[PATCH] Updated dynamic E3DNow!/SSE2 port mixing

[Jussi Laako]

Hi,

I just updated the patch to process four samples at once. New patch
attached. The check for alignment can be dropped if the buffer alignment
is ensured (is it?).

I can commit this to the CVS if people think that it's a Good Idea(tm).


BR,

--
Jussi Laako <[hidden email]>

[Jesse Chappell]

Jussi wrote:

 > I just updated the patch to process four samples at once.
 > New patch attached. The check for alignment can be dropped
 > if the buffer alignment is ensured (is it?).
 >
 > I can commit this to the CVS if people think that it's a
 > Good Idea(tm).

Do you have any performance comparisons between the non sse2 mixing
and the new code?  Just curious...

jlc


-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
Jackit-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/jackit-devel

[Jussi Laako]

On Mon, 2005-08-15 at 17:54 -0400, Jesse Chappell wrote:
> Do you have any performance comparisons between the non sse2 mixing
> and the new code?  Just curious...

Attached is latest and more tuned version. Results for the attached
version, using 32 channels and period size 1024:

P4/3000z (Prescott), using SSE code:
gen_mix(): 81.6106 us
x86_mix(): 24.4454 us

Athlon XP 2400+, using E3DNow! code:
gen_mix(): 138.091 us
x86_mix(): 89.7967 us

Compile flags used: "-march=i686 -O3 -ffast-math"


BR,

--
Jussi Laako <[hidden email]>