DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 4AE2D81F03
Date: Tue, 28 Mar 2017 17:26:01 -0400
From: Luiz Capitulino <lcapitulino@redhat.com>
To: Rik van Riel <riel@redhat.com>
Cc: Wanpeng Li <kernellwp@gmail.com>,
        Frederic Weisbecker <fweisbec@gmail.com>, linux-kernel@vger.kernel.org
Subject: Re: [BUG nohz]: wrong user and system time accounting
Message-ID: <20170328172601.4d17256c@redhat.com>
In-Reply-To: <1490734912.8850.85.camel@redhat.com>
References: <20170323165512.60945ac6@redhat.com>
        <CANRm+CxcgSP2-x+A822DmHLvFLzFmTptS6oYwYtwVdErTpiB=Q@mail.gmail.com>
        <1490636129.8850.76.camel@redhat.com>
        <20170328132406.7d23579c@redhat.com>
        <20170328161454.4a5d9e8b@redhat.com>
        <1490734912.8850.85.camel@redhat.com>
Organization: Red Hat
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1732
Lines: 40

On Tue, 28 Mar 2017 17:01:52 -0400
Rik van Riel <riel@redhat.com> wrote:

> On Tue, 2017-03-28 at 16:14 -0400, Luiz Capitulino wrote:
> > On Tue, 28 Mar 2017 13:24:06 -0400
> > Luiz Capitulino <lcapitulino@redhat.com> wrote:  
> > > I'm starting to suspect that the nohz code may be programming
> > > the tick period to be shorter than 1ms when it re-activates
> > > the tick.  
> > 
> > And I think I was right, it looks like the nohz code is programming
> > the tick period incorrectly when restarting the tick. The patch below
> > fixes things for me, but I still have some homework todo and more
> > testing before posting a patch for inclusion. Could you guys test it?  
> 
> Your patch seems to work. I don't claim to understand why
> your patch makes a difference, but for this particular test
> case, on this particular setup, it seems to work...

I don't fully understand why either yet. I was looking for places
where nohz might be programming the tick period incorrectly and
I found that there's a case in tick_nohz_stop_sched_tick() where
tick_nohz_restart() is called only to reprogram the tick timer,
not cancel the tick. In this case, ts->last_tick seems to be out
of date. Fixing this fixed accounting for me.

> > diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
> > index 7fe53be..9abe979 100644
> > --- a/kernel/time/tick-sched.c
> > +++ b/kernel/time/tick-sched.c
> > @@ -1152,6 +1152,7 @@ static enum hrtimer_restart
> > tick_sched_timer(struct hrtimer *timer)
> >         struct pt_regs *regs = get_irq_regs();
> >         ktime_t now = ktime_get();
> >  
> > +       ts->last_tick = now;
> >         tick_sched_do_timer(now);
> >  
> >         /*  
>