Daniel Mack wrote: > Calling prepare() on a stream which is already active ... would be a bug; snd_pcm_pre_prepare() in sound/core/pcm_native.c should prevent this. Why do you feel the need for this patch? > shouldn't touch the sample pointer positions. This code belongs into > the trigger callback, so move it there. The trigger callback should execute as fast as possible; any initialization that isn't need to directly start the stream should go into the prepare callback. Regards, Clemens _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxx http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
