clock_getres(CLOCK_REALTIME, .) may return an outdated and too high resolution

Corinna Vinschen corinna-cygwin@cygwin.com
Wed Mar 28 07:40:00 GMT 2012


On Mar 27 23:28, Christian Franke wrote:
> Corinna Vinschen wrote:
> >On Mar 27 20:59, Christian Franke wrote:
> >>>>Sorry, I probably forgot to mention that NtSetTimerResolution
> >>>>returns the same useless actual value than NtQueryTimerResolution.
> >>>>
> >>>>I would suggest:
> >>>>
> >>>>     status = NtSetTimerResolution (period, TRUE,&actual);
> >>>>     if (!NT_SUCCESS (status))
> >>>>       { ... return -1; }
> >>>>  -  minperiod = actual;
> >>>>  +  minperiod = period;
> >>>But that's not right.  The "actual" value is not useless, but the value
> >>>the resolution has actually been set to.
> >>No, again this is the minimum of all resolutions currently set by
> >>all processes.
> >>
> >>
> >>>   The OS just doesn't support
> >>>arbitrary values for the period.
> >>>
> >>Yes - but in 'actual' a smaller value than the value set for the
> >>current process may be returned.
> >Hmpf, ok.  Boy is that ugly.
> 
> Yes, aka broken by design :-)
> 
> There should be a function to query the actual setting for the
> current process only. But there is none.
> 
> 
> >   Is there a chance that actual is bigger
> >than period?  In that case we should perhaps set minperiod like this:
> >
> >   minperiod = MAX (actual, period);
> 
> According to some experiments this can only happen if period <
> finest. In this case NtSetTimerResolution also succeeds and returns
> actual == finest.

Since period < finest can't happen, I set minperiod to period now as
you suggested.  But, frankly, this would have been much simpler if
you would have send a patch right from the start ;)


Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple



More information about the Cygwin mailing list