Jussi Laako wrote:
On Sun, 2005-11-20 at 12:54 +1100, Dave Robillard wrote:
Out of curiosity, how expensive is this runtime architechture check?
It's done only once at initialization time and even there it's matter of
< 100 machine instructions.
At runtime the cost is doing integer comparis
On Sun, 2005-11-20 at 12:54 +1100, Dave Robillard wrote:
> Out of curiosity, how expensive is this runtime architechture check?
It's done only once at initialization time and even there it's matter of
< 100 machine instructions.
At runtime the cost is doing integer comparison.
--
Jussi Laako <
> Out of curiosity, how expensive is this runtime architechture check?
>
I don't think runtime detection is necessary if you compile both
library and app for the specific arch. A few ifdef's take care of the
selection at compile time.
If you want to provide a multiarch binary, I don't know how he
On Sun, 2005-20-11 at 00:23 +0200, Jussi Laako wrote:
> On Wed, 2005-11-16 at 00:05 +0100, Florian Schmidt wrote:
> > I think libDSP does prefetch and cache alignment, SIMD, yadayada :)
> >
> > I don't know though to which degree each one of the functions is
> > optimized. Best to ask Jussi himse
On Wed, 2005-11-16 at 00:05 +0100, Florian Schmidt wrote:
> I think libDSP does prefetch and cache alignment, SIMD, yadayada :)
>
> I don't know though to which degree each one of the functions is
> optimized. Best to ask Jussi himself (CC'ed) :)
Most of the time prefetch is left to compiler (wo
On Tue, 15 Nov 2005 14:49:01 +0100
Alfons Adriaensen <[EMAIL PROTECTED]> wrote:
> The only place where I've seen prefetch used explicitly is in Brutefir's
> sse and 3dnow routines which I recently modified for use in one of my own
> projects.
I think libDSP does prefetch and cache alignment, SIMD
> > On a related subject: How is level one cache replaced with new data,
> > should one (or ones compiler) decide to use some of the prefetch
> > instructions available from Intel PII and up? It would make sense to
> > fetch the next dataset while doing what has to be done "now". On the
> > other
On Tue, 15 Nov 2005 at 07:24 -0500, Paul Davis wrote:
> On Tue, 2005-11-15 at 11:58 +, James McDermott wrote:
> > > What are your thoughts? What is best practice on multichannel audio, or
> > > is it always application-specific?
> >
> > According to my experience and understanding:
> >
> > -n
On Tue, Nov 15, 2005 at 02:12:55PM +0100, Jens M Andreasen wrote:
(good to see you're back on line :-)
> On a related subject: How is level one cache replaced with new data,
> should one (or ones compiler) decide to use some of the prefetch
> instructions available from Intel PII and up? It would
On a related subject: How is level one cache replaced with new data,
should one (or ones compiler) decide to use some of the prefetch
instructions available from Intel PII and up? It would make sense to
fetch the next dataset while doing what has to be done "now". On the
other hand, overwriting the
On Tue, 2005-11-15 at 11:58 +, James McDermott wrote:
> > What are your thoughts? What is best practice on multichannel audio, or
> > is it always application-specific?
>
> According to my experience and understanding:
>
> -non-interleaved (multiple channels in separate arrays) is a bit
> eas
> What are your thoughts? What is best practice on multichannel audio, or
> is it always application-specific?
According to my experience and understanding:
-non-interleaved (multiple channels in separate arrays) is a bit
easier to code, but
-interleaved could give better performance (because the
I'm writing a library in ruby for dealing with audio data, and I'm faced
with a design decision.
For several reasons, the best thing to use in ruby for numerical data is
NArray[1] which is implemented in C for efficiency. So my code is
basically a wrapper around NArray which gives some more specif
13 matches
Mail list logo
