With the latest alse-lib release, several audio player (mpg123-alsa,
mpg321, ogg123) started using a huge amount of high cpu to play
anything. As an illustration, with mpg321 I get the numbers like the
following:
before after
real 2m32.188s 2m32.276s
user 0m2.212s 0m25.310s
sys 0m0.108s 0m47.259s
Through bisection, I've tracked this down to the following patch:
changeset: 2348:95e6e03f2e9d
user: tiwai
date: Mon Nov 12 12:01:16 2007 +0100
summary: Remove ugly hack in rate plugin poll_descriptors callback
If I re-add this hack to the 1.0.16 release, everything works find for
me. Should this be considered a regression? It looks like several apps
relied on the old behaviour. If this is considered an application bug,
perhaps alsa-lib could be made to print out a warning?
I'm using the snd_usb_audio driver with a C-Media usb device (id
0d8c:000c). I don't have a custom asoundrc file. I'm not sure what
information would be useful here, ask anything and I'll follow up :)
-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Alsa-user mailing list
Alsa-user@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/alsa-user
