Dear Pavel Machek:
2.5.32 without CONFIG_PM , it work fine in my machine.
2.5.32 with CONFIG_PM, Oops in add_timer.
Here is Oops , it is hardcopy from CRT.
c0119cd5
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0060:[<c0119cd5>] Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010086
eax: cf902ec4 ebx: cf902ec4 ecx: 00000012 edx: 00000000
esi: cf902ec4 edi: cf902d60 ebp: 00003000 esp: cd647e64
Stack: 0000300c d1864121 cf902ec4 cff79c00
Call Trace: [<d1864121>] [<d18630c0>] [<c016eda3>] [<c106ee69>] [<c016ee86>]
Code: 89 5a 04 89 13 89 43 04 89 18 51 9d eb 14 51 9d 8b 44 24 04
>>EIP; c0119cd5 <add_timer+a5/cc> <=====
>>eax; cf902ec4 <_end+f66ca38/11574bd4>
>>ebx; cf902ec4 <_end+f66ca38/11574bd4>
>>esi; cf902ec4 <_end+f66ca38/11574bd4>
>>edi; cf902d60 <_end+f66c8d4/11574bd4>
>>esp; cd647e64 <_end+d3b19d8/11574bd4>
Trace; d1864121 <END_OF_CODE+213de/????>
Trace; d18630c0 <END_OF_CODE+2037d/????>
Trace; c016eda3 <pci_pm_resume_device+1b/20>
Trace; c106ee69 <_end+dd89dd/11574bd4>
Trace; c016ee86 <pci_pm_resume_bus+3a/4c>
Code; c0119cd5 <add_timer+a5/cc>
00000000 <_EIP>:
Code; c0119cd5 <add_timer+a5/cc> <=====
0: 89 5a 04 mov %ebx,0x4(%edx) <=====
Code; c0119cd8 <add_timer+a8/cc>
3: 89 13 mov %edx,(%ebx)
Code; c0119cda <add_timer+aa/cc>
5: 89 43 04 mov %eax,0x4(%ebx)
Code; c0119cdd <add_timer+ad/cc>
8: 89 18 mov %ebx,(%eax)
Code; c0119cdf <add_timer+af/cc>
a: 51 push %ecx
Code; c0119ce0 <add_timer+b0/cc>
b: 9d popf
Code; c0119ce1 <add_timer+b1/cc>
c: eb 14 jmp 22 <_EIP+0x22> c0119cf7 <add_timer+c7/cc>
Code; c0119ce3 <add_timer+b3/cc>
e: 51 push %ecx
Code; c0119ce4 <add_timer+b4/cc>
f: 9d popf
Code; c0119ce5 <add_timer+b5/cc>
10: 8b 44 24 04 mov 0x4(%esp,1),%eax
--
--- Hu Gang