On Mon, Nov 27, 2000 at 09:10:05AM +0100, Matthias Andree wrote:
> My previous mail was a tad too early. After I moved the bus back to the
> Tekram DC-390U (sym53c8xx driver), I caught an oops again, and
> rescan-scsi-bus.sh caught a SIGSEGV. There seems to be a kernel bug somewhere
> in that SCSI handling stuff:
>
> Unable to handle kernel NULL pointer dereference at virtual address 00000052
> current->tss.cr3 = 07106000, %%cr3 = 07106000
> *pde = 00000000
> Oops: 0002
> CPU: 0
> EIP: 0010:[sr_finish+112/388]
> EFLAGS: 00010206
> eax: 00000000 ebx: 00000054 ecx: 00000005 edx: c7f9cc00
> esi: c629bd50 edi: c7f9cc40 ebp: 00000001 esp: c629bd3c
> ds: 0018 es: 0018 ss: 0018
> Process rescan-scsi-bus (pid: 2846, process nr: 23, stackpage=c629b000)
> Stack: 00000000 c7f88340 c02a9960 c629bd4c 00307273 c01fedaf c7f88700 c7555de0
> c629becc c01fedf3 00000000 00000000 c7f88340 00000000 c6e132c0 c02cb0e0
> c6e132c8 c0132793 c6e132c0 c629bda4 c629bda4 c0001a00 00020000 c63532c0
> Call Trace: [scan_scsis+491/1076] [scan_scsis+559/1076] [get_new_inode+147/312]
>[iget4+121/132] [vsprintf+720/764] [wake_up_process+64/76] [__wake_up+59/68]
> Code: 80 48 52 20 a1 8c 99 2a c0 80 4c 18 16 08 a1 8c 99 2a c0 80
> Using defaults from ksymoops -t elf32-i386 -a i386
This time, I do not suspect the low-level driver. It mey very well be, that
there's a subtile bug WRT add/remove-single-device housekeeping in the SCSI
layer.
You told that you did parallel rescan-scsi-bus.sh calls. That may have
confused the kernel. Did the above oops happen without such abuse?
Regards,
--
Kurt Garloff <[EMAIL PROTECTED]> [Eindhoven, NL]
Physics: Plasma simulations <[EMAIL PROTECTED]> [TU Eindhoven, NL]
Linux: SCSI, Security <[EMAIL PROTECTED]> [SuSE Nuernberg, FRG]
(See mail header or public key servers for PGP2 and GPG public keys.)
PGP signature
