From: Peter Zijlstra <peterz@infradead.org>
To: Ankur Arora <ankur.a.arora@oracle.com>
Cc: bigeasy@linutronix.de, tglx@linutronix.de, mingo@kernel.org,
linux-kernel@vger.kernel.org, juri.lelli@redhat.com,
vincent.guittot@linaro.org, dietmar.eggemann@arm.com,
rostedt@goodmis.org, bsegall@google.com, mgorman@suse.de,
vschneid@redhat.com, efault@gmx.de
Subject: Re: [PATCH 2/5] sched: Add Lazy preemption model
Date: Tue, 8 Oct 2024 16:48:29 +0200 [thread overview]
Message-ID: <20241008144829.GG14587@noisy.programming.kicks-ass.net> (raw)
In-Reply-To: <877cajdu8x.fsf@oracle.com>
On Mon, Oct 07, 2024 at 10:43:58PM -0700, Ankur Arora wrote:
> > @@ -519,7 +525,7 @@ static inline bool preempt_model_rt(void
> > */
> > static inline bool preempt_model_preemptible(void)
> > {
> > - return preempt_model_full() || preempt_model_rt();
> > + return preempt_model_full() || preempt_model_lazy() || preempt_model_rt();
> > }
>
> In addition to preempt_model_preemptible() we probably also need
>
> static inline bool preempt_model_minimize_latency(void)
> {
> return preempt_model_full() || preempt_model_rt();
> }
>
> for spin_needbreak()/rwlock_needbreak().
>
> That would make the behaviour of spin_needbreak() under the lazy model
> similar to none/voluntary.
That whole thing needs rethinking, for one the preempt_model_rt() one
doesn't really make sense anymore at the end of this.
> > --- a/kernel/sched/core.c
> > +++ b/kernel/sched/core.c
> > @@ -1078,6 +1078,9 @@ static void __resched_curr(struct rq *rq
> >
> > lockdep_assert_rq_held(rq);
> >
> > + if (is_idle_task(curr) && tif == TIF_NEED_RESCHED_LAZY)
> > + tif = TIF_NEED_RESCHED;
> > +
>
> Tasks with idle policy get handled at the usual user space boundary.
> Maybe a comment reflecting that?
is_idle_task() != SCHED_IDLE. This is about the idle task, which you
want to force preempt always. But I can stick a comment on.
> > @@ -5598,6 +5627,10 @@ void sched_tick(void)
> > update_rq_clock(rq);
> > hw_pressure = arch_scale_hw_pressure(cpu_of(rq));
> > update_hw_load_avg(rq_clock_task(rq), rq, hw_pressure);
> > +
> > + if (dynamic_preempt_lazy() && tif_test_bit(TIF_NEED_RESCHED_LAZY))
> > + resched_curr(rq);
> > +
>
> So this works for SCHED_NORMAL. But, does this do the right thing for
> deadline etc other scheduling classes?
Yeah, only fair.c uses resched_curr_laz8(), the others still use
resched_curr() and will work as if Full.
So that is: SCHED_IDLE, SCHED_BATCH and SCHED_NORMAL/OTHER get the lazy
thing, FIFO, RR and DEADLINE get the traditional Full behaviour.
next prev parent reply other threads:[~2024-10-08 14:48 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-07 7:46 [PATCH 0/5] sched: Lazy preemption muck Peter Zijlstra
2024-10-07 7:46 ` [PATCH 1/5] sched: Add TIF_NEED_RESCHED_LAZY infrastructure Peter Zijlstra
2024-10-09 12:18 ` Sebastian Andrzej Siewior
2024-10-09 13:01 ` Peter Zijlstra
2024-11-06 10:48 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2024-10-07 7:46 ` [PATCH 2/5] sched: Add Lazy preemption model Peter Zijlstra
2024-10-08 5:43 ` Ankur Arora
2024-10-08 14:48 ` Peter Zijlstra [this message]
2024-10-09 8:50 ` Sebastian Andrzej Siewior
2024-10-09 9:14 ` Peter Zijlstra
2024-10-09 9:19 ` Sebastian Andrzej Siewior
2024-10-15 14:37 ` Shrikanth Hegde
2024-10-25 10:42 ` Sebastian Andrzej Siewior
2024-10-22 16:44 ` Shrikanth Hegde
2024-10-25 13:19 ` Sebastian Andrzej Siewior
2024-10-29 18:57 ` Shrikanth Hegde
2024-11-06 10:48 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2024-10-07 7:46 ` [PATCH 3/5] sched: Enable PREEMPT_DYNAMIC for PREEMPT_RT Peter Zijlstra
2024-10-08 13:24 ` Sebastian Andrzej Siewior
2024-10-08 14:40 ` Peter Zijlstra
2024-10-10 6:52 ` Christoph Hellwig
2024-10-10 7:50 ` Peter Zijlstra
2024-11-06 10:48 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2024-10-07 7:46 ` [PATCH 4/5] sched, x86: Enable Lazy preemption Peter Zijlstra
2024-11-06 10:48 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2024-10-07 7:46 ` [PATCH 5/5] sched: Add laziest preempt model Peter Zijlstra
2024-10-08 5:59 ` Ankur Arora
2024-10-08 14:23 ` Thomas Gleixner
2024-10-08 14:40 ` Peter Zijlstra
2024-10-08 15:07 ` Sebastian Andrzej Siewior
2024-10-07 8:33 ` [PATCH 0/5] sched: Lazy preemption muck Sebastian Andrzej Siewior
2024-10-08 4:58 ` Mike Galbraith
2024-10-08 15:32 ` Sebastian Andrzej Siewior
2024-10-09 4:40 ` Ankur Arora
2024-10-09 6:20 ` Sebastian Andrzej Siewior
2024-10-09 7:23 ` Ankur Arora
2024-10-09 8:02 ` Peter Zijlstra
2024-10-09 8:45 ` Sebastian Andrzej Siewior
2024-10-09 14:01 ` Steven Rostedt
2024-10-09 20:13 ` Thomas Gleixner
2024-10-09 20:43 ` Steven Rostedt
2024-10-09 21:06 ` Thomas Gleixner
2024-10-09 21:19 ` Steven Rostedt
2024-10-09 23:16 ` Thomas Gleixner
2024-10-09 23:29 ` Steven Rostedt
2024-10-10 1:20 ` Thomas Gleixner
2024-10-10 10:23 ` David Laight
2024-10-13 19:02 ` Thomas Gleixner
2024-10-14 8:21 ` David Laight
2024-10-10 3:12 ` Tianchen Ding
2024-10-10 7:47 ` Thomas Gleixner
2024-10-09 7:30 ` Ankur Arora
2024-10-09 7:46 ` Peter Zijlstra
2024-10-09 11:07 ` Sebastian Andrzej Siewior
2024-10-17 12:36 ` Mike Galbraith
2024-11-07 17:21 ` Thomas Meyer
2024-11-08 0:59 ` Mike Galbraith
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=20241008144829.GG14587@noisy.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=ankur.a.arora@oracle.com \
--cc=bigeasy@linutronix.de \
--cc=bsegall@google.com \
--cc=dietmar.eggemann@arm.com \
--cc=efault@gmx.de \
--cc=juri.lelli@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@suse.de \
--cc=mingo@kernel.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=vincent.guittot@linaro.org \
--cc=vschneid@redhat.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