From de5bdff7a72acc281219be2b8edeeca1fd81c542 Mon Sep 17 00:00:00 2001 From: Hiroshi Shimamoto Date: Thu, 16 Feb 2012 14:52:21 +0900 Subject: sched: Make initial SCHED_RR timeslace DEF_TIMESLICE Current the initial SCHED_RR timeslice of init_task is HZ, which means 1s, and is not same as the default SCHED_RR timeslice DEF_TIMESLICE. Change that initial timeslice to the DEF_TIMESLICE. Signed-off-by: Hiroshi Shimamoto [ s/DEF_TIMESLICE/RR_TIMESLICE/g ] Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/4F3C9995.3010800@ct.jp.nec.com Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/sched/rt.c') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index f42ae7fb5ec5..f70206c2c802 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -1972,7 +1972,7 @@ static void task_tick_rt(struct rq *rq, struct task_struct *p, int queued) if (--p->rt.time_slice) return; - p->rt.time_slice = DEF_TIMESLICE; + p->rt.time_slice = RR_TIMESLICE; /* * Requeue to the end of queue if we are not the only element @@ -2000,7 +2000,7 @@ static unsigned int get_rr_interval_rt(struct rq *rq, struct task_struct *task) * Time slice is 0 for SCHED_FIFO tasks */ if (task->policy == SCHED_RR) - return DEF_TIMESLICE; + return RR_TIMESLICE; else return 0; } -- cgit v1.2.3 From 42c62a589f1ccbf38a02cb732231f9c2fccc5ab0 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 18 Oct 2011 22:03:48 +0200 Subject: sched/rt: Keep period timer ticking when rt throttling is active When a runqueue is throttled we cannot disable the period timer because that timer is the only way to undo the throttling. We got stale throttling entries when a rq was throttled and then the global sysctl was disabled, which stopped the timer. Signed-off-by: Peter Zijlstra [ Added changelog ] Signed-off-by: Thomas Gleixner Link: http://lkml.kernel.org/n/tip-nuj34q52p6ro7szapuz84i0v@git.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'kernel/sched/rt.c') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index f70206c2c802..6d1eb0be1870 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -778,12 +778,9 @@ static inline int balance_runtime(struct rt_rq *rt_rq) static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun) { - int i, idle = 1; + int i, idle = 1, throttled = 0; const struct cpumask *span; - if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF) - return 1; - span = sched_rt_period_mask(); for_each_cpu(i, span) { int enqueue = 0; @@ -818,12 +815,17 @@ static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun) if (!rt_rq_throttled(rt_rq)) enqueue = 1; } + if (rt_rq->rt_throttled) + throttled = 1; if (enqueue) sched_rt_rq_enqueue(rt_rq); raw_spin_unlock(&rq->lock); } + if (!throttled && (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF)) + return 1; + return idle; } @@ -884,7 +886,8 @@ static void update_curr_rt(struct rq *rq) if (unlikely((s64)delta_exec < 0)) delta_exec = 0; - schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); + schedstat_set(curr->se.statistics.exec_max, + max(curr->se.statistics.exec_max, delta_exec)); curr->se.sum_exec_runtime += delta_exec; account_group_exec_runtime(curr, delta_exec); -- cgit v1.2.3 From 7abc63b1bd412f7655b62ef3e35c3c11c5134636 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 18 Oct 2011 22:03:48 +0200 Subject: sched/rt: Do not throttle when PI boosting When a runqueue has rt_runtime_us = 0 then the only way it can accumulate rt_time is via PI boosting. That causes the runqueue to be throttled and replenishing does not change anything due to rt_runtime_us = 0. So avoid that situation by clearing rt_time and skip the throttling alltogether. Signed-off-by: Peter Zijlstra [ Changelog ] Signed-off-by: Thomas Gleixner Link: http://lkml.kernel.org/n/tip-7x70cypsotjb4jvcor3edctk@git.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 20 ++++++++++++++++++-- 1 file changed, 18 insertions(+), 2 deletions(-) (limited to 'kernel/sched/rt.c') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 6d1eb0be1870..7f7e7cdcb472 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -857,8 +857,24 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq) return 0; if (rt_rq->rt_time > runtime) { - rt_rq->rt_throttled = 1; - printk_once(KERN_WARNING "sched: RT throttling activated\n"); + struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq); + + /* + * Don't actually throttle groups that have no runtime assigned + * but accrue some time due to boosting. + */ + if (likely(rt_b->rt_runtime)) { + rt_rq->rt_throttled = 1; + printk_once(KERN_WARNING "sched: RT throttling activated\n"); + } else { + /* + * In case we did anyway, make it go away, + * replenishment is a joke, since it will replenish us + * with exactly 0 ns. + */ + rt_rq->rt_time = 0; + } + if (rt_rq_throttled(rt_rq)) { sched_rt_rq_dequeue(rt_rq); return 1; -- cgit v1.2.3 From 3ccf3e8306156a28213adc720aba807e9a901ad5 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 27 Feb 2012 10:47:00 +0100 Subject: printk/sched: Introduce special printk_sched() for those awkward moments There's a few awkward printk()s inside of scheduler guts that people prefer to keep but really are rather deadlock prone. Fudge around it by storing the text in a per-cpu buffer and poll it using the existing printk_tick() handler. This will drop output when its more frequent than once a tick, however only the affinity thing could possible go that fast and for that just one should suffice to notify the admin he's done something silly.. Signed-off-by: Peter Zijlstra Cc: Linus Torvalds Cc: Andrew Morton Link: http://lkml.kernel.org/n/tip-wua3lmkt3dg8nfts66o6brne@git.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'kernel/sched/rt.c') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 7f7e7cdcb472..b60dad720173 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -864,8 +864,14 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq) * but accrue some time due to boosting. */ if (likely(rt_b->rt_runtime)) { + static bool once = false; + rt_rq->rt_throttled = 1; - printk_once(KERN_WARNING "sched: RT throttling activated\n"); + + if (!once) { + once = true; + printk_sched("sched: RT throttling activated\n"); + } } else { /* * In case we did anyway, make it go away, -- cgit v1.2.3 From 1b028abc779b67b699daff55e27d2432f8d92666 Mon Sep 17 00:00:00 2001 From: Michael J Wang Date: Mon, 19 Mar 2012 22:26:19 +0000 Subject: sched/rt: Improve pick_next_highest_task_rt() Avoid extra work by continuing on to the next rt_rq if the highest prio task in current rt_rq is the same priority as our candidate task. More detailed explanation: if next is not NULL, then we have found a candidate task, and its priority is next->prio. Now we are looking for an even higher priority task in the other rt_rq's. idx is the highest priority in the current candidate rt_rq. In the current 3.3 code, if idx is equal to next->prio, we would start scanning the tasks in that rt_rq and replace the current candidate task with a task from that rt_rq. But the new task would only have a priority that is equal to our previous candidate task, so we have not advanced our goal of finding a higher prio task. So we should avoid the extra work by continuing on to the next rt_rq if idx is equal to next->prio. Signed-off-by: Michael J Wang Acked-by: Steven Rostedt Reviewed-by: Yong Zhang Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/2EF88150C0EF2C43A218742ED384C1BC0FC83D6B@IRVEXCHMB08.corp.ad.broadcom.com Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/sched/rt.c') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index b60dad720173..44af55e6d5d0 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -1428,7 +1428,7 @@ static struct task_struct *pick_next_highest_task_rt(struct rq *rq, int cpu) next_idx: if (idx >= MAX_RT_PRIO) continue; - if (next && next->prio < idx) + if (next && next->prio <= idx) continue; list_for_each_entry(rt_se, array->queue + idx, run_list) { struct task_struct *p; -- cgit v1.2.3