From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
To: Scott Wood <swood@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Steven Rostedt <rostedt@goodmis.org>,
Peter Zijlstra <peterz@infradead.org>,
Juri Lelli <juri.lelli@redhat.com>,
Daniel Bristot de Oliveira <bristot@redhat.com>,
Clark Williams <williams@redhat.com>,
linux-kernel@vger.kernel.org, linux-rt-users@vger.kernel.org
Subject: Re: [PATCH RT 5/8] sched/deadline: Reclaim cpuset bandwidth in .migrate_task_rq()
Date: Tue, 17 Sep 2019 17:10:39 +0200 [thread overview]
Message-ID: <20190917151039.ukplro66tzgxg6xt@linutronix.de> (raw)
In-Reply-To: <20190727055638.20443-6-swood@redhat.com>
On 2019-07-27 00:56:35 [-0500], Scott Wood wrote:
> With the changes to migrate disabling, ->set_cpus_allowed() no longer
> gets deferred until migrate_enable(). To avoid releasing the bandwidth
> while the task may still be executing on the old CPU, move the subtraction
> to ->migrate_task_rq().
I assume this is still valid for the !RT case.
Could someone from the scheduler/DL department please comment on this
one?
> Signed-off-by: Scott Wood <swood@redhat.com>
> ---
> kernel/sched/deadline.c | 67 +++++++++++++++++++++++--------------------------
> 1 file changed, 31 insertions(+), 36 deletions(-)
>
> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
> index c18be51f7608..2f18d0cf1b56 100644
> --- a/kernel/sched/deadline.c
> +++ b/kernel/sched/deadline.c
> @@ -1606,14 +1606,42 @@ static void yield_task_dl(struct rq *rq)
> return cpu;
> }
>
> +static void free_old_cpuset_bw_dl(struct rq *rq, struct task_struct *p)
> +{
> + struct root_domain *src_rd = rq->rd;
> +
> + /*
> + * Migrating a SCHED_DEADLINE task between exclusive
> + * cpusets (different root_domains) entails a bandwidth
> + * update. We already made space for us in the destination
> + * domain (see cpuset_can_attach()).
> + */
> + if (!cpumask_intersects(src_rd->span, p->cpus_ptr)) {
> + struct dl_bw *src_dl_b;
> +
> + src_dl_b = dl_bw_of(cpu_of(rq));
> + /*
> + * We now free resources of the root_domain we are migrating
> + * off. In the worst case, sched_setattr() may temporary fail
> + * until we complete the update.
> + */
> + raw_spin_lock(&src_dl_b->lock);
> + __dl_sub(src_dl_b, p->dl.dl_bw, dl_bw_cpus(task_cpu(p)));
> + raw_spin_unlock(&src_dl_b->lock);
> + }
> +}
> +
> static void migrate_task_rq_dl(struct task_struct *p, int new_cpu __maybe_unused)
> {
> struct rq *rq;
>
> - if (p->state != TASK_WAKING)
> + rq = task_rq(p);
> +
> + if (p->state != TASK_WAKING) {
> + free_old_cpuset_bw_dl(rq, p);
> return;
> + }
>
> - rq = task_rq(p);
> /*
> * Since p->state == TASK_WAKING, set_task_cpu() has been called
> * from try_to_wake_up(). Hence, p->pi_lock is locked, but
> @@ -2220,39 +2248,6 @@ static void task_woken_dl(struct rq *rq, struct task_struct *p)
> }
> }
>
> -static void set_cpus_allowed_dl(struct task_struct *p,
> - const struct cpumask *new_mask)
> -{
> - struct root_domain *src_rd;
> - struct rq *rq;
> -
> - BUG_ON(!dl_task(p));
> -
> - rq = task_rq(p);
> - src_rd = rq->rd;
> - /*
> - * Migrating a SCHED_DEADLINE task between exclusive
> - * cpusets (different root_domains) entails a bandwidth
> - * update. We already made space for us in the destination
> - * domain (see cpuset_can_attach()).
> - */
> - if (!cpumask_intersects(src_rd->span, new_mask)) {
> - struct dl_bw *src_dl_b;
> -
> - src_dl_b = dl_bw_of(cpu_of(rq));
> - /*
> - * We now free resources of the root_domain we are migrating
> - * off. In the worst case, sched_setattr() may temporary fail
> - * until we complete the update.
> - */
> - raw_spin_lock(&src_dl_b->lock);
> - __dl_sub(src_dl_b, p->dl.dl_bw, dl_bw_cpus(task_cpu(p)));
> - raw_spin_unlock(&src_dl_b->lock);
> - }
> -
> - set_cpus_allowed_common(p, new_mask);
> -}
> -
> /* Assumes rq->lock is held */
> static void rq_online_dl(struct rq *rq)
> {
> @@ -2407,7 +2402,7 @@ static void prio_changed_dl(struct rq *rq, struct task_struct *p,
> #ifdef CONFIG_SMP
> .select_task_rq = select_task_rq_dl,
> .migrate_task_rq = migrate_task_rq_dl,
> - .set_cpus_allowed = set_cpus_allowed_dl,
> + .set_cpus_allowed = set_cpus_allowed_common,
> .rq_online = rq_online_dl,
> .rq_offline = rq_offline_dl,
> .task_woken = task_woken_dl,
> --
> 1.8.3.1
>
Sebastian
next prev parent reply other threads:[~2019-09-17 15:10 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-27 5:56 [RT PATCH 0/8] migrate disable fixes and performance Scott Wood
2019-07-27 5:56 ` [PATCH RT 1/8] sched: migrate_enable: Use sleeping_lock to indicate involuntary sleep Scott Wood
2019-07-29 13:26 ` Steven Rostedt
2019-07-27 5:56 ` [PATCH RT 2/8] sched: __set_cpus_allowed_ptr: Check cpus_mask, not cpus_ptr Scott Wood
2019-09-17 14:57 ` Sebastian Andrzej Siewior
2019-09-17 15:23 ` Scott Wood
2019-07-27 5:56 ` [PATCH RT 3/8] sched: Remove dead __migrate_disabled() check Scott Wood
2019-07-27 5:56 ` [PATCH RT 4/8] sched: migrate disable: Protect cpus_ptr with lock Scott Wood
2019-09-26 16:39 ` Sebastian Andrzej Siewior
2019-09-26 16:52 ` Scott Wood
2019-09-27 12:19 ` Sebastian Andrzej Siewior
2019-09-27 20:02 ` Scott Wood
2019-07-27 5:56 ` [PATCH RT 5/8] sched/deadline: Reclaim cpuset bandwidth in .migrate_task_rq() Scott Wood
2019-09-17 15:10 ` Sebastian Andrzej Siewior [this message]
2019-09-27 8:11 ` Juri Lelli
2019-09-27 16:40 ` Scott Wood
2019-09-30 7:12 ` Juri Lelli
2019-09-30 16:24 ` Scott Wood
2019-10-01 8:52 ` Juri Lelli
2019-10-09 6:25 ` Scott Wood
2019-10-09 7:27 ` Juri Lelli
2019-10-09 19:12 ` Scott Wood
2019-10-10 8:18 ` Juri Lelli
2019-07-27 5:56 ` [PATCH RT 6/8] sched: migrate_enable: Set state to TASK_RUNNING Scott Wood
2019-09-17 15:31 ` Sebastian Andrzej Siewior
2019-09-17 17:54 ` Scott Wood
2019-07-27 5:56 ` [PATCH RT 7/8] sched: migrate_enable: Use select_fallback_rq() Scott Wood
2019-09-17 16:00 ` Sebastian Andrzej Siewior
2019-09-24 18:05 ` Scott Wood
2019-07-27 5:56 ` [PATCH RT 8/8] sched: Lazy migrate_disable processing Scott Wood
2019-09-17 16:50 ` Sebastian Andrzej Siewior
2019-09-17 17:06 ` Scott Wood
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=20190917151039.ukplro66tzgxg6xt@linutronix.de \
--to=bigeasy@linutronix.de \
--cc=bristot@redhat.com \
--cc=juri.lelli@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rt-users@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=swood@redhat.com \
--cc=tglx@linutronix.de \
--cc=williams@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;
as well as URLs for NNTP newsgroup(s).