Re: [crash/panic] Linux-2.6.0-test9

[Gary Wolfe <gpwolfe@cableone.net>]

Andrew Morton wrote:

>Gary Wolfe <gpwolfe@cableone.net> wrote:
>  
>
>>Greetings,
>>
>>I have:
>>
>>Asus P4C800 Deluxe w/2.4GHz P4 (no HT and not 800Mhz bus)
>>
>>Tried test8 and, now, test9 and both exhibit same problem.
>>
>>The issue seems to be related to the PnPBIOS support under the Plug and 
>>Play Kconfig category.  When enabled I get a crash of the form:
>>
>>Linux Plug and Play Support v0.97 (c) Adam Belay
>>PnPBIOS: Scanning system for PnP BIOS support...
>>PnPBIOS: Found PnP BIOS installation structure at 0xc00f5350
>>PnPBIOS: PnP BIOS version 1.0, entry 0xf0000:0x5f3a, dseg 0xf0000
>>general protection fault: 0000 [#1]
>>CPU: 0
>>EIP: 0098:[<00002b60>] Not tainted
>>EFLAGS: 00010083
>>EIP is at 0x2b60
>>eax: 000023d6  ebx: 0000007a  ecx: 00010000  edx: 00000001
>>esi: dfed244e  edi: 0000006d  ebp: dfed0000  esp: dfed9eda
>>    
>>
>
>Your stack pointer became misaligned.  I thought Manfred fixed that?
>You don't have nmi_watchdog enabled on the kernel boot command line
>do you?
>  
>

I do not.  The command line, through grub after the kernel spec is, ro 
root=/dev/hda5 hdc=ide-scsi

By the way I thought of some more information that may, or may not, 
matter in this case.  This is all based on RH90, upgraded based on the 
Documentation/Changes file of course, and I'm using gcc version  3.2.2 
200330222 (Red Hat Linux 3.2.2-5).

Thanks for the response,
Gary

>  
>
>>ds: 00b0  es: 00b0  ss: 0068
>>Process Swapper (pid:1 threadinfo=dfed8000 task=c151b900)
>>Stack: 000003d6 23d629d2 00000000 815d006d 0000d3ff 00010001 9f2c80fc 
>>006dd408
>>9f2c0000 d3ff9f08 00060001 61f990c0c 010c010c 007b6054 0000007b 00a08000
>>601000b0 00a85fd6 00000082 000b0000 00010090 00a80000 00b00000 00a00001
>>Call Trace:
>>
>>Code: Bad EIP value.
>><0>Kernel panic: Attempted to kill init!
>>
>>....blinking  cursor and nothing else.
>>
>>If I remove the PNPBIOS option I get past this point.  I would be happy 
>>to email my full .config should anyone wish to look at it.  I'd also be 
>>happy to test any patches anyone may have for this issue.
>>
>>    
>>
>
>
>  
>