On Tue, 26 Jan 2016, Prarit Bhargava wrote:
> On 01/26/2016 03:34 PM, Thomas Gleixner wrote:
> > Just for the record: The fast timekeeper stops across the inner workings of
> > suspend/resume. But I think that doesn't matter much.
>
> I've had to debug s/r before using the sysfs interface. It is nice to have
> that
> with a proper timestamp because a lot of the failures are in the driver
> suspend
> and resume paths. I could easily change the options to:
The timekeeper suspends/resumes in the syscore section, so the driver part is
not affected.
> 0 = disable
> 1 = local clock (which along with 0 would maintain the current state of
> things)
> 2 = monotonic
> 3 = real
Yes, that works.
Btw, you need to document, that REAL can occasionally provide a wrong value on
32bit machines in case that a concurrent update of the realtime offset happens
(settimeofday, leap seconds ...). That's because we access real_offset w/o the
seqcount protection. It's probably hard to achieve, but it might happen.
Thanks,
tglx
