Thank you all for your thoughtful comments.
I've read about the effects of cache on CPUs, but admittedly neglected
to consider that - good to be reminded of that.
As far as I have read about isolated cores, the kernel doesn't schedule
anything to run on them (kernel 4.1.13), so I would expect that they
don't handle interrupts either (though I'll have to research that to be
sure). Cache hit/miss activity will certainly affect an isolated core's
performance tho.
Testing was also mentioned and I plan to be inventing all kinds of
torture tests for it so that I can find the edges of the envelope.
As for MIDI latency, the core activity for the Linux cores runtime
configuration will be as lightly loaded as possible, no unnecessary
services (likely headless), since the kernel controlled cores are where
the MIDI UART and sound card interrupts will be fielded (as I mentioned,
this is my understanding of how isolated cores work at least on this CPU).
Given the nature of the CPU, testing appears to be a very important part
of this process since what I want is a reliable and accurate musical
instrument with good voice count and good feature set.
"optimizing for worst case" - that's exactly where I'm headed. I'm
convinced the Rpi2 has the power to make a very nice electronic musical
instrument.
Thanks again people. I may be back with more DSP related questions once
I get this full going. Currently it's bits and pieces of
proof-of-concept code (that grows daily).
-- Scott Gravenhorst
_______________________________________________
dupswapdrop: music-dsp mailing list
music-dsp@music.columbia.edu
https://lists.columbia.edu/mailman/listinfo/music-dsp