss.cr3 = 01ee6000, Lr3 = 01ee6000
> *pde = 00000000
> Oops: 0000
> CPU: 0
> EIP: 0010:[<c012237c>]
> EFLAGS: 00010013
> eax: 00000000 ebx: c1eeff74 ecx: c1344874 edx: 00000000
> esi: 00000246 edi: c1eeff74 ebp: 00000001 esp: c1eeff68
> ds: 0018 es: 0018 ss: 0018
> Process update (pid: 9, process nr: 7, stackpage=c1eef000)
> Stack: c1344820 c1b72240 00000000 c1eee000 c1344874 c01228b2 c1344820 c1eee000
> 00000001 00000000 bffffbf4 c01255f8 00000000 00000000 c1eee000 00000001
> 00000000 00000000 c0125709 c1eee000 00000100 bffffe14 c0109426 00000001
> Call Trace: [<c01228b2>] [<c01255f8>] [<c0125709>] [<c0109426>]
> Code: 8b 42 04 39 d8 74 05 89 c2 eb f5 90 89 4a 04 56 9d 89 e0 25
>
> Using `/boot/System.map' to map addresses to symbols.
>
> >>EIP: c012237c <__wait_on_inode+64/8c>
> Trace: c01228b2 <sync_inodes+2e/5c>
> Trace: c01255f8 <sync_old_buffers+14/ec>
> Trace: c0125709 <sys_bdflush+39/a8>
> Trace: c0109426 <system_call+3a/40>
>
> Code: c012237c <__wait_on_inode+64/8c>
>
> non-SMP 2.1.40, Pentium 133, 32 MB, EIDE, ext2...
>
> --
> Posted by Zlatko Calusic E-mail: <Zlatko.Calusic@CARNet.hr>
> ---------------------------------------------------------------------
> "640K ought to be enough for anybody." Bill Gates '81
My solution is to upgrade e2fsbins. I was using e2fsck 1.04 and
upgrade to 1.06 and then it works. (But it should not be that easy
to crash the kernel, with an old fsck.)
But no solution on my diald problem.
--
foo!