Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-15 Thread Oleg Nesterov
On 06/15, john stultz wrote: > > On Fri, 2007-06-15 at 19:52 +0400, Oleg Nesterov wrote: > > > > Could you please look at the message below? I sent it privately near a month > > ago, but I think these problems are not fixed yet. > > Hmm. Maybe you sent it to others on the cc list, as I can't find

Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-15 Thread john stultz
On Fri, 2007-06-15 at 19:52 +0400, Oleg Nesterov wrote: > john stultz wrote: > > > > The following additional patch should correct this issue. Although since > > we weren't actually hitting it, the issue is a bit theoretical, so I've > > not been able to prove it really fixes anything. > > Could y

Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-15 Thread Oleg Nesterov
john stultz wrote: > > The following additional patch should correct this issue. Although since > we weren't actually hitting it, the issue is a bit theoretical, so I've > not been able to prove it really fixes anything. Could you please look at the message below? I sent it privately near a month

Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-14 Thread john stultz
On Tue, 2007-06-05 at 19:17 -0700, john stultz wrote: > Hey Ingo, > So we've been seeing the following trace fairly frequently on our SMP > boxes when running kernbench: > > BUG: at kernel/softirq.c:639 __tasklet_action() > > Call Trace: > [] dump_trace+0xaa/0x32a > [] show_trace+0x41/0x5

Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-06 Thread john stultz
On Wed, 2007-06-06 at 11:45 +0200, Ingo Molnar wrote: > * john stultz <[EMAIL PROTECTED]> wrote: > > > This fix avoids this race by making sure *after* we've locked the > > tasklet that the STATE_SCHED bit is set before adding it to the list. > > > > Does it look ok to you? > > ah - nice!! What

Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-06 Thread Jesper Juhl
On 06/06/07, john stultz <[EMAIL PROTECTED]> wrote: --- 2.6-rt.orig/kernel/softirq.c2007-06-05 18:30:54.0 -0700 +++ 2.6-rt/kernel/softirq.c 2007-06-05 18:36:44.0 -0700 @@ -544,10 +544,17 @@ static void inline __tasklet_common_schedule(struct tasklet_struct *t, struct

Re: [PATCH -rt] Fix TASKLET_STATE_SCHED WARN_ON()

2007-06-06 Thread Ingo Molnar
* john stultz <[EMAIL PROTECTED]> wrote: > This fix avoids this race by making sure *after* we've locked the > tasklet that the STATE_SCHED bit is set before adding it to the list. > > Does it look ok to you? ah - nice!! What would be the worst-case effect of this bug? (besides the WARN_ON()?