On 07/29/2009 11:01 AM, Anne Wilson wrote:
On Wednesday 29 July 2009 18:53:15 Anne Wilson wrote:
> On Wednesday 29 July 2009 18:32:52 jack craig wrote:
>
>> On 07/29/2009 10:17 AM, Antonio M wrote:
>>
>>> after latest updates, any sound application when palys streams from
>>> the net crashes after a short time.
>>> No idea what is causing such a problem
>>>
>> explore recent threads, look at your /var/log/messages and see if our
>> recent discovery is also your problem?
>>
>> i have a work around if you are getting alsa bug reports in your
>> messages file.
>>
>> hth, jackc...
>>
> How about a clue, Jack? So much hokum has been talked that I've been
> skipping the sound threads, particularly as everything else plays
> correctly. I thought or a moment that I was going to get a whole track
> played (my problem is in amarok, but exactly the same as the rhythmbox
> one). This is what happened:
>
> Jul 29 18:45:33 anne-laptop pulseaudio[26616]: ratelimit.c: 7106 events
> suppressed
> Jul 29 18:45:33 anne-laptop pulseaudio[26616]: alsa-sink.c: Increasing
> minimal latency to 56.00 ms
> Jul 29 18:48:26 anne-laptop pulseaudio[26616]: alsa-sink.c: Increasing
> wakeup watermark to 46.00 ms
> Jul 29 18:48:31 anne-laptop pulseaudio[26616]: ratelimit.c: 6194 events
> suppressed
> Jul 29 18:48:36 anne-laptop pulseaudio[26616]: ratelimit.c: 6014 events
> suppressed
> Jul 29 18:48:38 anne-laptop kernel: kio_thumbnail[808]: segfault at 1320 ip
> 00a4f1de sp bfa0848c error 4 in libgcc_s-4.4.0-20090506.so.1[a2e000+2a000]
> ^C
>
> The moment of the 'Increasing wakeup watermark' message is when it started
> to race and crackle.
>
> Is this what you were expecting to see? I so, what's the workaround?
>
> Just out of curiosity, I'll do the same experiment while playing it in
> kaffeine, as I have never seen it happen there. It will be interesting to
> see whether the same messages appear.
>
>
Yes, in kaffeine the same thing happened. Everything was going swimmingly
until
Jul 29 18:58:51 anne-laptop pulseaudio[26616]: ratelimit.c: 4191 events
suppressed
Jul 29 18:58:51 anne-laptop pulseaudio[26616]: alsa-sink.c: Increasing wakeup
watermark to 56.00 ms
Jul 29 18:58:56 anne-laptop pulseaudio[26616]: ratelimit.c: 6390 events
suppressed
Jul 29 18:59:01 anne-laptop pulseaudio[26616]: ratelimit.c: 5688 events
suppressed
So it's an alsa bug?
Anne
sorry, i didnt want to pollute the group with WAG's if its not the same
issue.
see
https://bugzilla.redhat.com/show_bug.cgi?id=514213
for defect details.
this is apparently fixed in the mainline and due next release.
i am curious if you dont also get the messages entry about the alsa bug for
snd_pcm_delay() ?
anyway, the fix that worked for me, posted earlier, ...
to disable the g-f feature. do this, in
/etc/pulse/default.pa,
change, ...
load-module module-hal-detect
to
load-module module-hal-detect tsched=0
kill off the existing pulseaudio daemon so it restarts and reads the config update.
hth, jackc...
--
jack craig
jackc(a)LinuxLightHouse.com
831-684-1375 (Office)
831-596-6924 (cell)
IM: jackcraigaptos (AIM)