Re: [PATCH v8 27/33] sched/deadline: Start a period at a deboost out of the class
From: Mathieu Desnoyers
Date: Fri Nov 20 2026 - 08:55:16 EST
> Does task_non_contending() still arm a 0-ns inactive_timer instead of
> releasing the bandwidth at once when dl_deadline == dl_period?
Yes. v9 will test zerolag_time <= 0, so the bandwidth will be released
at once.
> What happens if p is asleep with p->dl.dl_non_contending already set when
> it is deboosted out of SCHED_DEADLINE?
The old timer stays armed. v7.4-rc2 has the same gap, and this patch did
not cover it. The task keeps its bandwidth for up to about the donor's
period.
In v9, switched_from_dl() will cancel that timer when the boost refilled
the runtime. task_non_contending() will then compute the 0-lag time from
the task's own period. If the timer is already running, its handler
releases the bandwidth. A new case, deboost_out_asleep_after_refill,
fails without the fix and passes with it.
Thanks,
Mathieu
--
Mathieu Desnoyers
EfficiOS Inc.
https://www.efficios.com