linux-kernel - Re: [RFC PATCH 1/3] sched/core: Provide a method to check if a task is PI-boosted.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <ZUSdLMQn0GZ+/T1E@localhost.localdomain>
Date:   Fri, 3 Nov 2023 08:11:40 +0100
From:   Juri Lelli <juri.lelli@...hat.com>
To:     Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Cc:     linux-kernel@...r.kernel.org, Ben Segall <bsegall@...gle.com>,
        Daniel Bristot de Oliveira <bristot@...hat.com>,
        Dietmar Eggemann <dietmar.eggemann@....com>,
        Frederic Weisbecker <frederic@...nel.org>,
        Ingo Molnar <mingo@...hat.com>,
        John Stultz <jstultz@...gle.com>, Mel Gorman <mgorman@...e.de>,
        Peter Zijlstra <peterz@...radead.org>,
        Stephen Boyd <sboyd@...nel.org>,
        Steven Rostedt <rostedt@...dmis.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Valentin Schneider <vschneid@...hat.com>,
        Vincent Guittot <vincent.guittot@...aro.org>
Subject: Re: [RFC PATCH 1/3] sched/core: Provide a method to check if a task
 is PI-boosted.

On 02/11/23 17:03, Sebastian Andrzej Siewior wrote:
> On 2023-11-02 10:30:50 [+0100], Juri Lelli wrote:
> > Hi,
> Hi,
> 
> > Apologies for noticing only now, but I believe this is still part of the
> > 6.6-rt patches and I've got the below question to ask.
> 
> I'm not sure if we want this. I do have an alternative solution in the
> meantime.

Ah, OK. Guess what below is moot then. :) Will be watching for the
alternative solution to show up.
> 
> > > --- a/kernel/sched/core.c
> > > +++ b/kernel/sched/core.c
> > > @@ -8886,6 +8886,21 @@ static inline void preempt_dynamic_init(void) { }
> > >  
> > >  #endif /* #ifdef CONFIG_PREEMPT_DYNAMIC */
> > >  
> > > +/*
> > > + * task_is_pi_boosted - Check if task has been PI boosted.
> > > + * @p:	Task to check.
> > > + *
> > > + * Return true if task is subject to priority inheritance.
> > > + */
> > > +bool task_is_pi_boosted(const struct task_struct *p)
> > > +{
> > > +	int prio = p->prio;
> > > +
> > > +	if (!rt_prio(prio))
> > > +		return false;
> > > +	return prio != p->normal_prio;
> > 
> > Does this need to also take DEADLINE tasks into consideration? We don't
> > change priority when they are boosted, only pi_se changes (please check
> > is_dl_boosted()).
> 
> If we want this, then probably yes.
> Isn't task_struct::prio for the DL (boosted) task set to 0? While the RT
> priority go from 1…MAX_RT_PRIO - 1.

But then we can also have DL on DL boosting and in this case only the
pi_se changes while prio stays the same. But, again, looks like this is
moot anyway.

Thanks!
Juri