[prev in list] [next in list] [prev in thread] [next in thread] 

List:       cobalt-developers
Subject:    Re: [cobalt-developers] RaQ3 kernel oops
From:       Tim Hockin <thockin () sun ! com>
Date:       2001-09-20 19:03:34
[Download RAW message or body]

Matthias Pigulla wrote:

> Now we're stuck with the following messages when booting. I can't figure out
> what this problem might be related to; we also checked it by installing the
> same HDD in another RaQ3 with some minor differences in the specs (no SCSI,
> just one eth interface), getting the same messages.

If the problem follows the disk, it almost HAS to be the disk.  What is
interesting is that the first stage kernel is the one that panicked.  Your
disk must have done something to make it very unhappy.  What happens when
you put a new disk in there?  Do you have a PCI card or anything else
different in the system?

> VFS: Mounted root (ext2 filesystem) readonly.
> Freeing unused kernel memory: 64k freed
> Unable to handle kernel NULL pointer dereference at virtual address 00000008
> current->tss.cr3 = 00101000, %cr3 = 00101000
> *pde = 00000000
> Oops: 0000
> CPU:    0
> EIP:    0010:[<c012b4dc>]
> EFLAGS: 00010282
> eax: 00000000   ebx: fffffff4   ecx: c0290140   edx: c0290090
> esi: c02900e0   edi: c0097dc0   ebp: c0003f44   esp: c0003f10
> ds: 0018   es: 0018   ss: 0018
> Process swapper (pid: 1, process nr: 1, stackpage=c0003000)
> Stack: c0003f44 00000000 c0237005 0000000b c0097e0c c012b6bc c0290060
> c0003f44
>        0000000b c0236020 ffffffe9 00000003 17ffffb4 c0237001 00000003
> 00006226
>        c012b83a c0237000 c0290060 00000001 c0236020 ffffffe9 c0106000
> 17ffffb4
> Call Trace: [<c012b6bc>] [<c012b83a>] [<c0106000>] [<c01243f8>] [<c0106000>]
> [<c
> 012463a>] [<c0109a04>]
>        [<c01db805>] [<c0106000>] [<c01060ab>] [<c010856b>]
> Code: 8b 40 08 ff d0 89 c3 83 c4 08 85 db 74 0e 56 e8 48 47 00 00
 
as best I can tell:

EIP is in cached_lookup() which does deal with filestystem lookups.  

> Any hints on how I could proceed? How exactly does the RaQ boot procedure
> look like? I noticed that there are two "stages" in the boot process, and it
> seems to me as if the first stage boots a kernel from ROM? As we're 

yes - a flash-loaded kernel reads the disk and loads the real kernel.

-- 
Tim Hockin
Systems Software Engineer
Sun Microsystems, Cobalt Server Appliances
thockin@sun.com

_______________________________________________
cobalt-developers mailing list
cobalt-developers@list.cobalt.com
http://list.cobalt.com/mailman/listinfo/cobalt-developers

[prev in list] [next in list] [prev in thread] [next in thread]