1
0
mirror of https://github.com/paboyle/Grid.git synced 2025-06-11 03:46:55 +01:00

Updating to modify non-inlining permute routines and hopefully get better reg use and

enhance performance.
This commit is contained in:
Peter Boyle
2015-09-25 08:55:04 -07:00
parent 5ef42add2d
commit 64d64d1ab6
11 changed files with 212 additions and 141 deletions

View File

@ -35,10 +35,10 @@ icpc-avx-openmp-mpi)
CXX=icpc ../../configure --enable-simd=AVX CXXFLAGS="-mavx -fopenmp -O3 -I/opt/local/include/openmpi-mp/ -std=c++11" LDFLAGS=-L/opt/local/lib/openmpi-mp/ LIBS="-lmpi -lmpi_cxx -fopenmp -lgmp -lmpfr" --enable-comms=mpi
;;
icpc-avx-openmp)
CXX=icpc ../../configure --enable-simd=AVX CXXFLAGS="-mavx -fopenmp -O3 -std=c++11" LIBS="-fopenmp -lgmp -lmpfr" --enable-comms=mpi
CXX=icpc ../../configure --enable-precision=single --enable-simd=AVX CXXFLAGS="-mavx -fopenmp -O3 -std=c++11" LIBS="-fopenmp -lgmp -lmpfr" --enable-comms=mpi
;;
icpc-avx2)
CXX=icpc ../../configure --enable-simd=AVX2 CXXFLAGS="-mavx2 -mfma -O3 -std=c++11" LIBS="-lgmp -lmpfr" --enable-comms=none
CXX=icpc ../../configure --enable-simd=AVX2 CXXFLAGS="-march=core-avx2 -O3 -std=c++11" LIBS="-lgmp -lmpfr" --enable-comms=none
;;
icpc-avx512)
CXX=icpc ../../configure --enable-simd=AVX512 CXXFLAGS="-xCOMMON-AVX512 -O3 -std=c++11" --host=none LIBS="-lgmp -lmpfr" --enable-comms=none
@ -50,7 +50,7 @@ icpc-mic-avx512)
CXX=icpc ../../configure --host=none --enable-simd=IMCI CXXFLAGS="-xCOMMON_AVX512 -O3 -std=c++11" LDFLAGS=-xCOMMON_AVX512 LIBS="-lgmp -lmpfr" --enable-comms=none
;;
clang-sse)
CXX=clang++ ../../configure --enable-simd=SSE4 CXXFLAGS="-msse4 -O3 -std=c++11" LIBS="-lgmp -lmpfr" --enable-comms=none
CXX=clang++ ../../configure --enable-precision=single --enable-simd=SSE4 CXXFLAGS="-msse4 -O3 -std=c++11" LIBS="-lgmp -lmpfr" --enable-comms=none
;;
clang-avx)
CXX=clang++ ../../configure --enable-simd=AVX CXXFLAGS="-mavx -O3 -std=c++11" LIBS="-lgmp -lmpfr" --enable-comms=none