Re: Audio Units and virtual memory
Re: Audio Units and virtual memory
- Subject: Re: Audio Units and virtual memory
- From: Benjamin Golinvaux <email@hidden>
- Date: Thu, 27 Feb 2003 13:55:45 +0100
you should have a look at the "mlock/munlock" calls that allow to
create a non-pageable area in memory.
look for 'mlock' in the list archives. It has been discussed a couple
of times
since this is maybe a bit extreme, perhaps you can have "warm up"
your buffers (in the main thread) before using them and check
whether it's enough..... (of course it won't solve any page out/page in
cycle problems)
Benjamin-
On Thursday, Feb 27, 2003, at 02:51 Europe/Paris, Glenn Olander wrote:
You've lost me here. There are no patch changes. The glitch occurs
during playback when touching a region of memory for the first
time while rendering, for example a region of a wave table.
(I say "first time", but I presume it could happen whenever that
memory is paged out).
This seems like an inherent danger of using virtual memory in
a real-time system. What strategies are people using to avoid
page faults while rendering audio?
- Glenn
Bill Stewart wrote:
Your "host" should be calling prepare instrument before it expects a
glitch free patch change to occur. I've been doing some work on the
MusicSequence / Player to actually fix this properly (which it now
is:)
_______________________________________________
coreaudio-api mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/coreaudio-api
Do not post admin requests to the list. They will be ignored.
_______________________________________________
coreaudio-api mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/coreaudio-api
Do not post admin requests to the list. They will be ignored.