On Mon, 23 Aug 2021 13:48:38 +0200, Arkadiusz Bokowy wrote:
By checking error code returned by the pointer callback, we can determine more precisely PCM state. Previous implementation assumed, that a software PCM can only produce overrun or underrun. It was impossible to mark software PCM as disconnected.
Signed-off-by: Arkadiusz Bokowy arkadiusz.bokowy@gmail.com
Please fix MUA setup.
Also...
src/pcm/pcm_ioplug.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/src/pcm/pcm_ioplug.c b/src/pcm/pcm_ioplug.c index c96104e9..0e27e121 100644 --- a/src/pcm/pcm_ioplug.c +++ b/src/pcm/pcm_ioplug.c @@ -81,11 +81,21 @@ static int snd_pcm_ioplug_hw_ptr_update(snd_pcm_t *pcm) } io->last_hw = (snd_pcm_uframes_t)hw; } else {
- switch (hw) {
- case -ESTRPIPE:
- io->data->state = SND_PCM_STATE_SUSPENDED;
- break;
- case -ENODEV:
- io->data->state = SND_PCM_STATE_DISCONNECTED;
- break;
- default:
- io->data->state = SND_PCM_STATE_XRUN;
- } if (io->data->state == SND_PCM_STATE_DRAINING) snd_pcm_ioplug_drop(pcm); else io->data->state = SNDRV_PCM_STATE_XRUN;
- return -EPIPE;
- return hw;
Be careful about the handling of the draining case.
thanks,
Takashi