2006-08-28 09:15:24

by Joerg Platte

[permalink] [raw]
Subject: 2.6.17.11: soft lockup detected on CPU#0

Hi,

I'm regulary suspending my notebook. After resume skype typically produces a
lot of processor load when trying to call somebody and I have to restart it
to make it work again. Today, with a new kernel, skype seems to cause a soft
lockup resulting in a freeze for a few seconds:

BUG: soft lockup detected on CPU#0!
<c012dd51> softlockup_tick+0x85/0x99 <c011b6d4>
update_process_times+0x35/0x57
<c01053b1> timer_interrupt+0x3e/0x69 <c012ddee> handle_IRQ_event+0x23/0x4c
<c012de8f> __do_IRQ+0x78/0xd1 <c01042cb> do_IRQ+0x19/0x24
<c0102cae> common_interrupt+0x1a/0x20
BUG: soft lockup detected on CPU#0!
<c012dd51> softlockup_tick+0x85/0x99 <c011b6d4>
update_process_times+0x35/0x57
<c01053b1> timer_interrupt+0x3e/0x69 <c012ddee> handle_IRQ_event+0x23/0x4c
<c012de8f> __do_IRQ+0x78/0xd1 <c01042cb> do_IRQ+0x19/0x24
<c0102cae> common_interrupt+0x1a/0x20

I know, skype is closed source and cannot be debugged, but even a closed
source userland program should not be able to cause a lockup. Is the
information given above useful to find the problem inside the kernel? If not,
what can I do to provide better debug information? I'm using a vanilla
2.6.17.11 kernel:
Linux ibm 2.6.17.11 #1 PREEMPT Sat Aug 26 18:55:43 CEST 2006 i686 GNU/Linux

regards,
Jörg