philipp/nixpkgs
1
0
Fork 0
Code Issues Pull requests Projects Releases Activity
%!d() commits 3 branches 0 tags 1.4 GiB
Commit graph

19 commits

Author SHA1 Message Date
Peter Simons 83df975ba8 atlas: allow users to override the CPU configuration used to build ATLAS 2013-02-06 12:27:23 +01:00
Peter Simons 3334f33daf atlas: update to version 3.10.1 
This patch fixes issue #297.
 2013-02-06 11:31:49 +01:00
Eelco Dolstra 7fd42a8380 Get rid of the "sf" alias for "sourceforge" 2013-01-14 12:00:32 +01:00
Lluís Batlle i Rossell 8f50680a05 I hope now I guess fine how to tell ATLAS not to accept any simd for i686. 
svn path=/nixpkgs/trunk/; revision=33793
 2012-04-15 12:44:48 +00:00
Lluís Batlle i Rossell 2e753bc4cf Typo again in atlas. 
svn path=/nixpkgs/trunk/; revision=33791
 2012-04-15 08:11:12 +00:00
Lluís Batlle i Rossell 023cd3b7bd Oh, I removed -fPIC in my last atlas commit. I didn't mean to. 
svn path=/nixpkgs/trunk/; revision=33790
 2012-04-15 07:58:58 +00:00
Lluís Batlle i Rossell 30d7a7fbe1 I think we better disable SIMD for i686. Or it can detect whatever the machine has. 
svn path=/nixpkgs/trunk/; revision=33789
 2012-04-14 22:00:12 +00:00
Lluís Batlle i Rossell 60ecc500f9 Trying to fix atlas i686. I set Pentium II and having sse/sse2 at the same time for i686... 
I think Pentium II does not have sse.


svn path=/nixpkgs/trunk/; revision=33787
 2012-04-14 18:58:30 +00:00
Lluís Batlle i Rossell 4250dd8104 I made atlas build for fixed architectures. 
We saw a crash in many computers, in the octave check phase, where octave crashed.
It was due to atlas being built for AMD Family 10h, which has a special SSE
trick that others computer don't have.

For x86_64, atlas is for K7. And for i686, PII.


svn path=/nixpkgs/trunk/; revision=33780
 2012-04-13 18:46:24 +00:00
Michael Raskin 2f97078c96 Support shared library build 
svn path=/nixpkgs/trunk/; revision=33612
 2012-04-05 12:18:08 +00:00
Peter Simons e321903ff1 Atlas aborts the build some kind of CPU frequency scaling is detected on the 
build machine because that feature offsets the performance timings. We ignore
that check, however, because with binaries being pre-built on Hydra those
timings aren't accurate for the local machine in the first place. The build log
might show something such as the following:

 | It appears you have cpu throttling enabled, which makes timings
 | unreliable and an ATLAS install nonsensical.  Aborting.
 | See ATLAS/INSTALL.txt for further information
 | Ignoring CPU throttling by user override!

svn path=/nixpkgs/trunk/; revision=32506
 2012-02-23 11:33:26 +00:00
Lluís Batlle i Rossell 8f7485d784 Making atlas do the check phase. 
svn path=/nixpkgs/trunk/; revision=32493
 2012-02-22 22:40:35 +00:00
Lluís Batlle i Rossell e4c4ab35e5 I think I fixed atlas for i686. It wants '-b 32'. Why, I don't know. 
svn path=/nixpkgs/trunk/; revision=32471
 2012-02-22 11:37:10 +00:00
Lluís Batlle i Rossell 2847a37fa9 Updating atlas to 3.9.67, the latest development version 
svn path=/nixpkgs/trunk/; revision=32458
 2012-02-21 14:49:36 +00:00
Eelco Dolstra 6609710409 * Get rid of many instances of "args: with args;", and other coding 
guidelines violations.
* Updated libsamplerate to 0.1.7.

svn path=/nixpkgs/trunk/; revision=22782
 2010-07-28 11:55:54 +00:00
Marc Weber 94ba737704 fix atlas by Tobias Neumann 
svn path=/nixpkgs/trunk/; revision=15685
 2009-05-21 17:39:17 +00:00
Lluís Batlle i Rossell 645279d544 Updating atlas 
svn path=/nixpkgs/trunk/; revision=14957
 2009-04-08 22:08:05 +00:00
Marc Weber d97a7e08bc make atlas build again (tested on amd64) 
svn path=/nixpkgs/trunk/; revision=13560
 2008-12-02 22:55:33 +00:00
Pjotr Prins 2391d84867 Two math packages 
svn path=/nixpkgs/trunk/; revision=12955
 2008-10-05 09:01:59 +00:00