On Wed, 03.06.09 18:17, Mark Greenwood (fatgerman at ntlworld.com) wrote: > > Even if pulse suspends itself, it should be able to unsuspend OK unless > > something has snuck in and grabbed the device. > > > > When it's in this locked state, the output of "pacmd list-sinks" will > > confirm if it's in the suspended state and you should be able to issue a > > It does, and it is suspended. > > > "pacmd suspend-sink 0 false" to unsuspend sink with index 0. If it is > > locked into a suspend state, this should try and unsuspend it. > > 'pacmd suspend-sink 0 false' gives > Failed to parse suspend switch > and the output is still suspended this should work. it might be a bit simpler to simply issue "suspend 0" in pacmd, which suspend/resumes *all* cards instead of just one. > But starting something playing and listing sinks shows the state > RUNNING. Then PA after all did resume the device. Do you hear anything then? > > You can also try hitting the mute button twice in pavucontrol to see if > > it is another occurrence of the infamous "undetected mute on startup bug." > > Aha! That got it. After that the state went to IDLE and now I can > hear audio. While audio is playing it goes to RUNNING and then back > to IDLE, then back to SUSPENDED when I close pavucontrol. > > So looks like this is the cause. Can I help with debugging? Dunno if I understand the issue. Are you saying the mute status simply got stuck in some way? Lennart -- Lennart Poettering Red Hat, Inc. lennart [at] poettering [dot] net http://0pointer.net/lennart/ GnuPG 0x1A015CC4
