Re: Oops while running systemtap on the p6 machine against the kernel version 2.6.36-rc7-git3

[fche at redhat dot com (Frank Ch. Eigler)]

divya <dipraksh@linux.vnet.ibm.com> writes:

> While running systemtap tests on the p6 machine , against the kernel
> version 2.6.36-rc7-git3 Oops occured , here are the call trace

Did the oops happen during a systemtap module startup vs. operation
vs. shutdown?  stap -V version string?

>  BUG: spinlock bad magic on CPU#6, stapio/20398
> -- 0:conmux-control -- time-stamp -- Oct/13/10  2:49:18 --res
>  lock: c000000000fcfa18, .magic: 00000000, .owner:<none>/-1, .owner_cpu: 0
> Call Trace:
> [c0000001effbfab0] [c000000000011934] .show_stack+0x6c/0x16c (unreliable)
> [c0000001effbfb60] [c0000000002c9274] .spin_bug+0xb0/0xd4
> [c0000001effbfbf0] [c0000000002c953c] .do_raw_spin_lock+0x48/0x184
> [c0000001effbfc90] [c00000000054af78] ._raw_spin_lock+0x10/0x24
> [c0000001effbfd00] [d000000003015908] .__stp_time_timer_callback+0x94/0x13c [stap_75ce6f84d34f8665c9a6b8e27fb9ea95_818798]
> [...]
> kernel BUG at kernel/timer.c:681!
> Oops: Exception in kernel mode, sig: 5 [#1]
> SMP NR_CPUS=1024 NUMA pSeries
> [...]
> [c0000001effbfc50] [c0000001effbfd00] 0xc0000001effbfd00 (unreliable)
> [c0000001effbfd00] [d00000000301597c] .__stp_time_timer_callback+0x108/0x13c [stap_75ce6f84d34f8665c9a6b8e27fb9ea95_818798]
> [c0000001effbfdc0] [c00000000009c2f8] .run_timer_softirq+0x1d8/0x2a8


We have had occasional problems in the past with something like this:
http://sourceware.org/PR10651, but it never was tracked down to a
systemtap bug per se, as opposed to suspicions that the kernel was not
satisfying one of its guarantees w.r.t. del_timer_sync().

- FChE