Frequent crashes of PulseAudio

Dimi Paun dimi at lattica.com
Mon Apr 7 20:52:39 UTC 2008


On Mon, 2008-04-07 at 22:39 +0200, Lennart Poettering wrote:
> 
> If you get a CPU time limit abort than this can have two reasons:
> 
> 1) You got an awfully slow CPU (very unlikely, if you bought your
> machine in the last 10 years and is x86 or x86-64)
> 
> 2) Something causes PA to enter an endless loop which causes it to eat
> 100% CPU.
> 
> In all the cases I managed to track down #2 is caused by a faulty ALSA
> driver where poll() signalled readiness while there actually was
> nothing to do. This happened with drivers which required the "multi"
> libasound module, such as emu10k.

Well, 1) clearly doesn't apply to me, I'm running on Core Duo 6400.

As for 2) it is true that I have noticed PA start using 100% CPU for
a few seconds before exiting. Randomly. A lot of times I find it gone
in the morning, with no music playing during the night.

As for sound card, I'm using a bog-standard Intel Corporation 82801H
(ICH8 Family) HD Audio Controller (rev 02).

-- 
Dimi Paun <dimi at lattica.com>
Lattica, Inc.




More information about the fedora-devel-list mailing list