[LAU] oh yoshimi, one more time

[cal]

Folderol wrote:
> [ ... ]
>>> I found it happens with 038 too.
>>> and:
>>> It seems to depend on the instrument used:
>>> The default sine-like tone doesn't cause an error,
>>> but when using Rhodes/SynthRhodes2, pressing two keys at once
>>> zombifies the jack process.
>>> Is this reproducable?
>> Yep! I've already apologised to another poster that yesterday wasn't one
>> of my better days for clear-headedness, so same goes here.
>>  
>>> Maybe one of the internal synths/effects is still not rt-safe.
>> Looks to me like a straight deadlock, ie a lower level section waiting on a lock
>> while the lock is already held by it's caller. I'll work on it.
>>
>> Cheers, and thanks for your patience. There are days when I fully qualify as a
>> muddle headed wombat.
> 
> I find this all very strange. I don't seem to be having these problems.
> I wonder what is different about my setup. I don't for one moment
> think it's luck - luck and I are bitter enemies :o

And I believe you are quite right. In messing with the deadlock issue on
parameter save (0.054-pre4 only) I got all kinds of things confused. In
fact I can _not_ replicate the reported issue on Rhodes/SynthRhodes2 (or any
other instrument). The forearm test and the "two keys simultaneously" both
test fine for me ... well, as well as the forearm test ever does.

It's been a muddled couple of days here.

cheers, Cal