Subject: Building a universal binary of gsl - msg#00023

List: lib.gsl.general

Date: Prev Next Index Thread: Prev Next Index

I am looking for the most straightforward way to build a universal binary of
gsl on OSX. Simply passing CFLAGS/LDFLAGS="-arch i386 -arch ppc" doesn't
work due to incompatible flags in config.h, as far as I can tell. I was able
to make it work using this method:

tar -xzf gsl-1.9.tar.gz && mv gsl-1.9 gsl-1.9_intel && cp -r
gsl-1.9_intel gsl-1.9_ppc
cd gsl-1.9_intel && ./configure --disable-dependency-tracking
--disable-shared --enable-static
cd .. && cd gsl-1.9_ppc
env CFLAGS="-arch ppc" LDFLAGS="-arch ppc" ./configure
--disable-dependency-tracking --enable-shared --disable-static
--build=powerpc-apple-darwin8.10.0
cd .. && cd gsl-1.9_intel
./configure --disable-dependency-tracking --enable-shared --disable-static
lipo -create .libs/libgsl.0.10.0.dylib
../gsl-1.9_ppc/.libs/libgsl.0.10.0.dylib -output
.libs/libgsl.0.10.0.dylib
lipo -create cblas/.libs/libgslcblas.0.0.0.dylib
../gsl-1.9_ppc/cblas/.libs/libgslcblas.0.0.0.dylib -output
cblas/.libs/libgslcblas.0.0.0.dylib
sudo make install-strip

I would be interested in a more straightforward method of creating a
universal binary out of the same configuration..

Cheers,
Brian

Find Programming Jobs at git.net
(osdir sister site)

Thread at a glance:

Previous Message by Date: (click to view message preview)

How to return a pointer of gsl_vector

Let's say we have this program ////////////////////////////////////////////////////////////////// double one_function(); int main() { gsl_vector *tmp; tmp=one_function(); printf("Testing %f\n",gsl_vector_get(tmp,50)); } double one_function () { gsl_vector *tmp2=gsl_calloc_vector(100); gsl_vector_set(tmp2,50,-999); return *tmp2; } //////////////////////////////////////////// Obviously the program won't run. I need this because i have some gsl vectors in one class defined as private ( eg :private gsl_vector *tmp ) And i want to create a function in :public in order to return the vectors to main.

Next Message by Date: click to view message preview

Re: Double vs Long Double

At Fri, 5 Oct 2007 18:31:01 +0000 (GMT), dave wrote: > With xmm registers present in an x86, fp can be computed either using the NPX > math unit > or else using the xmm registers. Would changing 'double' to 'DOUBLE' > everwhere in the GSL > source and then adding -DDOUBLE='long double' or -DDOUBLE='double' to the > CFLAGS work as a > means of generating either a library that used NPX temp real mode or a > library using XMM > double mode? No, it's not that simple unfortunately. Without support from the compiler it's not possible to work with extended precision for every operation. -- Brian Gough Network Theory Ltd, Publishing Free Software Manuals --- http://www.network-theory.co.uk/

Previous Message by Thread: click to view message preview

Next Message by Thread: click to view message preview

GSL installation question

I have downloaded a version the CYGWIN project onto my home computer so I can compile C++ programs using the g++ compiler. I also have a need to use the GSL library for one of my projects. I went to a Cygwin mirror site and downloaded and installed the necessary files. When I tried to compile a program using functions contained in gsl_rng.h and gsl_randist.h the linker returned an error. It complained that it coulded find refereces to the functions and/or constant variables I declared in my code. I am running with Window XP. Do you know of any bug trying to run GSL in this enivornment or have any ideas on what could be the issue. Thanks Jeffrey H. Jewell Jeffrey H. Jewell Boeing Software Engineer - GPS System Simulator (719)380-5940 jeffrey.h.jewell@xxxxxxxxxx

Web Hosting Reviews from OSDir.com Sister Site iBizWebHosting.com