Re: [RFC][PATCH 3/7] sched: Allow balance callbacks for check_class_changed()

From: Peter Zijlstra <peterz@infradead.org>
To: Kirill Tkhai <tkhai@yandex.ru>
Cc: "umgwanakikbuti@gmail.com" <umgwanakikbuti@gmail.com>,
	"mingo@elte.hu" <mingo@elte.hu>,
	"ktkhai@parallels.com" <ktkhai@parallels.com>,
	"rostedt@goodmis.org" <rostedt@goodmis.org>,
	"juri.lelli@gmail.com" <juri.lelli@gmail.com>,
	"pang.xunlei@linaro.org" <pang.xunlei@linaro.org>,
	"oleg@redhat.com" <oleg@redhat.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"luca.abeni@unitn.it" <luca.abeni@unitn.it>
Subject: Re: [RFC][PATCH 3/7] sched: Allow balance callbacks for check_class_changed()
Date: Wed, 3 Jun 2015 12:45:34 +0200	[thread overview]
Message-ID: <20150603104534.GB3644@twins.programming.kicks-ass.net> (raw)
In-Reply-To: <226151433328161@web20o.yandex.ru>

On Wed, Jun 03, 2015 at 01:42:41PM +0300, Kirill Tkhai wrote:
> [resend]
> 
> 03.06.2015, 10:32, "Peter Zijlstra" <peterz@infradead.org>:
> > On Tue, Jun 02, 2015 at 07:27:19PM +0300, Kirill Tkhai wrote:
> >>>  + * task_dead_dl() will cancel our timer if we happen to die while
> >>>  + * its still pending.
> >>  task_dead_dl() is called for tasks of deadline class only. So if we do that,
> >>  the timer may be executed after final task's dead.
> >
> > Indeed; sleep deprived brain misses the obvious :/
> 
> I have an idea, but it require small hrtimer modification. We may use
> task_struct counters {get,put}_task_struct(), when we're starting or cancelling
> the timer, and in timer handler. But it require to return back the commit:

Yeah, already working on that. :-)