Hi Con,
Well, something in the scheduler (I presume) has made 2.4.21-ck3 sink
like a rock on my production system:
ksymoops 2.4.8 on i586 2.4.21-ck3. Options used
-V (default)
-k /proc/ksyms (default)
-l /proc/modules (default)
-o /lib/modules/2.4.21-ck3/ (default)
-m /boot/System.map-2.4.21-ck3 (default)
Warning: You did not tell me where to find symbol information. I will
assume that the log matches the kernel and modules that are running
right now and I'll use the default options above for symbol resolution.
If the current kernel and/or modules do not match the log, you can get
more accurate output by telling me the kernel version and where to find
map, modules, ksyms etc. ksymoops -h explains the options.
Unable to handle kernel paging request at virtual address 363ec88c
c0113c9f
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[schedule+95/960] Not tainted
EFLAGS: 00010097
eax: 00000001 ebx: 010a5765 ecx: c94f0000 edx: c94f0000
esi: c02b1b00 edi: 363ec858 ebp: c94f1eec esp: c94f1ed0
ds: 0018 es: 0018 ss: 0018
Process dhcpd3 (pid: 23479, stackpage=c94f1000)
Stack: d09206a0 c94f1f3c d5e9fa2c c94f0000 010a5765 c94f1ef8 cddb09a0 00000100
c011ec12 c94f1ef8 c02bac2c d617bef8 010a5765 c94f0000 c011eba0 cddb09a0
00000000 00000008 00000000 00000009 c014300d 00000304 c94f0000 0000752a
Call Trace: [schedule_timeout+82/160] [process_timeout+0/32] [do_select+237/512] [sys_select+775/1216] [system_call+51/64]
Code: 8b 5f 34 89 f9 83 c1 2c ff 0b 8b 51 04 8b 47 2c 89 50 04 89
Using defaults from ksymoops -t elf32-i386 -a i386
>>ecx; c94f0000 <_end+9214d80/18551de0>
>>edx; c94f0000 <_end+9214d80/18551de0>
>>esi; c02b1b00 <runqueues+0/940>
>>ebp; c94f1eec <_end+9216c6c/18551de0>
>>esp; c94f1ed0 <_end+9216c50/18551de0>
Code; 00000000 Before first symbol
00000000 <_EIP>:
Code; 00000000 Before first symbol
0: 8b 5f 34 mov 0x34(%edi),%ebx
Code; 00000003 Before first symbol
3: 89 f9 mov %edi,%ecx
Code; 00000005 Before first symbol
5: 83 c1 2c add $0x2c,%ecx
Code; 00000008 Before first symbol
8: ff 0b decl (%ebx)
Code; 0000000a Before first symbol
a: 8b 51 04 mov 0x4(%ecx),%edx
Code; 0000000d Before first symbol
d: 8b 47 2c mov 0x2c(%edi),%eax
Code; 00000010 Before first symbol
10: 89 50 04 mov %edx,0x4(%eax)
Code; 00000013 Before first symbol
13: 89 00 mov %eax,(%eax)
1 warning issued. Results may not be reliable.
2.4.21-ck1 was fine under nearly the same circumstances. I rebuilt with
a newer -ck when I was configuring my new ADSL bridge to work with Linux
pppoe, but I doubt that's related.
Anyway, it drove my wireless card driver nuts too, probably due to some
busted interrupts. It kept printing a SW TICK STUCK? message. I'll
revert to vanilla for now. :(
-Josh
--
Using words to describe magic is like using a screwdriver to cut roast beef.
-- Tom Robbins
On Fri, 25 Jul 2003 15:18, Joshua Kwan wrote:
> 2.4.21-ck1 was fine under nearly the same circumstances. I rebuilt with
> a newer -ck when I was configuring my new ADSL bridge to work with Linux
> pppoe, but I doubt that's related.
>
> Anyway, it drove my wireless card driver nuts too, probably due to some
> busted interrupts. It kept printing a SW TICK STUCK? message. I'll
> revert to vanilla for now. :(
I suspect it's the variable Hz that is exported as 100 but runs as 1000 in
2.4-ck. I suggest you build without the variable Hz and tuning so it runs at
100Hz and try again.
Con
On Sat, Jul 26, 2003 at 01:06:45PM +1000, Con Kolivas wrote:
> I suspect it's the variable Hz that is exported as 100 but runs as 1000 in
> 2.4-ck. I suggest you build without the variable Hz and tuning so it runs at
> 100Hz and try again.
I only have patch-1000_O1_PE_LL_0306300059_2.4.21-ck3.bz2 applied, so
this is presumably not an issue.
-Josh
--
Using words to describe magic is like using a screwdriver to cut roast beef.
-- Tom Robbins