From: tip-bot for Peter Zijlstra <tipbot@zytor.com>
To: linux-tip-commits@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, hpa@zytor.com,
umgwanakikbuti@gmail.com, peterz@infradead.org, mingo@kernel.org,
tglx@linutronix.de
Subject: [tip:sched/hrtimers] sched: Allow balance callbacks for check_class_changed()
Date: Thu, 18 Jun 2015 16:01:09 -0700 [thread overview]
Message-ID: <tip-4c9a4bc89a9cca8128bce67d6bc8870d6b7ee0b2@git.kernel.org> (raw)
In-Reply-To: <20150611124742.615343911@infradead.org>
Commit-ID: 4c9a4bc89a9cca8128bce67d6bc8870d6b7ee0b2
Gitweb: http://git.kernel.org/tip/4c9a4bc89a9cca8128bce67d6bc8870d6b7ee0b2
Author: Peter Zijlstra <peterz@infradead.org>
AuthorDate: Thu, 11 Jun 2015 14:46:39 +0200
Committer: Thomas Gleixner <tglx@linutronix.de>
CommitDate: Fri, 19 Jun 2015 00:25:26 +0200
sched: Allow balance callbacks for check_class_changed()
In order to remove dropping rq->lock from the
switched_{to,from}()/prio_changed() sched_class methods, run the
balance callbacks after it.
We need to remove dropping rq->lock because its buggy,
suppose using sched_setattr()/sched_setscheduler() to change a running
task from FIFO to OTHER.
By the time we get to switched_from_rt() the task is already enqueued
on the cfs runqueues. If switched_from_rt() does pull_rt_task() and
drops rq->lock, load-balancing can come in and move our task @p to
another rq.
The subsequent switched_to_fair() still assumes @p is on @rq and bad
things will happen.
By using balance callbacks we delay the load-balancing operations
{rt,dl}x{push,pull} until we've done all the important work and the
task is fully set up.
Furthermore, the balance callbacks do not know about @p, therefore
they cannot get confused like this.
Reported-by: Mike Galbraith <umgwanakikbuti@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: ktkhai@parallels.com
Cc: rostedt@goodmis.org
Cc: juri.lelli@gmail.com
Cc: pang.xunlei@linaro.org
Cc: oleg@redhat.com
Cc: wanpeng.li@linux.intel.com
Link: http://lkml.kernel.org/r/20150611124742.615343911@infradead.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
kernel/sched/core.c | 25 ++++++++++++++++++++++---
1 file changed, 22 insertions(+), 3 deletions(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index b610ef9..ef546e3 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1001,7 +1001,11 @@ inline int task_curr(const struct task_struct *p)
}
/*
- * Can drop rq->lock because from sched_class::switched_from() methods drop it.
+ * switched_from, switched_to and prio_changed must _NOT_ drop rq->lock,
+ * use the balance_callback list if you want balancing.
+ *
+ * this means any call to check_class_changed() must be followed by a call to
+ * balance_callback().
*/
static inline void check_class_changed(struct rq *rq, struct task_struct *p,
const struct sched_class *prev_class,
@@ -1010,7 +1014,7 @@ static inline void check_class_changed(struct rq *rq, struct task_struct *p,
if (prev_class != p->sched_class) {
if (prev_class->switched_from)
prev_class->switched_from(rq, p);
- /* Possble rq->lock 'hole'. */
+
p->sched_class->switched_to(rq, p);
} else if (oldprio != p->prio || dl_task(p))
p->sched_class->prio_changed(rq, p, oldprio);
@@ -1491,8 +1495,12 @@ ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags)
p->state = TASK_RUNNING;
#ifdef CONFIG_SMP
- if (p->sched_class->task_woken)
+ if (p->sched_class->task_woken) {
+ /*
+ * XXX can drop rq->lock; most likely ok.
+ */
p->sched_class->task_woken(rq, p);
+ }
if (rq->idle_stamp) {
u64 delta = rq_clock(rq) - rq->idle_stamp;
@@ -3100,7 +3108,11 @@ void rt_mutex_setprio(struct task_struct *p, int prio)
check_class_changed(rq, p, prev_class, oldprio);
out_unlock:
+ preempt_disable(); /* avoid rq from going away on us */
__task_rq_unlock(rq);
+
+ balance_callback(rq);
+ preempt_enable();
}
#endif
@@ -3661,11 +3673,18 @@ change:
}
check_class_changed(rq, p, prev_class, oldprio);
+ preempt_disable(); /* avoid rq from going away on us */
task_rq_unlock(rq, p, &flags);
if (pi)
rt_mutex_adjust_pi(p);
+ /*
+ * Run balance callbacks after we've adjusted the PI chain.
+ */
+ balance_callback(rq);
+ preempt_enable();
+
return 0;
}
next prev parent reply other threads:[~2015-06-18 23:01 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-06-11 12:46 [PATCH 00/18] sched: balance callbacks v4 Peter Zijlstra
2015-06-11 12:46 ` [PATCH 01/18] sched: Replace post_schedule with a balance callback list Peter Zijlstra
2015-06-11 15:32 ` Kirill Tkhai
2015-06-18 23:00 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 02/18] sched: Use replace normalize_task() with __sched_setscheduler() Peter Zijlstra
2015-06-18 23:00 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 03/18] sched: Allow balance callbacks for check_class_changed() Peter Zijlstra
2015-06-18 23:01 ` tip-bot for Peter Zijlstra [this message]
2015-06-11 12:46 ` [PATCH 04/18] sched,rt: Remove return value from pull_rt_task() Peter Zijlstra
2015-06-18 23:01 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 05/18] sched,rt: Convert switched_{from,to}_rt() / prio_changed_rt() to balance callbacks Peter Zijlstra
2015-06-18 23:01 ` [tip:sched/hrtimers] sched, rt: Convert switched_{from, to}_rt() " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 06/18] sched,dl: Remove return value from pull_dl_task() Peter Zijlstra
2015-06-18 23:02 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 07/18] sched,dl: Convert switched_{from,to}_dl() / prio_changed_dl() to balance callbacks Peter Zijlstra
2015-06-18 23:02 ` [tip:sched/hrtimers] sched, dl: Convert switched_{from, to}_dl() " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 08/18] hrtimer: Remove HRTIMER_STATE_MIGRATE Peter Zijlstra
2015-06-18 22:18 ` [tip:timers/core] " tip-bot for Oleg Nesterov
2015-06-11 12:46 ` [PATCH 09/18] hrtimer: Fix hrtimer_is_queued() hole Peter Zijlstra
2015-06-18 22:18 ` [tip:timers/core] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 10/18] seqcount: Rename write_seqcount_barrier() Peter Zijlstra
2015-06-18 22:19 ` [tip:timers/core] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 11/18] seqcount: Introduce raw_write_seqcount_barrier() Peter Zijlstra
2015-06-11 15:33 ` Paul E. McKenney
2015-06-11 21:45 ` Paul E. McKenney
2015-06-12 7:08 ` Peter Zijlstra
2015-06-12 18:59 ` Oleg Nesterov
2015-06-17 12:29 ` Peter Zijlstra
2015-06-17 14:57 ` Paul E. McKenney
2015-06-17 15:11 ` Peter Zijlstra
2015-06-17 15:42 ` Paul E. McKenney
2015-06-17 16:58 ` Peter Zijlstra
2015-06-17 15:49 ` Peter Zijlstra
2015-06-17 16:37 ` Paul E. McKenney
2015-06-17 17:11 ` Peter Zijlstra
2015-06-17 18:02 ` Paul E. McKenney
2015-06-18 9:15 ` Peter Zijlstra
2015-06-18 9:40 ` Ingo Molnar
2015-06-18 10:40 ` Peter Zijlstra
2015-06-18 16:54 ` Paul E. McKenney
2015-06-18 17:10 ` Steven Rostedt
2015-06-18 17:51 ` Paul E. McKenney
2015-06-18 22:19 ` [tip:timers/core] seqcount: Introduce raw_write_seqcount_barrier( ) tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 12/18] hrtimer: Allow hrtimer::function() to free the timer Peter Zijlstra
2015-06-18 22:19 ` [tip:timers/core] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 13/18] sched,dl: Fix sched class hopping CBS hole Peter Zijlstra
2015-06-18 23:02 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 14/18] sched: Move code around Peter Zijlstra
2015-06-18 23:02 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 15/18] sched: Streamline the task migration locking a little Peter Zijlstra
2015-06-18 23:03 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 16/18] lockdep: Simplify lock_release() Peter Zijlstra
2015-06-18 23:03 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 17/18] lockdep: Implement lock pinning Peter Zijlstra
2015-06-18 23:03 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-06-11 12:46 ` [PATCH 18/18] sched,lockdep: Employ " Peter Zijlstra
2015-06-18 23:04 ` [tip:sched/hrtimers] " tip-bot for Peter Zijlstra
2015-12-29 5:41 ` [PATCH 00/18] sched: balance callbacks v4 Byungchul Park
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=tip-4c9a4bc89a9cca8128bce67d6bc8870d6b7ee0b2@git.kernel.org \
--to=tipbot@zytor.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-tip-commits@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=umgwanakikbuti@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.