Re: BUG: spinlockup in kacpid of 2.6.15-rc6 during boot

[Larry Finger <Larry.Finger@lwfinger.net>]

Adrian Bunk wrote:

> On Sun, Jan 01, 2006 at 10:58:19PM -0600, Larry Finger wrote:
>
>> I am getting an intermittent "BUG: spinlockup on CPU #0, kacpid/13, e3bcef9c" message when 
booting. I know it occurs from a cold boot, but am not sure about from a warm reboot.
>>
>> I don't know where in the boot sequence it occurs as I only have the text display on the screen, 
which is as follows:
>
>
>
> You can set a higher resolution using the vga= boot parameter (see 
Documentation/kernel-parameters.txt in the kernel sources).


Thanks - I'll set as high as I can.

>> ...... unknown material that rolled off the top of the screen
>> _spinlock + 0x1b/0x30
>> exit_io_context + 0x25/0x90
>> do_exit + 0x54/0x430
>> die + 0x17c/0x180
>> do_page_fault + 0x209/0x62e
>> error_code + 0x4f/0x54
>> show_stack + 0x9c/0x0e
>> show_registers + 0x18f/0x230
>> die + 0xfa/0x180
>> do_page_fault + 0x209/0x62e
>> BUG: spinlockup on CPU #0, kacpid/13, e3bcef9c
>> dump_stack + 0x1e/0x20
>> __spin_lock_debug + 0xb6/0xf0
>> _raw_spin_lock + 0x67/0x90
>> _spinlock + 0x1b/0x30
>> exit_io_context + 0x25/0x90
>> do_exit + 0x54/0x430
>> die + 0x17c/0x180
>> do_page_fault + 0x209/0x62e
>> error_code + 0x4f/0x54
>> show_stack + 0x9c/0x0e
>> show_registers + 0x18f/0x230
>> die + 0xfa/0x180
>> do_page_fault + 0x209/0x62e
>>
>> The computer is an HP ze1115 notebook with a mobile K7 processor. The beginning lines of a 
normal dmesg are:
>>
>> Linux version 2.6.15-rc6 (root@linux) (gcc version 4.0.2 20050901 (prerelease) (SUSE Linux)) #4 
PREEMPT
>> ...
>
>
>
> IOW, 2.6.15-rc6 does sometimes boot, and sometimes it doesn't boot?


Yes. I have never had it fail two times in a row. Powering off and restarting has fixed it everytime 
so far.

> Can you give information regarding which kernel versions are affected and which aren't, and what 
influences whether 2.6.15-rc6 does boot or does not boot?


I have not had this problem (yet?) with 2.6.15-rc7 or 2.6.15; however, it occurs so infrequently 
that I'm not sure I can rule out these later versions. The case I reported involved a dirty reiserfs 
file system. I had been debugging the bcm43xx driver and the system had frozen during shutdown 
because the module could not be unloaded. After I powered off for about 1 second, I restarted. The 
bug dump occurred fairly quickly and I believe it is early in the reboot sequence.

Larry


-
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/