Re: [Jack-Devel] Jack 0.120 incorrect error recovery for -n 3

Le 19 avr. 2011 à 22:28, torbenh a écrit :

> On Tue, Apr 19, 2011 at 07:52:03AM -0400, Paul Davis wrote:
>> On Tue, Apr 19, 2011 at 6:57 AM, Stéphane Letz <[hidden]> wrote:
>> 
>>> So is this fix considered "official" ?  (before I report it again on jack2)
>> 
>> from chatting with torben yesterday, i think we're not quite done yet.
> 
> well... we probably need to write zeros instead of empty buffers.
> but if Stéphane has that code already, it should be fine.
> 
> i am wondering what you mean by official ?
> the old behaviour was clearly wrong. it was just that nobody really
> cared before. (i have still no real explanation of why things worked,
> but i know that i fixed a bug in the null cycle, and that bug might have
> been "fixing" the offsets, during the client kick.
> 
> or we had real xruns all the time anyways.
> 

Some weeks ago we had these bug reports about "incorrect" behaviour of JACK2 on ARM7  (but not with JACK1) and it appeared that JACK2 interaction with ALSA backend was a bit different compared to JACK1. So I switched to a model of "do not call backend write"  in case of error, just to follow JACK1 model, which appear to be incorrect anyway...

So should I switch back to the previous model?

I have to say that the way a given backend (ALSA , CoreAudio, PortAudio...) interacts with the engine is somewhat unspecified: for instance CoreAudio send a kAudioDeviceProcessorOverload notification form another thread (than the RT one) and reset the driver, ALSA detect the XRun in the RT callback (and restart the driver), as well as PortAudio...

Would it make sense to try to make that part clearer?

Stéphane 

1303289538.9042_0.ltw:2,a <5B4D6409-F541-46E0-A2E7-F1897B2A651B at grame dot fr>