compiling 3.1.4 on a cluster #12090

markcmiller86 · 2021-08-01T10:10:40Z

markcmiller86
Aug 1, 2021
Maintainer

Date: Mon, 12 Apr 2021 12:28:26 -0700
From: alex.razoumov at westgrid.ca (Alex Razoumov)
This post was imported from the visit-users@ornl.gov email archive

Hello,

Trying to compile VisIt 3.1.4 on our HPC cluster:

module load python/2.7.18 cmake/3.18.4 gcc/9.3.0 vtk/9.0.1
export PAR_COMPILER=mpicc
export PAR_INCLUDE='-I${EBROOTOPENMPI}/include'
export PAR_LIBS='-L${EBROOTOPENMPI}/lib -lmpi'
export CC=$EBROOTGCC/bin/gcc
export CXX=$EBROOTGCC/bin/g++
./build_visit3_1_4 --parallel --prefix $HOME/3.1.4 --system-cmake --system-python --mesagl --server-
components-only --llvm --thirdparty-path $HOME/3rdparty --cflags '-pthread' --alt-vtk-dir $EBROOTVTK

with

$ which mpicc
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler/gcc9/openmpi/4.0.3/bin/mpicc
$ echo $EBROOTOPENMPI
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler/gcc9/openmpi/4.0.3

I receive an error in build_visit3_1_4_log while compiling IceT:

-- Could NOT find MPI_C (missing: MPI_C_WORKS)
CMake Error at /cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/cmake/3.18.4/share/cmake-3.
18/Modules/FindPackageHandleStandardArgs.cmake:165?(message):
?Could NOT find MPI (missing: MPI_C_FOUND)
?Reason given by package: MPI component 'CXX' was requested, but language CXX is not enabled. MPI
component 'Fortran' was requested, but language Fortran is not enabled.

In the related icet-master-*/CMakeFiles/CMakeError.log I see:

make[1]: Entering directory '/tmp/razoumov/icet-
master-77c708f9090236b576669b74c53e9f105eedbd7e/CMakeFiles/CMakeTmp'
Building C object CMakeFiles/cmTC_38e51.dir/test_mpi.c.o
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/bin/gcc?-isystem
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler
/gcc9/openmpi/4.0.3/include -fPIC -m64 -fPIC -pthread -o CMakeFiles/cmTC_38e51.dir/test_mpi.c.o -c
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Co
re/cmake/3.18.4/share/cmake-3.18/Modules/FindMPI/test_mpi.c
Linking C executable cmTC_38e51
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/cmake/3.18.4/bin/cmake?-E
cmake_link_script CMakeFiles/cmTC_38e51.dir/link.txt --verbose=1
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/bin/gcc?-fPIC -m64 -fPIC
-pthread -rdynamic CMakeFiles/cmTC_38e51.dir/test_mpi.c.o
?-o cmTC_38e51 -L/tmp/razoumov/icet-master-77c708f9090236b576669b74c53e9f105eedbd7e
-Wl,-rpath,/tmp/razoumov/icet-master-77c708f9090236b576669b74c53e9f10
5eedbd7e /cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler/gcc9/openmpi/4.0.3/lib/l
ibmpi.so?-lfakempi
/cvmfs/soft.computecanada.ca/gentoo/2020/usr/x86_64-pc-linux-gnu/binutils-bin/2.33.1/ld.gold:?error:
cannot find -lfakempi
collect2: error: ld returned 1 exit status

Clearly, it is looking for MPI, but I explicitly set all the necessary variables above. Is my syntax
wrong somewhere?

Thank you,

---

Alex Razoumov
WestGrid Training & Visualization Coordinator, Compute Canada

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20210412/a0a6d162/attachment.htm>

markcmiller86 · 2021-08-01T10:10:41Z

markcmiller86
Aug 1, 2021
Maintainer Author

Date: Mon, 12 Apr 2021 13:10:04 -0700
From: notifications at github.com (Edward Rusu)
This post was imported from the visit-users@ornl.gov email archive

We're trying to [combine `visit-users` email list with GitHub issues](https://github.com/visit-
dav/live-customer-response/wiki/How-the-new-GitHub-visit-users-Email-Integration-Works).
When replying on `visit-users`, please reply **only** to emails with a *Subject*
line that includes `[visit-dav/live-customer-response]`.

---

Hello,

Trying to compile VisIt 3.1.4 on our HPC cluster:

module load python/2.7.18 cmake/3.18.4 gcc/9.3.0 vtk/9.0.1
export PAR_COMPILER=mpicc
export PAR_INCLUDE='-I${EBROOTOPENMPI}/include'
export PAR_LIBS='-L${EBROOTOPENMPI}/lib -lmpi'
export CC=$EBROOTGCC/bin/gcc
export CXX=$EBROOTGCC/bin/g++
./build_visit3_1_4 --parallel --prefix $HOME/3.1.4 --system-cmake --system-python --mesagl --server-
components-only --llvm --thirdparty-path $HOME/3rdparty --cflags '-pthread' --alt-vtk-dir $EBROOTVTK

with

$ which mpicc
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler/gcc9/openmpi/4.0.3/bin/mpicc
$ echo $EBROOTOPENMPI
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler/gcc9/openmpi/4.0.3

I receive an error in build_visit3_1_4_log while compiling IceT:

-- Could NOT find MPI_C (missing: MPI_C_WORKS)
CMake Error at /cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/cmake/3.18.4/share/cmake-3.
18/Modules/FindPackageHandleStandardArgs.cmake:165 (message):
Could NOT find MPI (missing: MPI_C_FOUND)
Reason given by package: MPI component 'CXX' was requested, but language CXX is not enabled. MPI
component 'Fortran' was requested, but language Fortran is not enabled.

In the related icet-master-*/CMakeFiles/CMakeError.log I see:

make[1]: Entering directory '/tmp/razoumov/icet-
master-77c708f9090236b576669b74c53e9f105eedbd7e/CMakeFiles/CMakeTmp'
Building C object CMakeFiles/cmTC_38e51.dir/test_mpi.c.o
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/bin/gcc -isystem
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler
/gcc9/openmpi/4.0.3/include -fPIC -m64 -fPIC -pthread -o CMakeFiles/cmTC_38e51.dir/test_mpi.c.o -c
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Co
re/cmake/3.18.4/share/cmake-3.18/Modules/FindMPI/test_mpi.c
Linking C executable cmTC_38e51
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/cmake/3.18.4/bin/cmake -E
cmake_link_script CMakeFiles/cmTC_38e51.dir/link.txt --verbose=1
/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/bin/gcc -fPIC -m64 -fPIC
-pthread -rdynamic CMakeFiles/cmTC_38e51.dir/test_mpi.c.o
-o cmTC_38e51 -L/tmp/razoumov/icet-master-77c708f9090236b576669b74c53e9f105eedbd7e
-Wl,-rpath,/tmp/razoumov/icet-master-77c708f9090236b576669b74c53e9f10
5eedbd7e
/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Compiler/gcc9/openmpi/4.0.3/lib/libmpi.so
-lfakempi
/cvmfs/soft.computecanada.ca/gentoo/2020/usr/x86_64-pc-linux-gnu/binutils-bin/2.33.1/ld.gold: error:
cannot find -lfakempi
collect2: error: ld returned 1 exit status

Clearly, it is looking for MPI, but I explicitly set all the necessary variables above. Is my syntax
wrong somewhere?

Thank you,

---

Alex Razoumov
WestGrid Training & Visualization Coordinator, Compute Canada

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/243

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20210412/a0fc33d9/attachment.htm>

0 replies

markcmiller86 · 2021-08-01T10:10:42Z

markcmiller86
Aug 1, 2021
Maintainer Author

Date: Mon, 12 Apr 2021 13:27:44 -0700
From: notifications at github.com (Edward Rusu)
This post was imported from the visit-users@ornl.gov email archive

Hi Alex,

Sorry for your trouble building visit. I believe the bug you are experiencing was captured in
[5318](https://github.com/visit-dav/visit/issues/5318), for which we have already [merged a
fix](https://github.com/visit-dav/visit/pull/5413) for our development branch. This fix does not
appear in 3.1.4 since that was released in November 2020 and the fix was applied in January 2021.
However, the fix is very simple, just a couple of changes to bv_icet.sh. Can you copy [these
changes](https://github.com/visit-
dav/visit/pull/5413/files#diff-9809ab91080cb4909e54d55f35bba43e0a8abe73afe2d985530fe3ae4e394aa4) to
your build script and let me know if it works for you?

Thanks,
Eddie

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/243#issuecomment-818176559

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20210412/6422cf05/attachment.htm>

0 replies

markcmiller86 · 2021-08-01T10:10:42Z

markcmiller86
Aug 1, 2021
Maintainer Author

Date: Mon, 12 Apr 2021 13:49:27 -0700
From: notifications at github.com (Kathleen Biagas)
This post was imported from the visit-users@ornl.gov email archive

Hi Alex,

You should know as well that VisIt 3.1.4 is not compatible with VTK-9. You need VTK-8.1.0. We
recommend using build_visit to build VTK since we apply necessary patches during the build.

Kathleen

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/243#issuecomment-818221376

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20210412/6141b5f2/attachment.htm>

0 replies

markcmiller86 · 2021-08-01T10:10:43Z

markcmiller86
Aug 1, 2021
Maintainer Author

Date: Fri, 16 Apr 2021 21:07:12 -0700
From: notifications at github.com (Edward Rusu)
This post was imported from the visit-users@ornl.gov email archive

Closed #243.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/243#event-XXXXXXXXXX

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20210416/0e83679f/attachment.htm>

0 replies

markcmiller86 · 2021-08-01T10:10:43Z

markcmiller86
Aug 1, 2021
Maintainer Author

Date: Thu, 22 Apr 2021 08:25:33 -0700
From: alex.razoumov at westgrid.ca (Alex Razoumov)
This post was imported from the visit-users@ornl.gov email archive

Hello,

I finally managed to compile VisIt 3.1.4 on our HPC cluster with

$ ./build_visit3_1_4 --parallel --prefix /path/to/installation --cmake --mesagl
? ? ?--llvm --thirdparty-path /path/to/3rdparty --vtk --python
? ? ?--netcdf --hdf5 --silo --server-components-only

The client-server mode works well. Contrary to my naive expectations, this did not build the CLI
tools:

$ /path/to/installation/visit -nowin -cli -s test.py
Version 3.1.4 of cli does not exist for the architecture linux-x86_64.

I would like to avoid having to build the full Qt version with the GUI, as we need only (1) the
server for client-server workflows and (2) the engine to run off-screen rendering scripts from the
command line without opening any windows. Is there a way to do that without compiling the full GUI?

Thank you,

---

Alex Razoumov
WestGrid Training & Visualization Coordinator, Compute Canada

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20210422/0d63a8ab/attachment.htm>

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

compiling 3.1.4 on a cluster #12090

{{title}}

Replies: 5 comments

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

compiling 3.1.4 on a cluster #12090

markcmiller86 Aug 1, 2021 Maintainer

Replies: 5 comments

markcmiller86 Aug 1, 2021 Maintainer Author

markcmiller86 Aug 1, 2021 Maintainer Author

markcmiller86 Aug 1, 2021 Maintainer Author

markcmiller86 Aug 1, 2021 Maintainer Author

markcmiller86 Aug 1, 2021 Maintainer Author

markcmiller86
Aug 1, 2021
Maintainer

markcmiller86
Aug 1, 2021
Maintainer Author

markcmiller86
Aug 1, 2021
Maintainer Author

markcmiller86
Aug 1, 2021
Maintainer Author

markcmiller86
Aug 1, 2021
Maintainer Author

markcmiller86
Aug 1, 2021
Maintainer Author