On Tue, Sep 10, 2013 at 06:59:57PM -0700, Andy Lutomirski wrote: > On 09/10/2013 06:08 AM, Peter Zijlstra wrote: > > In order to combine the preemption and need_resched test we need to > > fold the need_resched information into the preempt_count value. > > > > We keep the existing TIF_NEED_RESCHED infrastructure in place but at 3 > > sites test it and fold its value into preempt_count; namely: > > > > - resched_task() when setting TIF_NEED_RESCHED on the current task > > - scheduler_ipi() when resched_task() sets TIF_NEED_RESCHED on a > > remote task it follows it up with a reschedule IPI > > and we can modify the cpu local preempt_count from > > there. > > - cpu_idle_loop() for when resched_task() found tsk_is_polling(). > > > It looks like the intel_idle code can get confused if TIF_NEED_RESCHED > is set but the preempt resched bit is not -- the need_resched call > between monitor and mwait won't notice TIF_NEED_RESCHED. > > Is this condition possible? Ah indeed, I'll have to go find all idle loops out there it seems. Now if only they were easy to spot :/ I was hoping the generic idle thing was good enough, apparently not quite. Thanks for spotting that. -- To unsubscribe from this list: send the line "unsubscribe linux-arch" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
