From: Andrea Righi <arighi@nvidia.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: tj@kernel.org, linux-kernel@vger.kernel.org, mingo@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,
longman@redhat.com, hannes@cmpxchg.org, mkoutny@suse.com,
void@manifault.com, changwoo@igalia.com, cgroups@vger.kernel.org,
sched-ext@lists.linux.dev, liuwenfang@honor.com,
tglx@linutronix.de
Subject: Re: [PATCH 01/12] sched: Employ sched_change guards
Date: Tue, 7 Oct 2025 10:20:44 +0200 [thread overview]
Message-ID: <aOTNXPTyk4zth-1C@gpd4> (raw)
In-Reply-To: <20251006104526.613879143@infradead.org>
Hi Peter,
On Mon, Oct 06, 2025 at 12:44:03PM +0200, Peter Zijlstra wrote:
> As proposed a long while ago -- and half done by scx -- wrap the
> scheduler's 'change' pattern in a guard helper.
>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> Acked-by: Tejun Heo <tj@kernel.org>
> ---
...
> --- a/kernel/sched/sched.h
> +++ b/kernel/sched/sched.h
> @@ -3885,23 +3885,22 @@ extern void check_class_changed(struct r
> extern struct balance_callback *splice_balance_callbacks(struct rq *rq);
> extern void balance_callbacks(struct rq *rq, struct balance_callback *head);
>
> -#ifdef CONFIG_SCHED_CLASS_EXT
> -/*
> - * Used by SCX in the enable/disable paths to move tasks between sched_classes
> - * and establish invariants.
> - */
> -struct sched_enq_and_set_ctx {
Not necessarily for this patch, we can add it later, but I kinda liked the
comment that briefly explained how the context is used. Maybe having
something along these lines could be helpful?
/*
* Used to ensure the correct sequence of task state transitions, such as
* switching between sched_classes, changing CPU affinity, priority, or
* updating the queued/running state.
*/
> +struct sched_change_ctx {
> struct task_struct *p;
> - int queue_flags;
> + int flags;
> bool queued;
> bool running;
> };
>
> -void sched_deq_and_put_task(struct task_struct *p, int queue_flags,
> - struct sched_enq_and_set_ctx *ctx);
> -void sched_enq_and_set_task(struct sched_enq_and_set_ctx *ctx);
> +struct sched_change_ctx *sched_change_begin(struct task_struct *p, unsigned int flags);
> +void sched_change_end(struct sched_change_ctx *ctx);
>
> -#endif /* CONFIG_SCHED_CLASS_EXT */
> +DEFINE_CLASS(sched_change, struct sched_change_ctx *,
> + sched_change_end(_T),
> + sched_change_begin(p, flags),
> + struct task_struct *p, unsigned int flags)
> +
> +DEFINE_CLASS_IS_UNCONDITIONAL(sched_change)
>
> #include "ext.h"
>
> --- a/kernel/sched/syscalls.c
> +++ b/kernel/sched/syscalls.c
> @@ -64,7 +64,6 @@ static int effective_prio(struct task_st
>
> void set_user_nice(struct task_struct *p, long nice)
> {
> - bool queued, running;
> struct rq *rq;
> int old_prio;
>
> @@ -90,22 +89,12 @@ void set_user_nice(struct task_struct *p
> return;
> }
>
> - queued = task_on_rq_queued(p);
> - running = task_current_donor(rq, p);
> - if (queued)
> - dequeue_task(rq, p, DEQUEUE_SAVE | DEQUEUE_NOCLOCK);
> - if (running)
> - put_prev_task(rq, p);
> -
> - p->static_prio = NICE_TO_PRIO(nice);
> - set_load_weight(p, true);
> - old_prio = p->prio;
> - p->prio = effective_prio(p);
> -
> - if (queued)
> - enqueue_task(rq, p, ENQUEUE_RESTORE | ENQUEUE_NOCLOCK);
> - if (running)
> - set_next_task(rq, p);
> + scoped_guard (sched_change, p, DEQUEUE_SAVE | DEQUEUE_NOCLOCK) {
> + p->static_prio = NICE_TO_PRIO(nice);
> + set_load_weight(p, true);
> + old_prio = p->prio;
> + p->prio = effective_prio(p);
> + }
>
> /*
> * If the task increased its priority or is running and
> @@ -515,7 +504,7 @@ int __sched_setscheduler(struct task_str
> bool user, bool pi)
> {
> int oldpolicy = -1, policy = attr->sched_policy;
> - int retval, oldprio, newprio, queued, running;
> + int retval, oldprio, newprio;
> const struct sched_class *prev_class, *next_class;
> struct balance_callback *head;
> struct rq_flags rf;
> @@ -698,33 +687,25 @@ int __sched_setscheduler(struct task_str
> if (prev_class != next_class && p->se.sched_delayed)
> dequeue_task(rq, p, DEQUEUE_SLEEP | DEQUEUE_DELAYED | DEQUEUE_NOCLOCK);
>
> - queued = task_on_rq_queued(p);
> - running = task_current_donor(rq, p);
> - if (queued)
> - dequeue_task(rq, p, queue_flags);
> - if (running)
> - put_prev_task(rq, p);
> -
> - if (!(attr->sched_flags & SCHED_FLAG_KEEP_PARAMS)) {
> - __setscheduler_params(p, attr);
> - p->sched_class = next_class;
> - p->prio = newprio;
> - }
> - __setscheduler_uclamp(p, attr);
> - check_class_changing(rq, p, prev_class);
> + scoped_guard (sched_change, p, queue_flags) {
>
> - if (queued) {
> - /*
> - * We enqueue to tail when the priority of a task is
> - * increased (user space view).
> - */
> - if (oldprio < p->prio)
> - queue_flags |= ENQUEUE_HEAD;
> + if (!(attr->sched_flags & SCHED_FLAG_KEEP_PARAMS)) {
> + __setscheduler_params(p, attr);
> + p->sched_class = next_class;
> + p->prio = newprio;
> + }
> + __setscheduler_uclamp(p, attr);
> + check_class_changing(rq, p, prev_class);
>
> - enqueue_task(rq, p, queue_flags);
> + if (scope->queued) {
> + /*
> + * We enqueue to tail when the priority of a task is
> + * increased (user space view).
> + */
> + if (oldprio < p->prio)
> + scope->flags |= ENQUEUE_HEAD;
> + }
> }
> - if (running)
> - set_next_task(rq, p);
>
> check_class_changed(rq, p, prev_class, oldprio);
>
>
>
Thanks,
-Andrea
next prev parent reply other threads:[~2025-10-07 8:21 UTC|newest]
Thread overview: 74+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-10-06 10:44 [PATCH 00/12] sched: Cleanup the change-pattern and related locking Peter Zijlstra
2025-10-06 10:44 ` [PATCH 01/12] sched: Employ sched_change guards Peter Zijlstra
2025-10-07 8:20 ` Andrea Righi [this message]
2025-10-08 6:51 ` Peter Zijlstra
2025-10-08 6:58 ` Andrea Righi
2025-10-07 16:58 ` Valentin Schneider
2025-10-08 14:02 ` Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 02/12] sched: Re-arrange the {EN,DE}QUEUE flags Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 03/12] sched: Fold sched_class::switch{ing,ed}_{to,from}() into the change pattern Peter Zijlstra
2025-10-09 13:30 ` Dietmar Eggemann
2025-10-09 13:54 ` Peter Zijlstra
2025-10-09 14:09 ` Peter Zijlstra
2025-10-09 16:50 ` Dietmar Eggemann
2025-10-13 10:23 ` Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] sched/deadline: Prepare for switched_from() change tip-bot2 for Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] sched: Fold sched_class::switch{ing,ed}_{to,from}() into the change pattern tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 04/12] sched: Cleanup sched_delayed handling for class switches Peter Zijlstra
2025-10-07 15:22 ` Vincent Guittot
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 05/12] sched: Move sched_class::prio_changed() into the change pattern Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2026-01-12 20:44 ` [PATCH 05/12] " Pierre Gondois
2026-01-13 4:12 ` K Prateek Nayak
2026-01-13 10:45 ` Pierre Gondois
2026-01-13 11:05 ` K Prateek Nayak
2026-01-13 11:53 ` Peter Zijlstra
2026-01-13 11:56 ` Peter Zijlstra
2026-01-13 13:07 ` Pierre Gondois
2026-01-13 13:10 ` Pierre Gondois
2026-01-15 21:01 ` [tip: sched/urgent] sched/deadline: Avoid double update_rq_clock() tip-bot2 for Peter Zijlstra
2026-01-13 11:47 ` [PATCH 05/12] sched: Move sched_class::prio_changed() into the change pattern Peter Zijlstra
2026-01-14 6:47 ` K Prateek Nayak
2026-01-14 10:23 ` Peter Zijlstra
2026-01-14 13:05 ` Peter Zijlstra
2026-01-14 14:04 ` luca abeni
2026-01-14 14:20 ` Juri Lelli
2026-01-14 15:25 ` luca abeni
2026-01-15 8:24 ` Peter Zijlstra
2026-01-15 9:05 ` Peter Zijlstra
2026-01-15 13:13 ` Pierre Gondois
2026-01-15 13:56 ` Juri Lelli
2026-01-15 21:00 ` [tip: sched/urgent] sched/deadline: Use ENQUEUE_MOVE to allow priority change tip-bot2 for Peter Zijlstra
2026-01-15 21:00 ` [tip: sched/urgent] sched: Deadline has dynamic priority tip-bot2 for Peter Zijlstra
2026-01-15 21:01 ` [tip: sched/urgent] sched: Audit MOVE vs balance_callbacks tip-bot2 for Peter Zijlstra
2026-01-15 21:01 ` [tip: sched/urgent] sched: Fold rq-pin swizzle into __balance_callbacks() tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 06/12] sched: Fix migrate_disable_switch() locking Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 07/12] sched: Fix do_set_cpus_allowed() locking Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-24 14:58 ` [REGRESSION] Deadlock during CPU hotplug caused by abfc01077df6 Jan Polensky
2025-10-06 10:44 ` [PATCH 08/12] sched: Rename do_set_cpus_allowed() Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 09/12] sched: Make __do_set_cpus_allowed() use the sched_change pattern Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 10/12] sched: Add locking comments to sched_class methods Peter Zijlstra
2025-10-07 9:54 ` Juri Lelli
2025-10-08 7:04 ` Peter Zijlstra
2025-10-08 7:33 ` Greg Kroah-Hartman
2025-10-08 9:43 ` Juri Lelli
2025-10-08 10:06 ` Greg Kroah-Hartman
2025-10-08 14:34 ` Steven Rostedt
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 11/12] sched: Match __task_rq_{,un}lock() Peter Zijlstra
2025-10-07 20:44 ` Tejun Heo
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-06 10:44 ` [PATCH 12/12] sched: Cleanup the sched_change NOCLOCK usage Peter Zijlstra
2025-10-16 9:33 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2025-10-07 8:25 ` [PATCH 00/12] sched: Cleanup the change-pattern and related locking Andrea Righi
2025-10-07 9:55 ` Juri Lelli
2025-10-07 15:23 ` Vincent Guittot
2025-10-07 20:46 ` Tejun Heo
2025-10-08 13:54 ` Valentin Schneider
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=aOTNXPTyk4zth-1C@gpd4 \
--to=arighi@nvidia.com \
--cc=bsegall@google.com \
--cc=cgroups@vger.kernel.org \
--cc=changwoo@igalia.com \
--cc=dietmar.eggemann@arm.com \
--cc=hannes@cmpxchg.org \
--cc=juri.lelli@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=liuwenfang@honor.com \
--cc=longman@redhat.com \
--cc=mgorman@suse.de \
--cc=mingo@kernel.org \
--cc=mkoutny@suse.com \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=sched-ext@lists.linux.dev \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
--cc=vincent.guittot@linaro.org \
--cc=void@manifault.com \
--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