Re: [linux-audio-user] Correct CFLAGS for libDSP and libConvolve?

On Monday 19 February 2007, James Stone wrote:
> On Sun, Feb 18, 2007 at 09:39:55PM -0800, Ken Restivo wrote:
> > .columbia.edu
> > Errors-To: linux-audio-user-bounces@email-addr-hidden
> > X-Virus-Scanned: by amavisd-new at music.columbia.edu
> > Status: O
> > Content-Length: 1692
> > Lines: 38
> >
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > I decided I'd really like to try these convolution reverbs that people on
> > other platforms have been using for nearly 10 years, apparently.
> >
> > So I grabbed jack_convolve, libconvolve, and libDSP.
> >
> > Anyone want to venture a guess as to what GCC flags to use, or, better
> > yet, how to cheat and make autotools tell me, even though this package
> > doesn't use it?
>
> You need to use an old version of gcc. 3.4 works. The assembly
> bit has completely changed for gcc 4.x and libDSP is not going to
> be updated (AFAIK)

I might just rip out the relevant libDSP parts and put them directly into
jack_convolve. Anyways, if you don't succeed in building with libDSP, you can
still use the naive C implementation of the complex multiply-and-add function
(see the README in libconvolve and the Makefile)..

jack_/libconvolve themself should work perfectly fine with newer gcc's..

Flo

P.S.: Maybe anybody has some ready to use optimized complex multiply-and-add
function lying around? Let me know..

-- 
Palimm Palimm!
http://tapas.affenbande.org