[LAU] A surprisingly stupid RT priority question

[Paul Davis]

simple answer: non-RT things need more buffering between them and the
hardware. find out how to do that.


On Sun, Dec 2, 2012 at 4:00 PM, Ken Restivo <ken at restivo.org> wrote:

> OK, I know I've been using Linux audio for 6 years now, and gigged and
> recorded with it extensively for most of those, yadda yadda. But it seems
> I've had an embarassingly huge hole in my knowledge the whole time.
>
> I was under the impression that, in order to use real time
> priorities/permissions and Ingo kernels, it was required for the process
> ITSELF early in the main() routine of its source code, to make some system
> calls to claim RT priority. In fact, I specifically remember reading or
> even writing source code in C which did that (probably based on JACK sample
> code). I don't recall the name of the syscall, but it was something obvious
> and well-documented.
>
> Also, I remember that it was also required that the software behave
> properly in order to be real-time capable, something about callbacks taking
> some predictable amount of time. Or perhaps that was only a JACK
> requirement.
>
> Why am I asking these dumb-ass questions now? Because I've been playing
> around with Liquidsoap and Airtime for some radio stations, and I'm
> obsessed with getting them as rock-solid on cheap/free/old hardware as I'd
> been able to get with my gigging and studio synths.
>
> It physically pains me to hear audio stuttering because Apache is running
> on the same box. It seems an outrage to me. Shouldn't happen. Ever. I used
> to record and mix multi-track songs in Ardour with tons of soft-synths
> WHILE A KERNEL COMPILE WAS GOING ON THE SAME MACHINE without a single
> glitch. I expect no less.
>
> So I asked on the Liquidsoap list, and I got only shrugs and a pointer to
> the Gentoo page in response (why? I have no idea. I use Debian, and that's
> irrelevant to the question at hand anyway.).
>
> So what's the deal? Is there a way to give Ingo-approved preemptive RT
> priority to things that aren't real-time apps and aren't specifically
> architected for that? What, if anything, would break?
>
> And, if I wanted to hack Liquidsoap (which would require learning ML,
> which wouldn't be a bad thing to know anyway), is it even possible to get
> it RT-capable, or are there low-level C system calls required in order to
> make that work?
>
> Sorry for the long and obscure question, but it's been bothering me for a
> while, and I figured someone here would know the answer, or where I might
> find it.
>
> Thanks.
>
>
> -ken
> _______________________________________________
> Linux-audio-user mailing list
> Linux-audio-user at lists.linuxaudio.org
> http://lists.linuxaudio.org/listinfo/linux-audio-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxaudio.org/pipermail/linux-audio-user/attachments/20121202/1f3f4546/attachment.html>