rpress;637018 Wrote: 
> Ah, so I've been looking into this.  I added some code to http_recv so
> that it would yield the thread if the audio->output_buffer is too low. 
> This worked very well and fixed it 90%, but it seems like there is still
> something else that is tying things up occasionally.


I've been looking at the changes to the pa_callback, there's still a
lock/unlock pair happening in the slimaudio_buffer_available function.
I believe that the buffering also needs to be changed to use a
ringbuffer implementation which doesn't require locks to access the
buffer.

As you're using the slimaudio_buffer_available in http_recv as well,
that could be causing the last 10%.

Another possibility is the decoder thread spinning too long for the
slower CPU.

Sorry it's all just suggestions and nothing concrete.


-- 
ralphy

Ralphy

*4*-Classics, *2*-Booms, *12*-Squeezeslaves
'Squeezeslave' (http://code.google.com/p/squeezeslave/) 'donations'
(https://www.paypal.com/cgi-bin/webscr?cmd=_donations&business=LL5P6365KQEXN&lc=CA&item_name=Squeezeslave&currency_code=USD&bn=PP%2dDonationsBF%3abtn_donate_SM%2egif%3aNonHosted)
always appreciated.
------------------------------------------------------------------------
ralphy's Profile: http://forums.slimdevices.com/member.php?userid=3484
View this thread: http://forums.slimdevices.com/showthread.php?t=83362

_______________________________________________
discuss mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/discuss

Reply via email to