2.4.1-pre7 SMP oops

Paul Flinders (P.Flinders@ftel.co.uk)
Wed, 24 Jan 2001 12:49:16 GMT


My Dual 866Mhz PIII died whilst compiling gcc with

ksymoops -m /boot/System.map-2.4.1-pre7 < ~/oops
ksymoops 0.7c on i686 2.4.1-pre7. Options used
-V (default)
-k /proc/ksyms (default)
-l /proc/modules (default)
-o /lib/modules/2.4.1-pre7/ (default)
-m /boot/System.map-2.4.1-pre7 (specified)

Warning (compare_maps): ksyms_base symbol acpi_clear_event_R__ver_acpi_clear_event not found in System.map. Ignoring ksyms_base entry
[rather a lot of similar warnings]
Unable to handle kernel NULL pointer derefernce at virtual address 00000000
c011f9e3
*pde = 00000000
Oops: 0002
EIP: 0010 : [<c011f9e3>]
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010086
eax: 00000000 ebx: c0274000 ecx: 00000000 edx: 00000000
esi: 00000000 edi: 00000000 ebp: 00000000 esp: c0275f70
ds: 0018 es: 0018 ss: 0018
Process swapper (pid: 0, stackpage=c0275000)
Stack: 00000000 00000000 00000000 00000000 c01137a6 00000000 c01072a0 c0274000
c01072a0 c0274000 c0212e3c c01072a0 c0274000 c0274000 c0274000 ?
Call trace: <c01137a6> <c01072a0> <c01072a0> <c01072a0> <c01072a0> <c0100018>
<c01072cc> <c0107332> <c0105000> <c01001cf>
Code: 6c 24 14 89 ef 56 83 f7 01 57 55 53 e8 f4 fe ff 83 c4 10

>>EIP; c011f9e3 <update_process_times+f/94> <=====
Trace; c01137a6 <smp_apic_timer_interrupt+e6/f4>
Trace; c01072a0 <default_idle+0/34>
Trace; c01072a0 <default_idle+0/34>
Trace; c01072a0 <default_idle+0/34>
Trace; c01072a0 <default_idle+0/34>
Trace; c0100018 <startup_32+18/cb>
Trace; c01072cc <default_idle+2c/34>
Trace; c0107332 <cpu_idle+3e/54>
Trace; c0105000 <empty_bad_page+0/1000>
Trace; c01001cf <L6+0/2>
Code; c011f9e3 <update_process_times+f/94>
00000000 <_EIP>:
Code; c011f9e3 <update_process_times+f/94> <=====
0: 6c insb (%dx),%es:(%edi) <=====
Code; c011f9e4 <update_process_times+10/94>
1: 24 14 and $0x14,%al
Code; c011f9e6 <update_process_times+12/94>
3: 89 ef mov %ebp,%edi
Code; c011f9e8 <update_process_times+14/94>
5: 56 push %esi
Code; c011f9e9 <update_process_times+15/94>
6: 83 f7 01 xor $0x1,%edi
Code; c011f9ec <update_process_times+18/94>
9: 57 push %edi
Code; c011f9ed <update_process_times+19/94>
a: 55 push %ebp
Code; c011f9ee <update_process_times+1a/94>
b: 53 push %ebx
Code; c011f9ef <update_process_times+1b/94>
c: e8 f4 fe ff 83 call 83ffff05 <_EIP+0x83ffff05> 4411f8e8 Before first symbol
Code; c011f9f4 <update_process_times+20/94>
11: c4 10 les (%eax),%edx

45 warnings issued. Results may not be reliable.

The code looks a bit odd - it doesn't quite align with the instructions in
update_process_times. The raw OOPS included "CPU: -117" so I guess that
something else randomly corrupted either current or current->processor and
update_process_times/update_one_process fell over as a result.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/