uo9xpb74
/
mathnet-numerics
forked from tsai/mathnet-numerics
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
Browse Source

Merge pull request #345 from lionpeloux/provider-ldp 

Update mkl native provider for OSX
netstandard
Christoph Ruegg 11 years ago
parent
55db448f90 58c6da4494
commit
1b0cc26c7d
2 changed files with 28 additions and 3 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 25
      docs/content/MKL.md
  2. 6
      src/NativeProviders/OSX/mkl_build.sh

25
docs/content/MKL.md
View File

@ -103,6 +103,31 @@ or just copy them e.g. to `/usr/lib`.
For details see Mono's [Interop with Native Libraries](http://www.mono-project.com/docs/advanced/pinvoke/#mac-os-x-framework-and-dylib-search-path).
To build the MKL native provider :
1. Make sure you've a valid [intel MKL](https://software.intel.com/en-us/qualify-for-free-software/academicresearcher) licence installed on your mac (look at opt/intel). If not, you can get a free trial on intel's web site.
1. Open the terminal
2. cd to the folder mathnet-numerics/src/NativeProviders/OSX
3. Run the .sh script by typping sh mkl_build.sh
4. ... wait for the build
[lang=sh]
lionel:~ Lionel$ cd /Users/Lionel/Public/Git/GitHub/mathnet-numerics/src/NativeProviders/OSX
lionel:OSX Lionel$ ls
mkl_build.sh
lionel:OSX Lionel$ sh mkl_build.sh
5. Check the /x86 and /x64 folders in mathnet-numerics/out/MKL : you should now find the `libiomp5.dylib` and `MathNet.Numerics.MKL.dll` libaries.
6. you need to add the path to the generated libraies in your `DYLD_LIBRARY_PATH` environnement variable (which you can move to the folder of you choice before). To do that, open your /Users/Lionel/.bas_profile.sh file with a text editor and add the following statements.
[lang=sh]
export DYLD_LIBRARY_PATH=$DYLD_LIBRARY_PATH:/Users/Lionel/../mathnet-numerics/out/MKL/OSX/x64
export DYLD_LIBRARY_PATH=$DYLD_LIBRARY_PATH:/Users/Lionel/../mathnet-numerics/out/MKL/OSX/x86
Of course replace `Lionel` by your account login.
Have a look a the example down this page to compare MKL-provider vs. managed-provider performances.
F# Interactive
--------------

6
src/NativeProviders/OSX/mkl_build.sh
View File

@ -1,15 +1,15 @@
export INTEL=/opt/intel
export MKL=$INTEL/mkl
export OPENMP=$INTEL/composerxe/lib
export OPENMP=$INTEL/lib
export OUT=../../../out/MKL/OSX
mkdir -p $OUT/x64
mkdir -p $OUT/x86
clang++ -D_M_X64 -DGCC -m64 --shared -fPIC -o $OUT/x64/MathNet.Numerics.MKL.dll -I$MKL/include -I../Common ../MKL/memory.c ../MKL/capabilities.cpp ../MKL/vector_functions.c ../MKL/blas.c ../MKL/lapack.cpp $MKL/lib/libmkl_intel_lp64.a $MKL/lib/libmkl_core.a $MKL/lib/libmkl_intel_thread.a -L$OPENMP -liomp5 -lpthread -lm
clang++ -std=c++11 -D_M_X64 -DGCC -m64 --shared -fPIC -o $OUT/x64/MathNet.Numerics.MKL.dll -I$MKL/include -I../Common -I../MKL ../MKL/memory.c ../MKL/capabilities.cpp ../MKL/vector_functions.c ../Common/blas.c ../Common/lapack.cpp $MKL/lib/libmkl_intel_lp64.a $MKL/lib/libmkl_core.a $MKL/lib/libmkl_intel_thread.a -L$OPENMP -liomp5 -lpthread -lm
cp $OPENMP/libiomp5.dylib $OUT/x64/
clang++ -D_M_IX86 -DGCC -m32 --shared -fPIC -o $OUT/x86/MathNet.Numerics.MKL.dll -I$MKL/include -I../Common ../MKL/memory.c ../MKL/capabilities.cpp ../MKL/vector_functions.c ../MKL/blas.c ../MKL/lapack.cpp $MKL/lib/libmkl_intel.a $MKL/lib/libmkl_core.a $MKL/lib/libmkl_intel_thread.a -L$OPENMP -liomp5 -lpthread -lm
clang++ -std=c++11 -D_M_IX86 -DGCC -m32 --shared -fPIC -o $OUT/x86/MathNet.Numerics.MKL.dll -I$MKL/include -I../Common -I../MKL ../MKL/memory.c ../MKL/capabilities.cpp ../MKL/vector_functions.c ../Common/blas.c ../Common/lapack.cpp $MKL/lib/libmkl_intel_lp64.a $MKL/lib/libmkl_core.a $MKL/lib/libmkl_intel_thread.a -L$OPENMP -liomp5 -lpthread -lm
cp $OPENMP/libiomp5.dylib $OUT/x86/

Write Preview
Loading…
Cancel
Save