cpufreq oops with 2.4.20-pre7-ac3

From: Pierrick Hascoet (pierrick.hascoet@hydromel.net)
Date: Fri Sep 20 2002 - 09:24:58 EST


I caught an oops during the 2.4.20-pre7-ac3 boot sequence, this is related
to the last cpufreq update i guess, it's probably not an ac3 issue. The
cause it's a divide per 0 during cpufreq scaling.

The patch introduced here:
must be aware of the div per 0 possible case.

OTOH, I can reproduce it only when laptop boot with ac power on, so with a
866 cpu frequency, scaled by cpufreq.

Best regards,

Laptop: Compaq Evo N600c with ICH3-M chipset.

model name : Intel(R) Pentium(R) III Mobile CPU 866MHz
stepping : 1
cpu MHz : 664.460

batz:~# ksymoops --vmlinux=/usr/src/linux-2.4.20-pre7-ac3/vmlinux
--system-map=/usr/src/linux-2.4.20-pre7-ac3/System.map --no-ksyms
--no-lsmod < cpufreq.oops.txt
ksymoops 2.4.5 on i686 2.4.20-pre1-ac3. Options used
     -v /usr/src/linux-2.4.20-pre7-ac3/vmlinux (specified)
     -K (specified)
     -L (specified)
     -o /lib/modules/2.4.20-pre1-ac3/ (default)
     -m /usr/src/linux-2.4.20-pre7-ac3/System.map (specified)

No modules in ksyms, skipping objects
CPU: 0
EIP: 0010:[<c010b413>] Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax: 004bde9c ebx: 004bde9c ecx: 00000000 edx: 00000000
esi: 00000000 edi: c11bbfa8 ebp: 000021c5 esp: c11bbf20
ds: 0018 es: 0018 0018
Process swapper (pid: 1, stackpage=c11bb000)
Stack: c02227a0 c11bbfa8 00000001 c0223fb4 00000097 00000246 c11b4800 c11bbfa8
       c011edee c02227a0 00000001 c11bbfa8 00000000 c11bbfa8 000021c5 c012233c
       c027aad8 00000001 c11bbfa8 c11bbf00 00000001 c11bbfa8 00001001 00000000
Call Trace: [<c011edee>] [<c012233c>] [<c0111b65>] [<c0111ea8>] [<c0105023>]
Code: f7 f1 89 c3 0f af dd 89 d6 0f af f5 89 f0 31 d2 f7 f1 01 d8

>>EIP; c010b413 <time_cpufreq_notifier+167/204> <=====

>>EIP; c010b413 <time_cpufreq_notifier+167/204> <=====

>>eax; 004bde9c Before first symbol
>>ebx; 004bde9c Before first symbol
>>edi; c11bbfa8 <END_OF_CODE+f1d0f4/????>
>>ebp; 000021c5 Before first symbol
>>esp; c11bbf20 <END_OF_CODE+f1d06c/????>

Trace; c011edee <notifier_call_chain+1e/38>
Trace; c012233c <cpufreq_notify_transition+ec/104>
Trace; c0111b65 <speedstep_set_state+15d/168>
Trace; c0111ea8 <speedstep_detect_speeds+80/b8>
Trace; c0105023 <init+7/114>
Trace; c01054e8 <kernel_thread+28/38>

Code; c010b413 <time_cpufreq_notifier+167/204>
00000000 <_EIP>:
Code; c010b413 <time_cpufreq_notifier+167/204> <=====
   0: f7 f1 div %ecx <=====
Code; c010b415 <time_cpufreq_notifier+169/204>
   2: 89 c3 mov %eax,%ebx
Code; c010b417 <time_cpufreq_notifier+16b/204>
   4: 0f af dd imul %ebp,%ebx
Code; c010b41a <time_cpufreq_notifier+16e/204>
   7: 89 d6 mov %edx,%esi
Code; c010b41c <time_cpufreq_notifier+170/204>
   9: 0f af f5 imul %ebp,%esi
Code; c010b41f <time_cpufreq_notifier+173/204>
   c: 89 f0 mov %esi,%eax
Code; c010b421 <time_cpufreq_notifier+175/204>
   e: 31 d2 xor %edx,%edx
Code; c010b423 <time_cpufreq_notifier+177/204>
  10: f7 f1 div %ecx
Code; c010b425 <time_cpufreq_notifier+179/204>
  12: 01 d8 add %ebx,%eax

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

This archive was generated by hypermail 2b29 : Mon Sep 23 2002 - 22:00:30 EST