[LAU] oh yoshimi, one more time

[cal]

david wrote:
> [ ... ]
> Well, for what it's worth, here on a non-RT kernel but with RTPRIO set 
> properly in limits.conf, using jackdmp 1.9.2:
> 
> david at eurolap:~$ yoshimi
> Error, failed to start Alsa midi thread: 1

Define "properly", when a pthread_create asking for SCHED_FIFO returns '1',
indicating 'No permission to set the scheduling policy and parameters specified
in attr'.

> [ ... ]
> yoshimi: ../../src/xcb_lock.c:32: _XCBUnlockDisplay: Assertion 
> `dpy->xcb->partial_request == 0' failed.

I've no idea where an _XCBUnlockDisplay assert might come from. I've never ever
seen one so I don't believe it's my doing.

> [ ... ]
> david at eurolap:~$ yoshimi -A -a
> ALSA lib pcm_dmix.c:996:(snd_pcm_dmix_open) unable to open slave
> Error, alsa audio: failed to open alsa audio device:default: No such 
> file or directory
> Segmentation fault

Again, I don't see anything like this on my system with -A -a.
 
> david at eurolap:~$ yoshimi -A -j
> ALSA lib pcm_dmix.c:996:(snd_pcm_dmix_open) unable to open slave
> Error, alsa audio: failed to open alsa audio device:default: No such 
> file or directory
> Segmentation fault

Or with -A -j.

> [ ... ]
> 
> I don't think there should be a segfault just because it couldn't open 
> Alsa audio with JACK running ... ;-)

Agreed. I can't reproduce it on my system.
 
> Back to 0.036 for now.

Again, whatever works for you.

Cheers.