Re: [PATCH 1/3] sched, timer: Remove usages of ACCESS_ONCE in the scheduler

[Jason Low <jason.low2@hp.com>]

Hi Ingo,

On Wed, 2015-04-15 at 09:46 +0200, Ingo Molnar wrote:
> * Steven Rostedt <rostedt@goodmis.org> wrote:
> > You are correct. Now I'm thinking that the WRITE_ONCE() is not needed,
> > and just a:
> > 
> > 	p->mm->numa_scan_seq = READ_ONCE(p->numa_scan_seq) + 1;
> > 
> > Can be done. But I'm still trying to wrap my head around why this is
> > needed here. Comments would have been really helpful. We should make
> > all READ_ONCE() WRITE_ONCE and obsolete ACCESS_ONCE() have mandatory
> > comments just like we do with memory barriers.
> 
> So the original ACCESS_ONCE() barriers were misguided to begin with: I 
> think they tried to handle races with the scheduler balancing softirq 
> and tried to avoid having to use atomics for the sequence counter 
> (which would be overkill), but things like ACCESS_ONCE(x)++ never 
> guaranteed atomicity (or even coherency) of the update.
> 
> But since in reality this is only statistical sampling code, all these 
> compiler barriers can be removed I think. Peter, Mel, Rik, do you 
> agree?

Though in the read side for accessing things such as numa_scan_seq, we
still want to keep them in order to guarantee that numa_scan_seq is only
loaded once right?

static void task_numa_placement(struct task_struct *p)
{
    ...

    seq = ACCESS_ONCE(p->mm->numa_scan_seq);
    if (p->numa_scan_seq == seq)
        return;
    p->numa_scan_seq = seq;

    ...
}