[LAD] Timers, hpet, hrtimer

[fons at kokkinizita.net]

On Tue, Nov 10, 2009 at 01:16:52PM -0500, Paul Davis wrote:

> On Tue, Nov 10, 2009 at 1:06 PM,  <fons at kokkinizita.net> wrote:
> > On Tue, Nov 10, 2009 at 07:53:19PM +0200, Jussi Laako wrote:
> >
> >> Brokenness is that JACK is not interested on those sub-timers and the
> >> driver should allocate those only via corresponding ioctl() and give
> >> unlimited read access to the global timer value without failing early...
> >
> > If hpet services are available via the kernel (I assume in the
> > form of fine-grained sleep, usleep, timetouts, etc.) why should
> > Jack (or any app) use the HW timer directly ?
> 
> because we don't want to have to make a system call to get a high res
> timer value.
> JACK was written with the assumption that doing this is essentially free.

So you're just reading it, not waiting on it ? That indeed
should be 'free'. 

What I don't see is how e.g. MIDI generators can use it
in this way to get accurate event timing. Either they are
spinning (evil), or they'd need a system call anyway, and
it that case they should use kernel services instead.

Ciao,

-- 
FA

Io lo dico sempre: l'Italia è troppo stretta e lunga.