[PATCH v7 16/22] sched: Defer wakeup in ttwu() for unschedulable frozen tasks

[Will Deacon]

On Thu, May 27, 2021 at 04:10:16PM +0200, Peter Zijlstra wrote:
> On Tue, May 25, 2021 at 04:14:26PM +0100, Will Deacon wrote:
> > diff --git a/kernel/freezer.c b/kernel/freezer.c
> > index dc520f01f99d..8f3d950c2a87 100644
> > --- a/kernel/freezer.c
> > +++ b/kernel/freezer.c
> > @@ -11,6 +11,7 @@
> >  #include <linux/syscalls.h>
> >  #include <linux/freezer.h>
> >  #include <linux/kthread.h>
> > +#include <linux/mmu_context.h>
> >  
> >  /* total number of freezing conditions in effect */
> >  atomic_t system_freezing_cnt = ATOMIC_INIT(0);
> > @@ -146,9 +147,16 @@ bool freeze_task(struct task_struct *p)
> >  void __thaw_task(struct task_struct *p)
> >  {
> >  	unsigned long flags;
> > +	const struct cpumask *mask = task_cpu_possible_mask(p);
> >  
> >  	spin_lock_irqsave(&freezer_lock, flags);
> > -	if (frozen(p))
> > +	/*
> > +	 * Wake up frozen tasks. On asymmetric systems where tasks cannot
> > +	 * run on all CPUs, ttwu() may have deferred a wakeup generated
> > +	 * before thaw_secondary_cpus() had completed so we generate
> > +	 * additional wakeups here for tasks in the PF_FREEZER_SKIP state.
> > +	 */
> > +	if (frozen(p) || (frozen_or_skipped(p) && mask != cpu_possible_mask))
> >  		wake_up_process(p);
> >  	spin_unlock_irqrestore(&freezer_lock, flags);
> >  }
> > diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> > index 42e2aecf087c..6cb9677d635a 100644
> > --- a/kernel/sched/core.c
> > +++ b/kernel/sched/core.c
> > @@ -3529,6 +3529,19 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
> >  	if (!(p->state & state))
> >  		goto unlock;
> >  
> > +#ifdef CONFIG_FREEZER
> > +	/*
> > +	 * If we're going to wake up a thread which may be frozen, then
> > +	 * we can only do so if we have an active CPU which is capable of
> > +	 * running it. This may not be the case when resuming from suspend,
> > +	 * as the secondary CPUs may not yet be back online. See __thaw_task()
> > +	 * for the actual wakeup.
> > +	 */
> > +	if (unlikely(frozen_or_skipped(p)) &&
> > +	    !cpumask_intersects(cpu_active_mask, task_cpu_possible_mask(p)))
> > +		goto unlock;
> > +#endif
> > +
> >  	trace_sched_waking(p);
> >  
> >  	/* We're going to change ->state: */
> 
> OK, I really hate this. This is slowing down the very hot wakeup path
> for the silly freezer that *never* happens. Let me try and figure out if
> there's another option.

I'm definitely open to alternative suggestions here. An easy thing to do
would be move the 'flags' field up in task_struct so that the previous
access to state pulls it in for us.

Will