mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Mathieu Desnoyers" <mathieu.desnoyers@efficios.com>
Cc: linux-rt-devel@lists.linux.dev, tj@kernel.org,
	hannes@cmpxchg.org, mkoutny@suse.com,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v4 09/26] sched/deadline: Reset the deadline and the runtime with the parameters
Date: Sat, 07 Nov 2026 02:18:50 +0000	[thread overview]
Message-ID: <sashiko-outbox-202497@kernel.org> (raw)
In-Reply-To: <20261107014451.1487825-10-mathieu.desnoyers@efficios.com>

Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [Medium] sched/deadline: unreset donor period in switched_from_dl()
  when dl_runtime != 0
- [Medium] sched/deadline: donor deadline and runtime kept on deboost
  after setattr
--

commit e10fc72d0029b6302955712a86b151e84624b177
Author: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

sched/deadline: Reset the deadline and the runtime with the parameters

This commit resets dl_se->deadline and dl_se->runtime in
__dl_clear_params() and in switched_from_dl() when a deboosted task has no
dl_runtime of its own.

> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
> index b8b0076870a64..f389620e6fc9a 100644
> --- a/kernel/sched/deadline.c
> +++ b/kernel/sched/deadline.c
> @@ -3580,6 +3580,16 @@ static void switched_from_dl(struct rq *rq, struct task_struct *p)
>  	if (task_on_rq_queued(p) && p->dl.dl_runtime)
>  		task_non_contending(&p->dl, false);
>  
> +	/*
> +	 * Deboosted from priority inheritance with no parameters of its own:
> +	 * the deadline and the runtime are the donor's. A change into the
> +	 * class then starts a period from the parameters it is given.
> +	 */
> +	if (!p->dl.dl_runtime) {
> +		p->dl.deadline = 0;
> +		p->dl.runtime = 0;
> +	}

[Severity: Medium]
What happens in switched_from_dl() when a task that still carries pre-0-lag
parameters (p->dl.dl_runtime != 0) is PI-boosted by a SCHED_DEADLINE waiter
and then deboosted?

When the task is boosted before its inactive_timer fires, switched_to_dl()
cancels inactive_timer and replenish_dl_entity() overwrites p->dl.deadline
and p->dl.runtime from the donor's pi_params.  On deboost, p->dl.dl_runtime
is non-zero, so switched_from_dl() skips the reset and passes the
donor-modified deadline and runtime to task_non_contending(), mixing them
with the task's old dl_period and dl_runtime.

Depending on the donor's parameters, this either releases the task's old
bandwidth before its real 0-lag time or leaves the donor's future deadline
and runtime in place for a subsequent entry into SCHED_DEADLINE.

[Severity: Medium]
Can a task keep the donor's deadline and runtime if sched_setattr() gives
it SCHED_DEADLINE parameters while it is PI-boosted?

If a non-deadline task is boosted into dl_sched_class and then configured
as SCHED_DEADLINE via sched_setattr() before unlocking the rt_mutex, both
__sched_setscheduler() and the subsequent deboost in rt_mutex_setprio()
keep the task in dl_sched_class without calling switched_from_dl():

kernel/sched/core.c:rt_mutex_setprio() {
    ...
		} else {
			p->dl.dl_boosted = 0;
		}
    ...
}

When rt_mutex_setprio() clears dl_boosted, enqueue_dl_entity() only calls
setup_new_dl_entity() if dl_se->deadline is already in the past, so the
task runs unboosted on the donor's remaining runtime and deadline instead
of starting a period from its own parameters.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261107014451.1487825-1-mathieu.desnoyers@efficios.com?part=9

  reply	other threads:[~2026-11-07  2:18 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-07  1:44 [PATCH v4 00/26] sched/deadline: bandwidth accounting fixes and selftests Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 01/26] sched: Fix cpuset_mutex double unlock in __sched_setscheduler() Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 02/26] sched: Refuse SCHED_FLAG_KEEP_PARAMS that would change the policy or the class Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 03/26] sched/deadline: Do not admit bandwidth for parameters that are kept Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 04/26] sched: Let a task boosted by a deadline waiter change its policy Mathieu Desnoyers
2026-11-07  2:13   ` sashiko-bot
2026-11-07  1:44 ` [PATCH v4 05/26] sched/deadline: Do not divide by zero active CPUs of a root domain Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 06/26] sched: Take cpuset_mutex for a deadline policy change only with pi Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 07/26] sched/deadline: Do not lose a throttled task switched out and back Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 08/26] sched/deadline: Take a deadline boost from the waiter's rt_mutex node Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 09/26] sched/deadline: Reset the deadline and the runtime with the parameters Mathieu Desnoyers
2026-11-07  2:18   ` sashiko-bot [this message]
2026-11-07  1:44 ` [PATCH v4 10/26] sched/deadline: Release the bandwidth of a task that leaves the class asleep Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 11/26] sched/deadline: Reject SCHED_DEADLINE on exited tasks Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 12/26] sched/deadline: Keep departing bandwidth across a root domain rebuild Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 13/26] sched/deadline: Do not take a waking task's bandwidth off a runqueue it left Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 14/26] sched/deadline: Rebuild the root domains' accounting from the task list Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 15/26] sched/deadline: Leave a dying task out of a rebuild of the accounting Mathieu Desnoyers
2026-11-07  2:10   ` sashiko-bot
2026-11-07  1:44 ` [PATCH v4 16/26] sched/deadline: Record the root domain holding a task's bandwidth Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 17/26] sched/deadline: Move a task's bandwidth only where room was made Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 18/26] cgroup/cpuset: Wait for an attach before rebuilding the sched domains Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 19/26] cgroup/cpuset: Give back the bandwidth an attach did not move Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 20/26] sched/deadline: Credit a rebuild to the CPU recorded with the task's bandwidth Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 21/26] sched/deadline: Cancel the replenishment timer with the parameters Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 22/26] sched/deadline: Count the bandwidth a boosted task is given Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 23/26] sched/debug: Show the deadline bandwidth of each runqueue Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 24/26] selftests/sched: Add a test of SCHED_DEADLINE policy round trips Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 25/26] selftests/sched: Test SCHED_DEADLINE bandwidth across a root domain rebuild Mathieu Desnoyers
2026-11-07  1:44 ` [PATCH v4 26/26] selftests/sched: Add a test of SCHED_DEADLINE bandwidth accounting Mathieu Desnoyers

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=sashiko-outbox-202497@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=hannes@cmpxchg.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rt-devel@lists.linux.dev \
    --cc=mathieu.desnoyers@efficios.com \
    --cc=mkoutny@suse.com \
    --cc=sashiko-reviews@lists.linux.dev \
    --cc=tj@kernel.org \
    /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

all inboxes | Powered by JetHome®