guard prevents multiple A64FX build messages

Grid/qcd/action/fermion/implementation/WilsonKernelsAsmA64FX.h

@@ -39,16 +39,16 @@ Author: Nils Meyer  <nils.meyer@ur.de>  Regensburg University
 
 // enable A64FX body
 #define WILSONKERNELSASMBODYA64FX
-#pragma message("Including A64FX Dslash: WilsonKernelsAsmBodyA64FX.h")
+#pragma message("A64FX Dslash: WilsonKernelsAsmBodyA64FX.h")
 
     ///////////////////////////////////////////////////////////
     // If we are A64FX specialise the single precision routine
     ///////////////////////////////////////////////////////////
 #if defined(DSLASHINTRIN)
-#pragma message ("including A64FX Dslash: intrin")
+#pragma message ("A64FX Dslash: intrin")
 #include <simd/Fujitsu_A64FX_intrin_single.h>
 #else
-#pragma message ("including A64FX Dslash: asm")
+#pragma message ("A64FX Dslash: asm")
 #include <simd/Fujitsu_A64FX_asm_single.h>
 #endif
 

Grid/simd/Grid_a64fx-fixedsize.h

@@ -39,7 +39,7 @@
 
 // gcc 10 features
 #if __ARM_FEATURE_SVE_BITS==512
-#pragma message("Fixed-size SVE ACLE")
+#pragma message("building for A64FX / fixed SVE data type size")
 /* gcc 10.0.1 and gcc 10.1 bug using ACLE data types  CAS-159553-Y1K4C6
    workaround: use gcc's internal data types, bugfix expected for gcc 10.2
 typedef svbool_t    pred __attribute__((arm_sve_vector_bits(512)));

Grid/simd/Grid_vector_types.h

@@ -120,11 +120,15 @@ accelerator_inline Grid_half sfw_float_to_half(float ff) {
   #if defined(A64FX) || defined(A64FXFIXEDSIZE) // breakout A64FX SVE ACLE here
     #include <arm_sve.h>
     #if defined(A64FX) // VLA
+      #ifndef MSGVLA
       #pragma message("building for A64FX / SVE ACLE VLA")
+      #define MSGVLA
+      #endif
       #if defined(ARMCLANGCOMPAT)
+        #ifndef MSGCOMPAT
         #pragma message("applying armclang fix")
-      //#else
+        #define MSGCOMPAT
-        //#pragma message("not applying armclang fix")
+        #endif
       #endif
       #include "Grid_a64fx-2.h"
     #endif
@@ -132,7 +136,10 @@ accelerator_inline Grid_half sfw_float_to_half(float ff) {
       #include "Grid_a64fx-fixedsize.h"
     #endif
   #else
+    #ifndef MSGGEN
     #pragma message("building for GEN") // generic
+    #define MSGGEN
+    #endif
     #include "Grid_generic.h"
   #endif
 #endif

SVE_README.txt

@@ -8,6 +8,22 @@
 ../configure --with-lime=/home/men04359/lime/c-lime --without-hdf5 --enable-gen-simd-width=64 --enable-simd=GEN --enable-precision=double --enable-comms=none --enable-openmp CXX=g++-10.0.1 CC=gcc-10.0.1 CXXFLAGS="-std=c++11 -march=armv8-a+sve -msve-vector-bits=512 -fno-gcse -DA64FXFIXEDSIZE -DA64FXASM -DDSLASHINTRIN"
 
 
+* gcc 10.0.1 fixed-size ACLE (fjt) w/ MPI
+
+export OMPI_CC=gcc-10.0.1
+export OMPI_CXX=g++-10.0.1
+export MPICH_CC=gcc-10.0.1
+export MPICH_CXX=g++-10.0.1
+
+$ ../configure --without-hdf5 --enable-gen-simd-width=64 --enable-simd=GEN --enable-precision=double --enable-comms=mpi3 --enable-openmp CXX=mpiFCC CC=mpifcc CXXFLAGS="-std=c++11 -march=armv8-a+sve -msve-vector-bits=512 -fno-gcse -DA64FXFIXEDSIZE -DA64FXASM -DDSLASHINTRIN -DTOFU -I/opt/FJSVxtclanga/tcsds-1.2.24/include/mpi/fujitsu -lrt" LDFLAGS="-L/opt/FJSVxtclanga/tcsds-1.2.24/lib64 -lrt"
+
+works! but binaries do not finish when running via job scheduler. problem with MPI_finalize ?
+
+interactive login:  mpirun -np 1 ./Benchmark_wilson_sweep --dslash-asm
+                    [WARN] PLE 0610 plexec The process terminated with the signal.(rank=0)(nid=0xff010008)(sig=9) ?
+
+--------------------------------------------------------
+
 * armclang 20.0 VLA (merlin)
 
 ../configure --with-lime=/home/men04359/lime/c-lime --without-hdf5 --enable-gen-simd-width=64 --enable-simd=GEN --enable-precision=double --enable-comms=none --enable-openmp CXX=armclang++ CC=armclang CXXFLAGS="-std=c++11 -fno-unroll-loops -mllvm -vectorizer-min-trip-count=2 -march=armv8-a+sve -DARMCLANGCOMPAT -DA64FX -DA64FXASM -DDSLASHINTRIN" LDFLAGS=-static GRID_LDFLAGS=-static MPI_CXXLDFLAGS=-static
@@ -29,6 +45,14 @@ TODO check ARMCLANGCOMPAT
 TODO check ARMCLANGCOMPAT
 
 
+* armclang 20.1 VLA (fjt cluster)
+
+../configure --with-lime=$HOME/local --without-hdf5 --enable-gen-simd-width=64 --enable-simd=GEN --enable-precision=double --enable-comms=mpi3 --enable-openmp CXX=mpiFCC CC=mpifcc CXXFLAGS="-std=c++11 -mcpu=a64fx -DA64FX -DA64FXASM -DDSLASHINTRIN -DTOFU -I/opt/FJSVxtclanga/tcsds-1.2.24/include/mpi/fujitsu -lrt" LDFLAGS="-L/opt/FJSVxtclanga/tcsds-1.2.24/lib64"
+
+No ARMCLANGCOMPAT -> still correct ?
+
+--------------------------------------------------------
+
 * Fujitsu fcc
 
 ../configure --with-lime=$HOME/grid-a64fx/lime/c-lime --without-hdf5 --enable-gen-simd-width=64 --enable-simd=GEN --enable-precision=double --enable-comms=none --enable-openmp --with-mpfr=/home/users/gre/gre-1/grid-a64fx/mpfr-build/install CXX=FCC CC=fcc CXXFLAGS="-Nclang -Kfast -DA64FX -DA64FXASM -DDSLASHINTRIN"