On 1 June 2016 at 17:31, Dietmar Eggemann <dietmar.eggem...@arm.com> wrote: > On 30/05/16 16:52, Vincent Guittot wrote: >> The cfs_rq->avg.last_update_time is initialize to 0 with the main effect >> that the 1st sched_entity that will be attached, will keep its > > attached in .task_move_group ? > > I'm not sure if we can have a .switched_to() directly followed by a > .enqueue_task() into a cfs_rq with avg.last_update_time = 0.
I think it can happen as well if the task is not queued during the change of scheduling class because when we attach the task in switched_to, the task->se.avg.last_update_time will stay to 0. So when the task will be enqueued, it will be attached one more time > >> last_update_time set to 0 and will attached once again during the >> enqueue. >> Initialize cfs_rq->avg.last_update_time to 1 instead. > > Maybe worth mentioning in the header: > > This double se attaching for the first task which moves into the task > group owning this cfs_rq (.task_move_group() and .enqueue_task()) can > (obviously) only happen if CONFIG_FAIR_GROUP_SCHED is set. > > The reason for this is that we set 'se->avg.last_update_time = > cfs_rq->avg.last_update_time' in attach_entity_load_avg() and use > 'migrated = !sa->last_update_time' as a flag in > enqueue_entity_load_avg() to decide if we call attach_entity_load_avg() > (again) or only update se->avg . > > Tested-by: Dietmar Eggemann <dietmar.eggem...@arm.com> Thanks > >> Signed-off-by: Vincent Guittot <vincent.guit...@linaro.org> >> --- >> >> v3: >> - add initialization of load_last_update_time_copy for not 64bits system >> - move init into init_cfs_rq >> >> v2: >> - rq_clock_task(rq_of(cfs_rq)) can't be used because lock is not held >> >> kernel/sched/fair.c | 10 ++++++++++ >> 1 file changed, 10 insertions(+) >> >> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c >> index 218f8e8..86be9c1 100644 >> --- a/kernel/sched/fair.c >> +++ b/kernel/sched/fair.c >> @@ -8459,6 +8459,16 @@ void init_cfs_rq(struct cfs_rq *cfs_rq) >> cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime; >> #endif >> #ifdef CONFIG_SMP >> + /* >> + * Set last_update_time to something different from 0 to make >> + * sure the 1st sched_entity will not be attached twice: once >> + * when attaching the task to the group and one more time when >> + * enqueueing the task. >> + */ >> + cfs_rq->avg.last_update_time = 1; >> +#ifndef CONFIG_64BIT >> + cfs_rq->load_last_update_time_copy = 1; >> +#endif >> atomic_long_set(&cfs_rq->removed_load_avg, 0); >> atomic_long_set(&cfs_rq->removed_util_avg, 0); >> #endif