Messages in this thread | ![/](/images/icornerl.gif) | | Date | Wed, 27 Dec 2006 00:40:45 +0100 | From | Ingo Molnar <> | Subject | Re: 2.6.19.1-rt15: BUG in __tasklet_action at kernel/softirq.c:568 |
| |
* Steven Rostedt <rostedt@goodmis.org> wrote:
> On Wed, 2006-12-20 at 20:50 +0100, Ingo Molnar wrote: > > * Robert Crocombe <rcrocomb@gmail.com> wrote: > > > > > On 12/19/06, Ingo Molnar <mingo@elte.hu> wrote: > > > >yeah. This is something that triggers very rarely on certain boxes. Not > > > >fixed yet, and it's been around for some time. > > > > > > Is there anything you would like me to do to help diagnose this? > > > > to figure out what the bug is :-/ Below is the tasklet redesign patch - > > the bug must be in there somewhere. > > > +static inline int tasklet_tryunlock(struct tasklet_struct *t) > > +{ > > + return cmpxchg(&t->state, TASKLET_STATEF_RUN, 0) == TASKLET_STATEF_RUN; > > +} > > + > > This probably isn't it, but is cmpxchg available on all archs now?
yeah, it's probably not related, i saw these failures on plain i686 too, which definitely has cmpxchg support. The failures i saw happened on a hyperthreading CPU, so i guess it must be some sort of narrow race.
Ingo - 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/
| ![\](/images/icornerr.gif) |