Aymeric Mansoux said :
> Dan S said :
> > Hi -
> > 
> > I don't have firewire kit to test, I'm afraid, but maybe you could try
> > running the "latencytop" program (not installed by default, but
> > available in the repos) to detect the source of the xruns?
> 

me neither, can't really tell whats wrong without it :(

> sounds like this one should be added to the CD...
> 

yep true, I didn't even know it exists!

>  
> > It could be a question of IRQ priorities I imagine.
> 
> And/Or we need a more recent ALSA....
> 

I don't  think this uses ALSA  at all to be  honest, its a  firewire device that
uses the freebob  backend as far as  I understand. It could be  that between now
and versions that are in L&P and  Miso there have been regressions in the driver
code for  that device, so  one thing  one could try  is to backport  the freebob
driver from these versions to karmic and see if it works better.

> It would be good to try if a generic kernel + the ALSA backport modules
> package would make any difference.
> 
> @Karsten: did you check if we could get the ALSA backports working for
> the RT kernel we use? This would be really good to be able to keep this
> as fresh as possible.

I checked,  but its not trivial  as it requires quite  a lot of  tweaking in the
packaging  part  (it  also  uses  this  kernel  build  system  that  is  largely
undocumented (or at least not very  well documented to date). It should become a
priority  for  the next  sprint  to develop  a  kernel  engineering process  and
document it thoroughly.

> 
> @gusano: stupid test/question do you also have xruns when you run jackd
> from a console *without* X running? Regarding previous Puredyne
> releases, every download mirrors has them, just go to the parent
> directories and browse.




http://krgn.goto10.org --
tel :: +49 (0) 30.6233.475
mob :: +49 (0) 178.8540.102
pgp :: http://krgn.goto10.org/pubkey.asc

Attachment: signature.asc
Description: Digital signature

---
[email protected]
http://identi.ca/group/puredyne
irc://irc.goto10.org/puredyne

Reply via email to