From: Peter Zijlstra <peterz@infradead.org>
To: umgwanakikbuti@gmail.com, mingo@elte.hu
Cc: ktkhai@parallels.com, rostedt@goodmis.org, tglx@linutronix.de,
juri.lelli@gmail.com, pang.xunlei@linaro.org, oleg@redhat.com,
wanpeng.li@linux.intel.com, linux-kernel@vger.kernel.org,
peterz@infradead.org
Subject: [PATCH 00/14] sched: balance callbacks
Date: Fri, 05 Jun 2015 10:48:36 +0200 [thread overview]
Message-ID: <20150605084836.364306429@infradead.org> (raw)
Mike stumbled over a cute bug where the RT/DL balancing ops caused a bug.
The exact scenario is __sched_setscheduler() changing a (runnable) task from
FIFO to OTHER. In swiched_from_rt(), where we do pull_rt_task() we temporarity
drop rq->lock. This gap allows regular cfs load-balancing to step in and
migrate our task.
However, check_class_changed() will happily continue with switched_to_fair()
which assumes our task is still on the old rq and makes the kernel go boom.
Instead of trying to patch this up and make things complicated; simply disallow
these methods to drop rq->lock and extend the current post_schedule stuff into
a balancing callback list, and use that.
This survives Mike's testcase.
Changes since -v2:
- reworked the hrtimer patch. -- Kirill, tglx
- added lock pinning
Changes since -v1:
- make SMP=n build,
- cured switched_from_dl()'s cancel_dl_timer().
no real tests on the new parts other than booting / building kernels.
next reply other threads:[~2015-06-05 9:01 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-06-05 8:48 Peter Zijlstra [this message]
2015-06-05 8:48 ` [PATCH 01/14] sched: Replace post_schedule with a balance callback list Peter Zijlstra
2015-06-05 8:48 ` [PATCH 02/14] sched: Use replace normalize_task() with __sched_setscheduler() Peter Zijlstra
2015-06-05 8:48 ` [PATCH 03/14] sched: Allow balance callbacks for check_class_changed() Peter Zijlstra
2015-06-05 8:48 ` [PATCH 04/14] sched,rt: Remove return value from pull_rt_task() Peter Zijlstra
2015-06-05 8:48 ` [PATCH 05/14] sched,rt: Convert switched_{from,to}_rt() / prio_changed_rt() to balance callbacks Peter Zijlstra
2015-06-05 8:48 ` [PATCH 06/14] sched,dl: Remove return value from pull_dl_task() Peter Zijlstra
2015-06-05 8:48 ` [PATCH 07/14] sched,dl: Convert switched_{from,to}_dl() / prio_changed_dl() to balance callbacks Peter Zijlstra
2015-06-05 8:48 ` [PATCH 08/14] hrtimer: Allow hrtimer::function() to free the timer Peter Zijlstra
2015-06-05 9:48 ` Thomas Gleixner
2015-06-07 19:43 ` Oleg Nesterov
2015-06-07 22:33 ` Oleg Nesterov
2015-06-07 22:56 ` Oleg Nesterov
2015-06-08 8:06 ` Thomas Gleixner
2015-06-08 9:14 ` Peter Zijlstra
2015-06-08 10:55 ` Peter Zijlstra
2015-06-08 12:42 ` Peter Zijlstra
2015-06-08 14:27 ` Oleg Nesterov
2015-06-08 14:42 ` Peter Zijlstra
2015-06-08 15:49 ` Oleg Nesterov
2015-06-08 15:10 ` Peter Zijlstra
2015-06-08 15:16 ` Oleg Nesterov
2015-06-09 21:33 ` Oleg Nesterov
2015-06-09 21:39 ` Oleg Nesterov
2015-06-10 6:55 ` Peter Zijlstra
2015-06-10 7:46 ` Kirill Tkhai
2015-06-10 16:04 ` Oleg Nesterov
2015-06-11 7:31 ` Peter Zijlstra
2015-06-11 16:25 ` Kirill Tkhai
2015-06-10 15:49 ` Oleg Nesterov
2015-06-10 22:37 ` Peter Zijlstra
2015-06-08 14:03 ` Oleg Nesterov
2015-06-08 14:17 ` Peter Zijlstra
2015-06-08 15:10 ` [PATCH 0/3] hrtimer: HRTIMER_STATE_ fixes Oleg Nesterov
2015-06-08 15:11 ` [PATCH 2/3] hrtimer: turn newstate arg of __remove_hrtimer() into clear_enqueued Oleg Nesterov
2015-06-08 15:11 ` [PATCH 3/3] hrtimer: fix the __hrtimer_start_range_ns() race with hrtimer_active() Oleg Nesterov
2015-06-08 15:12 ` [PATCH 1/3] hrtimer: kill HRTIMER_STATE_MIGRATE, fix the race with hrtimer_is_queued() Oleg Nesterov
2015-06-08 15:35 ` [PATCH 0/3] hrtimer: HRTIMER_STATE_ fixes Peter Zijlstra
2015-06-08 15:56 ` Oleg Nesterov
2015-06-08 17:11 ` Thomas Gleixner
2015-06-08 19:08 ` Peter Zijlstra
2015-06-08 20:52 ` Oleg Nesterov
2015-06-08 15:10 ` [PATCH 1/3] hrtimer: kill HRTIMER_STATE_MIGRATE, fix the race with hrtimer_is_queued() Oleg Nesterov
2015-06-08 15:13 ` Oleg Nesterov
2015-06-05 8:48 ` [PATCH 09/14] sched,dl: Fix sched class hopping CBS hole Peter Zijlstra
2015-06-05 8:48 ` [PATCH 10/14] sched: Move code around Peter Zijlstra
2015-06-05 8:48 ` [PATCH 11/14] sched: Streamline the task migration locking a little Peter Zijlstra
2015-06-05 8:48 ` [PATCH 12/14] lockdep: Simplify lock_release() Peter Zijlstra
2015-06-05 8:48 ` [PATCH 13/14] lockdep: Implement lock pinning Peter Zijlstra
2015-06-05 9:55 ` Ingo Molnar
2015-06-11 11:37 ` Peter Zijlstra
2015-06-05 8:48 ` [PATCH 14/14] sched,lockdep: Employ " Peter Zijlstra
2015-06-05 9:57 ` Ingo Molnar
2015-06-05 11:03 ` Peter Zijlstra
2015-06-05 11:24 ` Ingo Molnar
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=20150605084836.364306429@infradead.org \
--to=peterz@infradead.org \
--cc=juri.lelli@gmail.com \
--cc=ktkhai@parallels.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=oleg@redhat.com \
--cc=pang.xunlei@linaro.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=umgwanakikbuti@gmail.com \
--cc=wanpeng.li@linux.intel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome