High res timer?

High res timer?

[Andre Eisenbach]

Hello!

We're currently using pth_usleep() as a timer for a userspace audio
application. However, it doesn't seem very accurate and reliable. Is
there a better (more accurate) timer that we can call form a userspace
application?

Thanks,
    Andre

Re: High res timer?

[Lee Revell]

On Tue, 2005-05-10 at 13:27 -0700, Andre Eisenbach wrote:
> Hello!
> 
> We're currently using pth_usleep() as a timer for a userspace audio
> application. However, it doesn't seem very accurate and reliable. Is
> there a better (more accurate) timer that we can call form a userspace
> application?

Why don't you just clock off the sound card?

Lee

Re: High res timer?

[Lee Revell]

On Tue, 2005-05-10 at 13:27 -0700, Andre Eisenbach wrote:
> Hello!
> 
> We're currently using pth_usleep() as a timer for a userspace audio
> application. However, it doesn't seem very accurate and reliable. Is
> there a better (more accurate) timer that we can call form a userspace
> application?
> 

Also, this is a bit OT for LKML.  Try linux-audio-dev at
music.columbia.edu.  You can probably find several good ways to do it in
the archives. 

Lee

Re: High res timer?

[Matthias-Christian Ott]

Andre Eisenbach wrote:
> Hello!
> 
> We're currently using pth_usleep() as a timer for a userspace audio
> application. However, it doesn't seem very accurate and reliable. Is
> there a better (more accurate) timer that we can call form a userspace
> application?
> 
> Thanks,
>     Andre
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 
What about nanosleep ()?

Matthias-Christian

Re: High res timer?

[Andre Eisenbach]

2005/5/10, Matthias-Christian Ott <matthias.christian@tiscali.de>:
>
> What about nanosleep ()?
> 

nanosleep() seems to have some latency very similar to usleep(). Isn't
usleep based on nanosleep()?

Here's what I get if I try to nanosleep for 5 secs (for testing):

-> 5.009952 s

The .009952 part varies, but is very close to that usually.

Thanks,
    Andre

Re: High res timer?

[Andre Eisenbach]

2005/5/10, Lee Revell <rlrevell@joe-job.com>:
> On Tue, 2005-05-10 at 13:27 -0700, Andre Eisenbach wrote:
> 
> Why don't you just clock off the sound card?
> 

I'll look into this.

Oh, and sorry for the off-topic :(

Thanks,
    Andre

Re: High res timer?

[Lee Revell]

On Tue, 2005-05-10 at 14:16 -0700, Andre Eisenbach wrote:
> 2005/5/10, Matthias-Christian Ott <matthias.christian@tiscali.de>:
> >
> > What about nanosleep ()?
> > 
> 
> nanosleep() seems to have some latency very similar to usleep(). Isn't
> usleep based on nanosleep()?
> 
> Here's what I get if I try to nanosleep for 5 secs (for testing):
> 
> -> 5.009952 s
> 
> The .009952 part varies, but is very close to that usually.

Is this a 2.4 kernel?  The resolution on 2.6 should be 1ms, not ~10ms.

Lee

Re: High res timer?

[Andre Eisenbach]

2005/5/10, Lee Revell <rlrevell@joe-job.com>:
> > Here's what I get if I try to nanosleep for 5 secs (for testing):
> >
> > -> 5.009952 s
> >
> > The .009952 part varies, but is very close to that usually.
> 
> Is this a 2.4 kernel?  The resolution on 2.6 should be 1ms, not ~10ms.

Yes indeed, that was a 2.4 kernel. On 2.6 (different machine though) I get:

-> 5.001064 s

Thanks,
    Andre

Re: High res timer?

[George Anzinger]

Andre Eisenbach wrote:
> Hello!
> 
> We're currently using pth_usleep() as a timer for a userspace audio
> application. However, it doesn't seem very accurate and reliable. Is
> there a better (more accurate) timer that we can call form a userspace
> application?

You don't say what resolution you want, but to ~ 1 ms, nanosleep or 
clock_nanosleep should work.  If you need better resolution, check out the HRT 
patch, see sig. below.

-- 
George Anzinger   george@mvista.com
High-res-timers:  http://sourceforge.net/projects/high-res-timers/