8.1 release installation error

Luca

unread,

Apr 6, 2021, 1:09:06 PM4/6/21

to cp2k

Dear Developers

I am trying to install the release 8.1 with intel compilers. Unfortunately, I get the error

cp2k-8.1/src/grid/grid_replay.c(107): error: expected an expression

for (int i = 0; i < 3; i++) {

I tested both the intel compilers 2018 and 2020 with the arch file Linux-x86-64-intel-minimal.psmp

Which way I can fix the issue?

Best Regards.

Luca

Krack Matthias (PSI)

unread,

Apr 6, 2021, 4:19:21 PM4/6/21

to cp2k

Dear Luca

just with that information, it is difficult to guess what is causing the problem. I have not seen such error yet. Could you provide more information about your system and how you compiled cp2k 8.1?

Best

Matthias

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/11df7c90-3b52-4051-9469-85d5f76ff1a4n%40googlegroups.com.

Luca

unread,

Apr 6, 2021, 6:36:43 PM4/6/21

to cp2k

Dear Matthias

thank you for your answer. What I did is

1) set the intel environment with

source /intel/compilers_and_libraries_2020.2.254/linux/bin/ifortvars.sh -arch intel64 -platform linux

source /intel/compilers_and_libraries_2020.2.254/linux/bin/iccvars.sh -arch intel64 -platform linux

source /intel/compilers_and_libraries_2020.2.254/linux/mkl/bin/mklvars.sh intel64

source /intel/compilers_and_libraries_2020.2.254/linux/mpi/intel64/bin/mpivars.sh intel64

2) Python3 is ~/LIBS/PYTHON/anaconda3/bin/python

3) make -j8 ARCH=Linux-x86-64-intel-minimal VERSION=psmp

the arch file is the one you can find into the arch dir. (I did not modify any line)

My OS is

NAME="CentOS Linux"

VERSION="7 (Core)"

ID="centos"

ID_LIKE="rhel fedora"

VERSION_ID="7"

PRETTY_NAME="CentOS Linux 7 (Core)"

ANSI_COLOR="0;31"

CPE_NAME="cpe:/o:centos:centos:7"

HOME_URL="https://www.centos.org/"

BUG_REPORT_URL="https://bugs.centos.org/"

I also tried with the intel compiler intel2018/compilers_and_libraries_2018.5.274 and the compilation always gets stuck at

mpiifort -c -O2 -g -qopenmp -traceback -xHost -D__FFTW3 -D__MAX_CONTR=4 -D__MKL -D__MPI_VERSION=3 -D__parallel -D__SCALAPACK -diag-disable=8291 -diag-disable=8293 -fpp -fp-model precise -free -funroll-loops -I/shared/Libs/intel2018/compilers_and_libraries_2018.5.274/linux/mkl/include -I/shared/Libs/intel2018/compilers_and_libraries_2018.5.274/linux/mkl/include/fftw -D__COMPILE_ARCH="\"Linux-x86-64-intel-minimal\"" -D__COMPILE_DATE="\"mar 6 apr 2021, 18.37.16, CEST\"" -D__COMPILE_HOST="\"ngrowth1\"" -D__COMPILE_REVISION="\"git:0b61f2f\"" -D__DATA_DIR="\"/home/ldau/CODES/INTEL/18/build/HUT/cp2k-8.1/data\"" -D__SHORT_FILE__="\"common/kahan_sum.F\"" -I'/home/ldau/CODES/INTEL/18/build/HUT/cp2k-8.1/src/common/' -I'/home/ldau/CODES/INTEL/18/build/HUT/cp2k-8.1/obj/Linux-x86-64-intel-minimal/psmp/exts/dbcsr' kahan_sum.F90

/home/ldau/CODES/INTEL/18/build/HUT/cp2k-8.1/src/grid/grid_replay.c(107): error: expected an expression

for (int i = 0; i < 3; i++) {

^

/home/ldau/CODES/INTEL/18/build/HUT/cp2k-8.1/src/grid/grid_replay.c(107): error: identifier "i" is undefined

for (int i = 0; i < 3; i++) {

^

Best Regards

Luca

Krack Matthias (PSI)

unread,

Apr 7, 2021, 5:36:29 AM4/7/21

to cp2k

Dear Luca

It seems that your cp2k installation stops with the compilation of the first C code file. It is possible that your Intel C compiler refers to a gcc compiler version which is too old like 4.x. Could you run after the setup of the Intel compiler environment: “mpiicc –v” and “icc –v”?

If that shows something like “icc version xx (gcc version 4.8.5 compatibility)” then the gcc compiler is probably too old. You may add the “-gcc-name <path_to_newer_gcc_version>” flag to point to a newer gcc version.

HTH

Matthias

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/8d71b7ac-d751-48df-b2bf-d2a99260940bn%40googlegroups.com.

Luca

unread,

Apr 8, 2021, 1:02:33 PM4/8/21

to cp2k

Dear Matthias

thank you for your advice. I will try to update the gcc release.

Best regards.

Luca

ganta.pra...@gmail.com

unread,

Apr 22, 2021, 6:24:41 PM4/22/21

to cp2k

Dear Dr. Matthias,

I also face similar issue. Can you please tell me where exactly should I add -gcc-name=/path/to/new/gcc/version. My archfile is as below:

CC = mpiicc -qopenmp

CXX = mpiicpc -qopenmp

CPP =

FC = mpiifort -qopenmp

LD = mpiifort -qopenmp

AR = ar -r

export OMP_STACKSIZE="50M"

include /home/xx/plumed/intel-hsw/lib/plumed/src/lib/Plumed.inc

EXTERNAL_OBJECTS=$(PLUMED_STATIC_DEPENDENCIES)

LIBXC_DIR = /home/xx/libxc/

LIBINT_DIR = /home/xx/libint/

ELPA_DIR = /home/xx/elpa/

DFLAGS = -D__MKL -D__FFTW3 -D__ELPA -D__LIBINT -D__LIBXC -D__parallel \

-D__SCALAPACK -D__BLACS -D__LIBINT_MAX_AM=6 -D__LIBDERIV_MAX_AM1=5 \

-D__PLUMED2 -D__MAX_CONTR=4 -D__MPI_VERSION=3

CFLAGS = $(DFLAGS)

FCFLAGS = $(DFLAGS) -O2 -g -traceback -fpp -free -no-ipo \

-I$(MKLROOT)/include -I$(MKLROOT)/include/fftw -I$(LIBINT_DIR)/include \

-I$(LIBXC_DIR)/include -I$(ELPA_DIR)/include/elpa_openmp-2020.11.001/elpa \

-I$(ELPA_DIR)/include/elpa_openmp-2020.11.001/modules

FCFLAGS2 = $(DFLAGS) -O1 -g -traceback -fpp -free -no-ipo \

-I$(MKLROOT)/include -I$(MKLROOT)/include/fftw -I$(LIBINT_DIR)/include \

-I$(LIBXC_DIR)/include -I$(ELPA_DIR)/include/elpa_openmp-2020.11.001/elpa \

-I$(ELPA_DIR)/include/elpa_openmp-2020.11.001/modules

FCFLAGS3 = $(DFLAGS) -O0 -g -traceback -fpp -free -no-ipo \

-I$(MKLROOT)/include -I$(MKLROOT)/include/fftw -I$(LIBINT_DIR)/include \

-I$(LIBXC_DIR)/include -I$(ELPA_DIR)/include/elpa_openmp-2020.11.001/elpa \

-I$(ELPA_DIR)/include/elpa_openmp-2020.11.001/modules

LDFLAGS = $(FCFLAGS) -L$(ELPA_DIR) -static-intel

LDFLAGS_C = $(FCFLAGS) -L$(ELPA_DIR) -static-intel -nofor_main

LIBS = -L$(ELPA_DIR)/lib -lelpa_openmp \

$(MKLROOT)/lib/intel64/libmkl_scalapack_lp64.a \

-Wl,--start-group $(MKLROOT)/lib/intel64/libmkl_intel_lp64.a \

$(MKLROOT)/lib/intel64/libmkl_sequential.a \

$(MKLROOT)/lib/intel64/libmkl_core.a \

$(MKLROOT)/lib/intel64/libmkl_blacs_intelmpi_lp64.a -Wl,--end-group \

-lpthread -lm -ldl \

$(LIBINT_DIR)/lib/libderiv.a $(LIBINT_DIR)/lib/libint.a -lstdc++ \

$(LIBXC_DIR)/lib/libxcf03.a $(LIBXC_DIR)/lib/libxc.a \

-lz -ldl -lstdc++ /home/xx/plumed/intel-hsw/lib/libplumed.a

# Required due to memory leak that occurs if high optimisations are used

mp2_optimize_ri_basis.o: mp2_optimize_ri_basis.F

$(FC) -c $(FCFLAGS) $<

ganta.pra...@gmail.com

unread,

Apr 23, 2021, 2:48:31 AM4/23/21

to cp2k

Dear Dr. Matthias,

I updated the below lines in Arch file with new GCC name and the compilation didn't stop as before.

CC = mpiicc -qopenmp -gcc-name=/opt/common/compiler/gnu/8.2.0/bin/gcc

CXX = mpiicpc -qopenmp -gcc-name=/opt/common/compiler/gnu/8.2.0/bin/gcc

CPP =

FC = mpiifort -qopenmp -gcc-name=/opt/common/compiler/gnu/8.2.0/bin/gcc

Thank you.

Reply all

Reply to author

Forward