Re: [LAU] A surprisingly stupid RT priority question

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: <linux-audio-user@...>
Date: Monday, December 3, 2012 - 4:48 pm

--f46d043c821a341c0604cff583f4
Content-Type: text/plain; charset=UTF-8

Hi all,

Interesting thread, and thanks for the information.

I have a related question: I have an EEEPC 1000HE netbook (used to run Arch
but I got tired of having to compile, or failing to compile, some extra
pacakges) that runs Xubuntu, usually with XFCE of course, but sometimes
with Fluxbox. Also have a bigger laptop, and HP pavilion7, running Mint
with Maya Desktop (I also switch it to Fluxbox sometimes).

Anyway, it seems the EEE has suboptimal IRQ hardware settings for RT low
latency audio...doing less /proc/interrupts shows that there's a lot of
sharing, like the audio and video card being on the same IRQ. Indeed,
running Csound FLTK gui interfaces and moving them while processing the
csound orchestra sometimes makes pops and crackle dropouts, while the same
thing on the better IRQ set HP Pavilion shows flawless performance.

Even if I run an outboard USB audio interface (Lexicon Lambda) on the EEE,
there is still suboptimal USB Irq settings to contend with.

I've always suspected IRQ settings, which are unfortunately hard-wired in
laptops, to be profoundly influential in RT performance, but I wanted the
audio guru programmers to comment.

Thanks,
AKJ

On Sun, Dec 2, 2012 at 8:27 PM, Ken Restivo wrote:

> On Sun, Dec 02, 2012 at 11:03:58PM +0100, Pedro Lopez-Cabanillas wrote:

--
Aaron Krister Johnson
http://www.akjmusic.com
http://www.untwelve.org

--f46d043c821a341c0604cff583f4
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi all,Interesting thread, and thanks for the informati=
on.I have a related question: I have an EEEPC 100=
0HE netbook (used to run Arch but I got tired of having to compile, or fail=
ing to compile, some extra pacakges) that runs Xubuntu, usually with XFCE o=
f course, but sometimes with Fluxbox. Also have a bigger laptop, and HP pav=
ilion7, running Mint with Maya Desktop (I also switch it to Fluxbox sometim=
es).
Anyway, it seems the EEE has suboptimal IRQ hardware se=
ttings for RT low latency audio...doing less /proc/interrupts shows that th=
ere's a lot of sharing, like the audio and video card being on the same=
IRQ. Indeed, running Csound FLTK gui interfaces and moving them while proc=
essing the csound orchestra sometimes makes pops and crackle dropouts, whil=
e the same thing on the better IRQ set HP Pavilion shows flawless performan=
ce.
Even if I run an outboard USB audio interface (Lexicon Lambda) on the E=
EE, there is still suboptimal USB Irq settings to contend with.<=
br>I've always suspected IRQ settings, which are unfortunate=
ly hard-wired in laptops, to be profoundly influential in RT performance, b=
ut I wanted the audio guru programmers to comment.=C2=A0
Thanks,AKJOn Sun, Dec 2, 2012 at 8:27 PM, Ken Restivo <ken@restivo.org> wrote:
On Sun, Dec 02, 2012 at 11=
:03:58PM +0100, Pedro Lopez-Cabanillas wrote:

igged and
it seems
time.
process
me system
ing or
ACK sample
ething obvious
/a>
but
ctly

ffers +
ncy so

Thanks everyone for the fantastic and detailed answers! Now it is all=
starting to make sense to me.

Yes, sched_foo were the ones I was thinking of, SCHED_FIFO being the priori=
ty for Ingo-ish RT, IIRC.

And yes, it was the requirement for determinsm (no malloc, blocking system =
calls, etc) that I remember from JACK callbacks, and thanks for confirming =
that it is a general requirement of RT audio in general, not of JACK in per=
ticular.

Buffering indeed is what sounds like the main issue for Liquidsoap. =C2=A0B=
ut I have no way of knowing how much buffer would actually be necessary to =
assure it NEVER skips. This is why I was thinking along the lines of some k=
ind of preemption to assure that never happens.

rtkit looks interesting, but if there were a way to this without it, I'=
d prefer that.

Low latency is NOT required for this application, just stability when starv=
ed for resources. Liquidsoap spends almost all of of its time decoding MP3s=
from disk or reading and decoding MP3 streams off of the internet.

-ken
_____________________=
__________________________
Linux-audio-user mailing list
Linux-audio-user@l=
ists.linuxaudio.org

http://lists.linuxaudio.org/listinfo/linux-audio-user
-- =
Aaron Krister Johnsonhttp://www.akjmusic.comhttp://www.untwelve.org

--f46d043c821a341c0604cff583f4--

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[LAU] A surprisingly stupid RT priority question, Ken Restivo, (Sun Dec 2, 9:01 pm)
Re: [LAU] A surprisingly stupid RT priority question, Paul Coccoli, (Sat Dec 8, 7:20 pm)
Re: [LAU] A surprisingly stupid RT priority question, Pedro Lopez-Cabanillas, (Sun Dec 2, 10:04 pm)
Re: [LAU] A surprisingly stupid RT priority question, Ken Restivo, (Mon Dec 3, 2:27 am)
Re: [LAU] A surprisingly stupid RT priority question, Aaron Krister Johnson, (Mon Dec 3, 4:48 pm)
Re: [LAU] A surprisingly stupid RT priority question, Harry van Haaren, (Sun Dec 2, 9:23 pm)