Messages in this thread | | | Subject | Re: [patch] Re: [regression bisect -next] BUG: using smp_processor_id() in preemptible [00000000] code: rmmod | From | Mike Galbraith <> | Date | Thu, 29 Oct 2009 10:19:27 +0100 |
| |
On Thu, 2009-10-29 at 10:14 +0100, Ingo Molnar wrote:
> hm, the problem is kthread_bind(). It is rummaging around in scheduler > internals without holding the runqueue lock - and this now got exposed. > Even though it is operating on (supposedly ...) inactive tasks, the guts > of that function should be moved into sched.c and it should be fixed to > have proper locking.
Yeah, I was thinking that nobody should ever be able to hit that without it being a bug.. but wimped out.
-Mike
| |