1.2.13 oops
1.3.71 boots + hangs after 2nd ramdisk in redhat set
2.0.7 oops (attached at the end)
i have read all the documents with some relevance
but found no suggestions. i found three articles
in the linux.setup newsgroup describing the exact same
problem under 1.2.13 but with no responses so far.
the machine is a gateway 2000 p5-60 with a mitsumi cdrom,
a 3com el3 network board (but the problems occur
with and without those boards in), and an ati-mach32 video board.
i have also tried to boot with slackware 3.0 and redhat 3.0.3
boot disks with no luck (same type of immediate crashes).
finally, please note that i am not an expert.
also i am not in the list, so if you have any
suggestions please remember to mail them to
me as well.
thanks in advance,
-- carlos
here is the oops: i am not copying the stack trace, but
if you want i can transcribe it.
the oops log:
Loading .......
pcbios_init: ... structure at 0x...
pcbios_init: ... entry at 0x...
pcbios_init: ... structure at 0x...
general protection: 5024
CPU: 0
EIP: 0010:[<0000004c>]
EFLAGS: 00010012
eax: 000f3070 ebx: 00000010 ecx: 00010001 edx: 00003000
esi: 00100000 edi: 00239604 ebp: 017fffff esp: 001b7724
ds: 0018 es: 0018 fs: 0018 gs: 0018 ss: 0018
Process swapper: (pid: 0, process nr: 0, stackpage=001b57a4)
Stack: ....
Call trace: [<0010d470>] [<0010d8af>] [<001a1976>] [<00109120>]
Code: cb a1 00 f0 39 e7 00 f0 59 f8 00 f0 2e e8 00 f0 d2 ef 00 f0
kfree of non-kmalloced memory: 001b77ec, next=00000000, order=0
kfree of non-kmalloced memory: 001b77dc, next=00000000, order=0
kfree of non-kmalloced memory: 001b7cf0, next=00000000, order=0
idle task may not sleep
idle task may not sleep
idle task may not sleep
idle task may not sleep
idle task may not sleep
then after several minutes the screen goes dark and
there is no way to return it from the deep
darkness.
[01:17:12](3)chilli:linux# ksymoops System.map < oops-log
Using `System.map' to map addresses to symbols.
Code: lret
Code: movl 0xe739f000,%eax
Code: addb %dh,%al
Code: popl %ecx
Code: clc
Code: addb %dh,%al
Code: call efd2f012 <gcc2_compiled.+efd2f012>
Code: addb %dh,%al
[01:17:16](3)chilli:linux#
0000004c seem pretty low according to the symbols listed by nm:
[01:20:02](3)chilli:linux# nm arch/i386/boot/compressed/vmlinux| sort | head
00001000 T startup_32
00001076 t move_routine_start
0000108c t move_routine_end
00001090 t gcc2_compiled.
00001090 t huft_build
00001720 t huft_free
00001750 t inflate_codes
00001b70 t inflate_stored
00001d20 t inflate_fixed
00001ea0 t inflate_dynamic
[01:20:09](3)chilli:linux#
Call trace (from the System.map):
0010d470 -
0010d460 T check_pcibios
0010d53c T pcibios_present
0010d8af -
0010d7ec T pcibios_init
0010d8bc t get_long
001a1976 -
001a1964 T pci_init
001a19e4 T csum_partial
00109120 -
00109074 T start_kernel
001091f8 t printf
this seems to point to a problem related to
the system detecting two pci bios structures?
perhaps this is a known bug?
or am i missing something?
in 1.2.13 a warning is issued before the crash,
but that warning was taken off from bios32.c later
on (although the two bios signatures are still detected).