Compilation BGW 4.0 Daint alps

[Andres Ortega]

Dear all, 

I am trying to compile BGW 4.0 in the daint alps. that has GH200 GPU and have ARM cores. 

I was wondering if maybe you could give some adviced to solve the following error 

ln -sf ../Common/sort_m.mod .
make[4]: Leaving directory '/users/aortega/BerkeleyGW-4.0/library'
cd Epsilon && make
make[4]: Entering directory '/users/aortega/BerkeleyGW-4.0/Epsilon'
cpp -C -nostdinc -nostdinc++ -I../Common  -DNVHPC -DNVHPC_API -DNVIDIA_GPU -DMPI  -DOMP -DUSESCALAPACK -DUNPACKED -DUSEFFTW3 -DOMP_TARGET -DOPENACC -DHDF5   algos_epsilon.f90 > algos_epsilon.p.f
cc1: warning: command-line option '-nostdinc++' is valid for C++/ObjC++ but not for C
algos_epsilon.f90:100:23: warning: missing terminating ' character
  100 |     ! Then check if we're using the GPU
      |                       ^
algos_epsilon.f90:103:16: warning: missing terminating " character
  103 |       call die("You have specified one or more GPU-accelerated algorithms, but&
      |                ^
algos_epsilon.f90:104:41: warning: missing terminating " character
  104 |                & the GPU is not enabled.", only_root_writes=.true.)
      |                                         ^
algos_epsilon.f90:107:45: warning: missing terminating ' character
  107 |     ! Finally, check for algorithms that don't play nicely with one another
      |                                             ^
mpif90 -Mfree -acc -mp=multicore,gpu -gpu=cc90  -cudalib=cublas,cufft -traceback -Minfo=all,mp,acc -gopt -traceback -I./Common -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include -I./Common -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include ../Common -I./Common -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include /user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include/ -I./Common -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include /user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include/ -c -fast -Mfree -Mlarge_arrays algos_epsilon.p.f -o algos_epsilon.o -module ./
algos_epsilon.p.f:
NVFORTRAN-F-0004-Unable to open MODULE file algos_common_m.mod (algos_epsilon.f90: 10)
NVFORTRAN/arm64 Linux 24.11-0: compilation aborted
make[4]: *** [../Common/common-rules.mk:282: algos_epsilon.o] Error 2
make[4]: Leaving directory '/users/aortega/BerkeleyGW-4.0/Epsilon'
make[3]: *** [Makefile:82: make-Epsilon] Error 2
make[3]: Leaving directory '/users/aortega/BerkeleyGW-4.0'
make[2]: *** [Makefile:115: all] Error 2
make[2]: Leaving directory '/users/aortega/BerkeleyGW-4.0'
make[1]: *** [Makefile:126: real] Error 2
make[1]: Leaving directory '/users/aortega/BerkeleyGW-4.0'
make: *** [Makefile:119: all-flavors] Error 2

the arch.mk file I am using is this one, 

COMPFLAG  = -DNVHPC -DNVHPC_API -DNVIDIA_GPU
PARAFLAG  = -DMPI  -DOMP
MATHFLAG  = -DUSESCALAPACK -DUNPACKED -DUSEFFTW3 -DOMP_TARGET -DOPENACC -DHDF5 # -DUSEPRIMME # -DUSEELPA # USEELPA_GPU -DHDF5
# DEBUGFLAG = -DDEBUG -DNVTX
#

NVCC=nvcc
NVCCOPT= -O3 -use_fast_math
CUDALIB= -lcufft -lcublasLt -lcublas -lcudart -lcuda -lnvToolsExt

FCPP = cpp -C -nostdinc -nostdinc++
#FCPP    = cpp  -P -ansi  -nostdinc  -C  -E  -std=c11 #-nostdinc++
F90free = mpif90 -Mfree -acc -mp=multicore,gpu -gpu=cc90  -cudalib=cublas,cufft -traceback -Minfo=all,mp,acc -gopt -traceback $(INCFLAG)

LINK = mpif90 -acc -mp=multicore,gpu -gpu=cc90 -cudalib=cublas,cufft -Minfo=mp,acc \
      -L$(HDF5_LDIR) -lhdf5_hl_fortran -lhdf5_hl -lhdf5_fortran -lhdf5 -lz -ldl


#LINK    = mpif90       -acc -mp=multicore,gpu -gpu=cc90  -cudalib=cublas,cufft -Minfo=mp,acc # -lnvToolsExt  
FOPTS   = -fast -Mfree -Mlarge_arrays
FNOOPTS = $(FOPTS)
MOD_OPT = -module
INCFLAG = -I./Common \
          -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include \
          -I/user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include


#INCFLAG = -I./

C_PARAFLAG  = -DPARA -DMPICH_IGNORE_CXX_SEEK
CC_COMP = mpicxx #-fopenmp
C_COMP  = mpicc
C_LINK  = mpicxx
C_OPTS  = -mp -fast
C_DEBUGFLAG =

REMOVE  = /bin/rm -f

FFTW_DIR= /user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/lib
FFTWLIB      = $(FFTW_DIR)/libfftw3.so \
               $(FFTW_DIR)/libfftw3_threads.so \
               $(FFTW_DIR)/libfftw3_omp.so \
               ${CUDALIB}  -lstdc++
FFTWINCLUDE  = /user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/fftw-3.3.10-zll77ggy54brpmsycpnzph4lshlfemn5/include/
PERFORMANCE  =

SCALAPACKLIB =
LAPACKLIB =

HDF5_LDIR    =  /user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/lib/
HDF5LIB      =  $(HDF5_LDIR)/libhdf5_hl_fortran.a \
                $(HDF5_LDIR)/libhdf5_hl.a \
                $(HDF5_LDIR)/libhdf5_fortran.a \
                $(HDF5_LDIR)/libhdf5.a -lz -ldl
HDF5INCLUDE  = /user-environment/linux-sles15-neoverse_v2/nvhpc-24.11/hdf5-1.14.5-nbohvzmuykfinybor3pgkx532ixdnggo/include/