11 Nov
2009
11 Nov
'09
4:01 p.m.
Folderol wrote:
On Wed, 11 Nov 2009 20:18:24 +0100
Frank Neumann <beachnase(a)web.de> wrote:
Ok. I think I'm starting to get the picture. At 0.34 I tried using the
threaded
fftw3 routines. From my observations, that hasn't really contributed anything
worthwhile performance wise, and it's starting to look like that move has brought
instability for some. Give me a moment or three and I'll drop it.
cheers,
Hi list,
I'd just like to chime in here -
Josh Lawrence <hardbop200(a)gmail.com> wrote:
[..]
This is very strange. I am running on an AMD dual core desktop machine
and stability is exceptional. However I'm not running ubuntu. I am
running 64studio 3 beta with some debian squeeze enhancements.
Are you running 'standard' ubuntu or have you added an RT kernel with
the appropriate adjustments to /etc/security/limits.conf ?
first blush, things are looking good but there
are xruns all over the
place, so I installed the linux-rt kernel from the ubuntu repos.
eventually, yoshimi zombies, so I reboot into my new kernel.
zombies at every patch switch. adjusted the OscilSize just for
giggles, but it didn't do any good. my only thoughts at this point is
that this netbook just might be too little power for this application.
I have been
trying every 5 or so updates of yoshimi in the last weeks, and for
me it's similar to Josh's experiences:
- On an Asus eeePC Netbook, I can set up things fine and can also play the default
Sine sound, but normally within a few patch change clicks, yoshimi zombifies.
- On an AMD64 2-core desktop PC, after setting things up the first MIDI event
received by yoshimi already causes the zombie..so no sound at all here.