Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 28 Apr 2022 12:00:52 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     "Eric W. Biederman" <ebiederm@xmission.com>
Cc:     rjw@rjwysocki.net, oleg@redhat.com, mingo@kernel.org,
        vincent.guittot@linaro.org, dietmar.eggemann@arm.com,
        rostedt@goodmis.org, mgorman@suse.de, bigeasy@linutronix.de,
        Will Deacon <will@kernel.org>, linux-kernel@vger.kernel.org,
        tj@kernel.org, linux-pm@vger.kernel.org
Subject: Re: [PATCH v2 1/5] sched,signal,ptrace: Rework TASK_TRACED,
 TASK_STOPPED state
Message-ID: <Ympl1D11gXAxF0s4@hirez.programming.kicks-ass.net>
References: <20220421150248.667412396@infradead.org>
 <20220421150654.757693825@infradead.org>
 <87fslzfmha.fsf@email.froward.int.ebiederm.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87fslzfmha.fsf@email.froward.int.ebiederm.org>
Precedence: bulk

On Tue, Apr 26, 2022 at 06:34:09PM -0500, Eric W. Biederman wrote:
> Peter Zijlstra <peterz@infradead.org> writes:
> 
> > Currently ptrace_stop() / do_signal_stop() rely on the special states
> > TASK_TRACED and TASK_STOPPED resp. to keep unique state. That is, this
> > state exists only in task->__state and nowhere else.
> >
> > There's two spots of bother with this:
> >
> >  - PREEMPT_RT has task->saved_state which complicates matters,
> >    meaning task_is_{traced,stopped}() needs to check an additional
> >    variable.
> >
> >  - An alternative freezer implementation that itself relies on a
> >    special TASK state would loose TASK_TRACED/TASK_STOPPED and will
> >    result in misbehaviour.
> >
> > As such, add additional state to task->jobctl to track this state
> > outside of task->__state.
> >
> > NOTE: this doesn't actually fix anything yet, just adds extra state.
> >
> > Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> 
> > --- a/kernel/signal.c
> > +++ b/kernel/signal.c
> > @@ -770,7 +773,9 @@ void signal_wake_up_state(struct task_st
> >  	 * By using wake_up_state, we ensure the process will wake up and
> >  	 * handle its death signal.
> >  	 */
> > -	if (!wake_up_state(t, state | TASK_INTERRUPTIBLE))
> > +	if (wake_up_state(t, state | TASK_INTERRUPTIBLE))
> > +		t->jobctl &= ~(JOBCTL_STOPPED | JOBCTL_TRACED);
> > +	else
> >  		kick_process(t);
> >  }
> 
> This hunk is subtle and I don't think it is actually what we want if the
> code is going to be robust against tsk->__state becoming TASK_FROZEN.

Oooh, indeed. Yes, let me go back to that resume based thing as you
suggest.

But first, let me go read all your patches :-)