Csound Csound-dev Csound-tekno Search About

Linux: --sched=N in its own little world

Date2007-10-29 22:12
From"Chuckk Hubbard"
SubjectLinux: --sched=N in its own little world
AttachmentsNone  None  

Date2007-10-29 22:24
From"Chuckk Hubbard"
SubjectRe: Linux: --sched=N in its own little world
AttachmentsNone  None  

Date2007-10-29 22:27
From"Chuckk Hubbard"
SubjectRe: Linux: --sched=N in its own little world
AttachmentsNone  None  

Date2007-10-30 13:12
From"Chuckk Hubbard"
SubjectRe: Linux: --sched=N in its own little world
AttachmentsNone  None  

Date2008-01-06 14:58
FromAndres Cabrera
Subject[Csnd] Re: Linux: --sched=N in its own little world
Hi Chuckk,

Have you found out more about this issue? I have briefly tested, and
there is no problem using --sched with jack, but I'm not sure if it may
be disabled internally.

Cheers,
Andrés

El mar, 30-10-2007 a las 15:12 +0200, Chuckk Hubbard escribió:
> Yet again, I discover something strange.
> 
> From html/UsingRealTime.html
> "DO NOT use "--sched" if you are using JACK for audio output."
> 
> and later on the same page:
> "An example of buffer settings for low latency on a fast system:
> jackd -d alsa -P -r 48000 -p 64 -n 4 -zt &
> csound -+rtaudio=jack -b 64 -B 256 [...]
> 
> with real time scheduling (as root):
> jackd -R -P 90 -d alsa -P -r 48000 -p 64 -n 2 -zt &
> csound --sched=80,90,10 -d -+rtaudio=jack -b 64 -B 192 [...]"
> 
> So it says DO NOT use "--sched" with "-+rtaudio=jack", but has an
> example using "--sched=N,MAXCPU,TIME" with "-+rtaudio=jack". 
> I know this whole setup is not common knowledge, and even the
> documentation for the Linux and Jack ends of things is poor, but can
> anyone straighten this all out for me?
> 
> If I'm using Csound with Jack, can I use "--sched=N" or not? 
> And it was my understanding that Jack clients should have slightly
> LOWER priority than jackd; but in this example they are set the same.
> What gives?
> 
> -Chuckk
> 
> On 10/30/07, Chuckk Hubbard  wrote:
>         Hi.
>         Debian 64 testing, 2.6.22.1, rt patch, limits.conf edited for
>         rtprio.
>         So I'm running a test .csd with these flags: 
>         
>         -d -+rtaudio=jack -o dac:alsa_pcm:playback_ --sched=99 -b256
>         -B512 
>         
>         
>         and when I check the csound processes in htop I see one that
>         runs PRI 20 and one -45.  If I switch to --sched=67, I still
>         see 20 and -45.  If I run --sched=37, it's 20 and -45.  If I
>         run --sched with no value, it's 20 and -45. 
>         Am I not checking the priority correctly, or am I not entering
>         the flag correctly?
>         
>         Thanky.
>         -Chuckk
>         
>         -- 
>         http://www.badmuthahubbard.com
> 
> 
> 
> -- 
> http://www.badmuthahubbard.com


Date2008-01-26 08:01
From"Chuckk Hubbard"
Subject[Csnd] Re: Re: Linux: --sched=N in its own little world
AttachmentsNone  

Date2008-01-28 21:06
FromAndres Cabrera
Subject[Csnd] Re: Linux: --sched=N in its own little world
Hi,

Do you have a suggestion for changing the manual, or should we leave it
as it is in this regard?

Cheers,
Andrés

El sáb, 26-01-2008 a las 10:01 +0200, Chuckk Hubbard escribió:
> Hi Andres.
> 
> I have stopped using --sched altogether, and I believe Csound still
> inherits jackd's priority.  At least that's what I heard.  At any
> rate, I usually have no problems with jackd rt priority set and
> real-time Csound without --sched; and on my laptop, if it wasn't set
> to realtime, I would have problems...
> -Chuckk
> 
> On Jan 6, 2008 4:58 PM, Andres Cabrera  wrote:
> > Hi Chuckk,
> >
> > Have you found out more about this issue? I have briefly tested, and
> > there is no problem using --sched with jack, but I'm not sure if it may
> > be disabled internally.
> >
> > Cheers,
> > Andrés
> >
> > El mar, 30-10-2007 a las 15:12 +0200, Chuckk Hubbard escribió:
> >
> > > Yet again, I discover something strange.
> > >
> > > From html/UsingRealTime.html
> > > "DO NOT use "--sched" if you are using JACK for audio output."
> > >
> > > and later on the same page:
> > > "An example of buffer settings for low latency on a fast system:
> > > jackd -d alsa -P -r 48000 -p 64 -n 4 -zt &
> > > csound -+rtaudio=jack -b 64 -B 256 [...]
> > >
> > > with real time scheduling (as root):
> > > jackd -R -P 90 -d alsa -P -r 48000 -p 64 -n 2 -zt &
> > > csound --sched=80,90,10 -d -+rtaudio=jack -b 64 -B 192 [...]"
> > >
> > > So it says DO NOT use "--sched" with "-+rtaudio=jack", but has an
> > > example using "--sched=N,MAXCPU,TIME" with "-+rtaudio=jack".
> > > I know this whole setup is not common knowledge, and even the
> > > documentation for the Linux and Jack ends of things is poor, but can
> > > anyone straighten this all out for me?
> > >
> > > If I'm using Csound with Jack, can I use "--sched=N" or not?
> > > And it was my understanding that Jack clients should have slightly
> > > LOWER priority than jackd; but in this example they are set the same.
> > > What gives?
> > >
> > > -Chuckk
> > >
> > > On 10/30/07, Chuckk Hubbard  wrote:
> > >         Hi.
> > >         Debian 64 testing, 2.6.22.1, rt patch, limits.conf edited for
> > >         rtprio.
> > >         So I'm running a test .csd with these flags:
> > >         
> > >         -d -+rtaudio=jack -o dac:alsa_pcm:playback_ --sched=99 -b256
> > >         -B512
> > >         
> > >
> > >         and when I check the csound processes in htop I see one that
> > >         runs PRI 20 and one -45.  If I switch to --sched=67, I still
> > >         see 20 and -45.  If I run --sched=37, it's 20 and -45.  If I
> > >         run --sched with no value, it's 20 and -45.
> > >         Am I not checking the priority correctly, or am I not entering
> > >         the flag correctly?
> > >
> > >         Thanky.
> > >         -Chuckk
> > >
> > >         --
> > >         http://www.badmuthahubbard.com
> > >
> > >
> > >
> > > --
> > > http://www.badmuthahubbard.com
> >
> >
> >
> > Send bugs reports to this list.
> > To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> >
> 
> 
> 
> -- 
> http://www.badmuthahubbard.com
> 
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"