Re: dynamic allocation in the CoreAudio DSP thread
Re: dynamic allocation in the CoreAudio DSP thread
- Subject: Re: dynamic allocation in the CoreAudio DSP thread
- From: Doug Wyatt <email@hidden>
- Date: Tue, 02 Jul 2013 17:10:31 -0700
Without going and digging through the code right now, I can think of two property change notifications that would legitimately get sent from that thread:
- device is running (happens before and after we care about not glitching)
- overload (we have already glitched)
On Jul 1, 2013, at 19:25 , Info <email@hidden> wrote:
> Well stack trace 0 is malloc so I disagree. That and the fact that is possible to tell as I am sure Jeff M. knows.
>
> From: Art Gillespie
> Sent: Monday, July 01, 2013 10:10 PM
> To: Taylor Holliday
> Cc: email@hidden
> Subject: Re: dynamic allocation in the CoreAudio DSP thread
>
> It's impossible to tell, really: `operator new` here could be overridden to use something other than `malloc`.
>
>
> On Mon, Jul 1, 2013 at 5:45 PM, Taylor Holliday <email@hidden> wrote:
> Hi CoreAudio,
>
> While making sure my DSP code isn't doing any dynamic allocation, I noticed that HALObject seems to be doing it... am I misunderstanding the following stack trace?
>
> https://dl.dropboxusercontent.com/u/45540223/Audulus/DynamicAlloc.tiff
>
> thanks!
> - Taylor
_______________________________________________
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