ELPA_GPU gitlab buildtest: test all real/complex SSE/AVX kernel

.gitlab-ci.yml

@@ -203,3 +203,80 @@ single-precision-mpi-openmp-ftimings-redirect-real-generic-simple-complex-generi
     - export OMP_NUJM_THREADS=2
     - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
     - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-sse_assembly-complex-sse_assembly-kernel-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-generic-kernel-only --with-complex-generic-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-sse_block2-complex-sse_block1-kernel-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-sse_block2-kernel-only --with-complex-sse_block1-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-sse_block4-complex-sse_block2-kernel-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-sse_block4-kernel-only --with-complex-sse_block2-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-sse_block6-complex-avx_block1-kernel-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-sse_block6-kernel-only --with-complex-avx_block1-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-avx_block2-complex-avx_block2-kernel-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-avx_block2-kernel-only --with-complex-avx_block2-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-avx_block4-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-avx_block4-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'
+
+single-precision-mpi-openmp-ftimings-redirect-real-avx_block6-jobs:
+  script:
+    - export LANG=C
+    - module load impi intel gcc mkl autotools
+    - ./autogen.sh
+    - ./configure CFLAGS="-O3 -mavx" CXXFLAGS="-O3 -mavx" FCFLAGS="-O3 -mavx" SCALAPACK_LDFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -Wl,-rpath,$MKL_HOME/lib/intel64" SCALAPACK_FCFLAGS="-L/afs/@cell/common/soft/intel/ics2015/15.0/mkl/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -I$MKLROOT/include/intel64/lp64" --enable-openmp --enable-single-precision --with-ftimings --with-redirect --with-real-avx_block6-kernel-only
+    - make -j 8
+    - export OMP_NUJM_THREADS=2
+    - export LD_LIBRARY_PATH=$MKL_HOME/lib/intel64:$LD_LIBRARY_PATH
+    - make check TEST_FLAGS='1500 50 16'