Aquila-Consortium/libsharp2
0
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

better documentation

Browse source
This commit is contained in:
Martin Reinecke 2019-01-21 11:14:19 +01:00
parent 90d2444bf1
commit e231a0e184
1 changed files with 13 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

26
COMPILE
View file

@ -28,14 +28,14 @@ to the C99 standard, you should still be able to compile libsharp with
Runtime CPU selection with gcc Runtime CPU selection with gcc
------------------------------ ------------------------------
When using a recent gcc (6.0 and newer) on an x86_64 platform, the build When using a recent gcc (6.0 and newer) or a recent clang (successfully tested
machinery will compile the time-critical functions for several different with versions 6 and 7) on an x86_64 platform, the build machinery can compile
architectures (SSE2, AVX, AVX2, FMA3, FMA4, AVX512F), and the appropriate the time-critical functions for several different architectures (SSE2, AVX,
implementation will be selected at runtime. AVX2, FMA3, FMA4, AVX512F), and the appropriate implementation will be selected
This only happens if you do _not_ explicitly specify a target architecture via at runtime.
the compiler flags. I.e., please do _not_ specify "-march=native" or This is enabled by passing "-DMULTIARCH" as part of the CFLAGS.
"-mtarget=avx" or similar if you want a portable binary that will run If this is enabled, please do _not_ specify "-march=native" or
efficiently on different x86_64 CPUs. "-mtarget=avx" or similar!
If you are compiling libsharp for a particular target CPU only, or if you are If you are compiling libsharp for a particular target CPU only, or if you are
using a different compiler, however, "-march-native" should be used. The using a different compiler, however, "-march-native" should be used. The
resulting binary will most likely not run on other computers, though. resulting binary will most likely not run on other computers, though.
@ -65,16 +65,16 @@ Example configure invocations
============================= =============================
GCC, OpenMP, portable binary: GCC, OpenMP, portable binary:
CFLAGS="-std=c99 -O3 -ffast-math -fopenmp" ./configure CFLAGS="-DMULTIARCH -std=c99 -O3 -ffast-math -fopenmp" ./configure
GCC, no OpenMP, portable binary: GCC, no OpenMP, portable binary:
CFLAGS="-std=c99 -O3 -ffast-math" ./configure CFLAGS="-DMULTIARCH -std=c99 -O3 -ffast-math" ./configure
Clang, OpenMP, nonportable binary: Clang, OpenMP, portable binary:
CC=clang CFLAGS="-std=c99 -O3 -march=native -ffast-math -fopenmp" ./configure CC=clang CFLAGS="-DMULTIARCH -std=c99 -O3 -ffast-math -fopenmp" ./configure
Intel C compiler, OpenMP, nonportable binary: Intel C compiler, OpenMP, nonportable binary:
CC=icc CFLAGS="-std=c99 -O3 -march=native -ffast-math -fopenmp" ./configure CC=icc CFLAGS="-std=c99 -O3 -march=native -ffast-math -fopenmp -D__PURE_INTEL_C99_HEADERS__" ./configure
MPI support, nonportable binary: MPI support, nonportable binary:
CC=mpicc CFLAGS="-DUSE_MPI -std=c99 -O3 -march=native -ffast-math" ./configure CC=mpicc CFLAGS="-DUSE_MPI -std=c99 -O3 -march=native -ffast-math" ./configure