Re: [PATCH v3] alarmtimer: fix unexpected rtc interrupt when system resume from S3

[Thomas Gleixner <tglx@linutronix.de>]

On Fri, 20 Nov 2015, John Stultz wrote:

> On Tue, Nov 17, 2015 at 4:08 AM,  <zhuo-hao.lee@intel.com> wrote:
> > From: zhuo-hao <zhuo-hao.lee@intel.com>
> >
> > Before the system go to suspend (S3), if user create a timer with clockid
> > CLOCK_REALTIME_ALARM/CLOCK_BOOTTIME_ALARM and set a "large" timeout value
> > to this timer. The function alarmtimer_suspend will be called to setup
> > a timeout value to RTC timer to avoid the system sleep over time. However,
> > if the system wakeup early than RTC timeout, the RTC timer will not be cleared.
> > And this will cause the hpet_rtc_interrupt come unexpectedly until the RTC
> > timeout. To fix this problem, just adding alarmtimer_resume to cancel the
> > RTC timer.
> >
> > This was noticed because the HPET RTC emulation fires an interrupt every
> > 16ms(=1/2^DEFAULT_RTC_SHIFT) up to the point where the alarm time is reached.
> > This program always hits this situation(https://lkml.org/lkml/2015/11/8/326),
> > if system wake up earlier than alarm time.
> 
> So thanks for the extra context here, and again I don't have an
> objection to this patch.
> 
> Although from the earlier discussion it still isn't quite clear to me:
> Why must the HPET RTC emulation need to fire the alarm every 16ms? Is
> that not something that can be fixed?

We probably can fix it with some surgery. OTOH that stuff is fragile
as hell and I rather avoid touching it, but I won't hinder someone
brave enough doing it :)
 
> I just want to make sure we're not hiding a deeper issue.

It's not a deeper issue. It's a - admittedly dumb - implementation
detail.

Thanks,

	tglx