Re: avoiding VM page swaps for big sample buffers
Re: avoiding VM page swaps for big sample buffers
- Subject: Re: avoiding VM page swaps for big sample buffers
- From: Shaun Wexler <email@hidden>
- Date: Wed, 29 Sep 2004 02:24:16 -0700
On Sep 29, 2004, at 1:59 AM, Mikael Hillborg wrote:
Does anyone know when it's useful to lock the memory, I mean where's
the limit? Is it worthwile to lock a few hundred KB? A couple of MB?
I lock down each active audio channel's I/O buffer when its device is
running, and unlock it when the device stops or the channel has no more
"connections" (in which case it's also freed). DSP buffers are also
locked appropriately, and unlocked when possible. For ongoing realtime
op's, I don't alloc/free anything between ioProc cycles either, but I
allow it for transient or CPU/fps-constrained op's.
As long as there's some free primary memory, then the buffer shouldn't
really be swapped out, should it?
No, but you can starve the rest of the system and cause swapfile
bashing which could end up blocking everything anyway.
--
Shaun Wexler
MacFOH
http://www.macfoh.com
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden