On Thu, Apr 02, 2009 at 06:36:27PM +0200, Lennart Poettering wrote: > > Ok, just some minutes after I wrote that it turned out that > > suspend/resume calls seem to be the culprit. If I'm fast enough to start > > a client right after the pulseaudio daemon came up, this doesn't happen. > > But once the daemon decided to send the device to sleep it won't > > recover. Is this a thing a driver has to implement? And is failing so > > hard if a driver doen't do that intended? ;) > > Suspend/resume (as mentioned in my previous mail) is implemented by > closing/reopening the device. When we reopen we try to configure the > same hwparams as initially, and that fails with EPERM. Ok, thanks. I'll try to get more insight on that and track things down. Daniel
