via8235 Audio on 2.4.21-pre
Greetings,
For the past two months, I've been testing Alan's updates to the via82cxxx_audio driver which allow the codecs attached to the VT8235 to work quite nicely. Recent updates have allowed the majority of applications to work, however any version of ESD fails (whilst the same configuration works fine with other sound drivers).
This has been a problem ever since the first via82cxxx_audio releases with vt8235 support, but at present, I'm using 2.4.21-pre4-ac4. When ESD is loaded, 3 of the five beeps are heard (unless the beeps are disabled, in which I can get about 0.25 seconds of audio to play through ESound) and then ESound will hang. Once esound is killed, I get the message:
via_audio: ignoring drain playback error -512
This has been tested on several motherboards with different codecs - most of which generally work well when using normal OSS applications. However, the VT1612A (Subsystem: VIA Technologies, Inc.: Unknown device 4161) which has a PCM2 volume control which behaves as the Master volume should; on this chip, the Master volume doesn't do anything at all.
Are there any debugging modes that I can activate on this driver to provide more valuable information? I'd really like to see this chip working as ALSA has had less success than the OSS driver with a few codecs attached to the VT8235.
Thanks,
Jonathan
Quoted 3 lines · Jonathan Thorpe
work quite nicely. Recent updates have allowed the majority of applications to work, however any version of ESD fails (whilst the same configuration works fine with other sound drivers).
esd is the one failure case I know about
Quoted 7 lines
This has been a problem ever since the first
via82cxxx_audioreleases with vt8235 support, but at present, I'm using 2.4.21-pre4-ac4. When ESD is loaded, 3 of the five beeps are heard (unless the beeps are disabled, in which I can get about 0.25 seconds of audio to play through ESound) and then ESound will hang. Once esound is killed, I get the message:
via_audio: ignoring drain playback error -512
The printk itself is actually a bug, but its providing clues. Something esd does seems to stop the audio engine running. You get one chunk then it all stops. I can duplicate it here too.
Alan