Re: Audio mixing in kernel ?
Re: Audio mixing in kernel ?
- Subject: Re: Audio mixing in kernel ?
- From: Matthew Xavier Mora <email@hidden>
- Date: Tue, 23 Jun 2009 07:08:17 -0700
The mixing code is run in the kernel during the audio io trap call.
The audio trap call explicitly saves and restores the fp register
because it knows that it will be dealing with floats. The kernel does
not save and restore fp registers when doing a context switch from
user to kernel. If the code getting called in the kernel touches those
registers it must save and store them.
Matt
Sent from my iPhone
On Jun 23, 2009, at 4:25 AM, Stéphane Letz <email@hidden> wrote:
Hi,
My understanding of CoreAudio layering is that audio mixing (when
several applications play an audio stream) is actually done in
kernel mode.
It that correct?
So does it mean that OSX actually save/restore the processor FPU
context between kernel thread switches?
Thanks
Stephane Letz
_______________________________________________
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
_______________________________________________
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