29 Jul
2005
29 Jul
'05
2:44 p.m.
On Fri, Jul 29, 2005 at 11:52:37AM +0200, Florian Schmidt wrote:
On Fri, 29 Jul 2005 05:11:22 +0200
Cesare <cesare(a)poeticstudios.com> wrote:
Hi Flo,
Is the wiki in a decently up to date state on this stuff?
I'll be away next week, but starting August 8th, I _will_ follow through
on chasing this down for the boxes I've got here.
For now, I'm going to disable the NICs and put in eepros.
--
Eric Dantan Rzewnicki | Systems Administrator
Technical Operations Division | Radio Free Asia
2025 M Street, NW | Washington, DC 20036 | 202-530-4900
CONFIDENTIAL COMMUNICATION
This e-mail message is intended only for the use of the addressee and
may contain information that is privileged and confidential. Any
unauthorized dissemination, distribution, or copying is strictly
prohibited. If you receive this transmission in error, please contact
network(a)rfa.org.
I've tried also with a vanilla 2.6.11.7 with
both the realtime preempt
and the realtime-lsm patch. And got even more xruns.
But maybe I haven't disabled some daemon or something else.
Hmm,
a realtime preempt kernel alone will not nessecarily do any good. You
need to tune the irq handlers. If your network irq handler has a higher
RT prio than the soundcard irq handler, xruns are preprogrammed.
If you run RP make sure that:
a] the soundcard irq has the highest prio in the system (i.e. around 90)
b] all other irq's are left near their default (around 50)
c] jackd's prio is inbetween (i use 70 with great results)
If you then still get xruns, the hw is to blame [imho]. Or maybe a bug
in RP patches. Use the latency diagnosis tools that the RP kernel
provides to find out what part of the kernel produces the delays.