Hi,
I had the same problem until I gave a look at RTirq config file
(/etc/default/rtirq on debian). It used to give high rtprio
to usb device with the following config line: "rtc snd usb ..."
But as I have a PCI audio card, I just had to remove 'usb' from it
and the 10min xrun problem is solved for me.
I still have a problem with my SCSI controller, but I chose
to start another thread.
I hope it can help somebody.
Bye,
Romain
Joe Hartley wrote:
On Fri, 28 Sep 2007 03:35:49 -0700
Martin Lynch <martin.lynch-fVOoFLC7IWo(a)public.gmane.org> wrote:
As I write this, Qjackctl shows 14:35 since the
last X Run!!!
I don't have this completely narrowed down as to the offending
program(s), however what I did was open the System Monitor and began
stopping programs I thought were either likely offenders or simply
unnecessary. One thing I didn't realize was that Evolution (linux mail
client), despite not being my default mail client and not being active
still had 4 programs running - remains to be seen if that's the culprit.
Here's what I currently have stopped:
ntpd
This is unlikely to be the culprit
evolution-alarm-notifier
evolution-data-server-1.10
evolution-exchange-storage
evolution
freshclam
update-notifier
gnome-screensaver
These all have a high chance of being part of the problem!
getty (there were 6 instances of this running,
all showing the identical
amount of memory being used - I left one on and stopped the other 5 - no
idea what this is - will research it further - but doubt this has
anything to do with the X run issue, since one is still running)
These handle
the logins for the terminals on the Alt-F1 through 6
screens. They are not the processes you're looking for, move along.
FYI: I'm now at over 22 mins without an X
run, still at my 2ms latency
setting and I happen to be watching an HD movie at the same time! (and
no, it didn't take me 8 mins to write this - had to have a pillsbury
turnover break for fortitude ; - )
Sounds like you found it! Congrats!