I patched this as much as I could, but it's still not compiling due an error in one of the solvers, which is written in Fortran. Help is really welcome.
That said, this release is so old that it's probably best to just use the git version.
Git Clone URL: | https://aur.archlinux.org/elmerfem.git (read-only, click to copy) |
---|---|
Package Base: | elmerfem |
Description: | A finite element software for multiphysical problems |
Upstream URL: | http://www.elmerfem.org |
Keywords: | CFD FEM simulation |
Licenses: | GPL |
Conflicts: | elmerfem-git |
Submitter: | saxonbeta |
Maintainer: | rmsc |
Last Packager: | rmsc |
Votes: | 12 |
Popularity: | 0.000014 |
First Submitted: | 2015-08-18 13:50 (UTC) |
Last Updated: | 2025-05-23 21:15 (UTC) |
I patched this as much as I could, but it's still not compiling due an error in one of the solvers, which is written in Fortran. Help is really welcome.
That said, this release is so old that it's probably best to just use the git version.
In newer versions of GCC, implicit function declaration in C code is regarded as error by default. -Wno-error=implicit-function-declaration
is required to compile Elmer FEM successfully.
ElmerGUI doesn't terminate when we exit it. We have to kill the process.
For some reason, some tests were failing for me, then I checked the logfile at the end of the output and found a ompi error that led me here: https://github.com/open-mpi/ompi/issues/5613.
Even tho it's closed, I modified my $PATH to remove /bin
(so command which
is forced to use /usr/bin
) and for some reason it worked.
@rmsc, if you depend on mumps-cmake, you can build this with -D WITH_Mumps=ON
@rmsc thankyou! this worked for me!
I think it's best to just fix the PKGBUILD. Just add this line at the end of the prepare()
function:
sed -i 's#mmg/mmg3d/libmmgtypesf.h#mmg/common/libmmgtypesf.h#g' elmerice/Solvers/CalvingRemeshMMG.F90
Actually, I got the same problem. The file libmmgtypesf.h is now located in mmg/common/ folder not in mmg/mmg3d/ . There is 2 options, you can change the include statement in the file ClavingRemeshMMG.F90 to the correct folder and do this whenever you encounter the error. Or, do what I did, after mmg was installed by the installation routine, and the routine stopped here with this error; I went to the already installed mmg include folder and I made a softlink in mmg/mmg3d to the mmg/common/libmmgtypesf.h and rerun the installation. This time without problems... mmg include folder for me was located in /usr/include/mmg/, however it might be in /usr/local/include/mmg/ or even somewhere else...
Getting this error
Warnung: Unverträgliche Typen zwischen Argument bei (1) und Argument bei (2) (REAL(8)/INTEGER(4)).
[ 59%] Building Fortran object elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/CalvingHydroInterp.F90.o
[ 59%] Building Fortran object elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/HydroRestart.F90.o
[ 59%] Building Fortran object elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/GMValid.F90.o
[ 59%] Building Fortran object elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/CalvingRemeshMMG.F90.o
/home/monsoon/.cache/yay/elmerfem/src/elmerfem-release-9.0/elmerice/Solvers/CalvingRemeshMMG.F90:65:2:
65 |
| 1
Schwerwiegender Fehler: mmg/mmg3d/libmmgtypesf.h: Datei oder Verzeichnis nicht gefunden
Kompilierung beendet.
make[2]: *** [elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/build.make:1089: elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/CalvingRemeshMMG.F90.o] Fehler 1
make[1]: *** [CMakeFiles/Makefile2:14558: elmerice/Solvers/CMakeFiles/ElmerIceSolvers.dir/all] Fehler 2
make: *** [Makefile:166: all] Fehler 2
==> FEHLER: Ein Fehler geschah in build().
Breche ab...
-> Fehler beim Erstellen: elmerfem-exit status 4
-> Failed to install the following packages. Manual intervention is required:
elmerfem - exit status 4
I can only find this information:
https://github.com/MmgTools/mmg/issues/66
Would love to have this on my CachyOS
Linux glassplanet 6.3.3-1-cachyos #1 SMP PREEMPT_DYNAMIC Wed, 17 May 2023 19:10:26 +0000 x86_64 GNU/Linux
Getting the following error:
[ 64%] Built target Mesh2D
Scanning dependencies of target mpi_stubs
[ 64%] Building Fortran object fem/src/CMakeFiles/mpi_stubs.dir/mpif_stubs.F90.o
[ 64%] Linking Fortran shared library libmpi_stubs.so
[ 64%] Built target mpi_stubs
Scanning dependencies of target Solver_TGT
[ 64%] Building Fortran object fem/src/CMakeFiles/Solver_TGT.dir/Solver.F90.o
[ 64%] Linking Fortran executable ElmerSolver_mpi
/usr/bin/ld: warning: libfmt.so.8, needed by /usr/lib/libvtkFiltersGeneral.so.1, not found (try using -rpath or -rpath-link)
/usr/bin/ld: /usr/lib/libvtkFiltersGeneral.so.1: undefined reference to `fmt::v8::vformat[abi:cxx11](fmt::v8::basic_string_view<char>, fmt::v8::basic_format_args<fmt::v8::basic_format_context<fmt::v8::appender, char> >)'
collect2: error: ld returned 1 exit status
make[2]: *** [fem/src/CMakeFiles/Solver_TGT.dir/build.make:117: fem/src/ElmerSolver_mpi] Error 1
make[1]: *** [CMakeFiles/Makefile2:15395: fem/src/CMakeFiles/Solver_TGT.dir/all] Error 2
make: *** [Makefile:166: all] Error 2
==> ERROR: A failure occurred in build().
Aborting...
Full build log: https://github.com/arch4edu/cactus/runs/7178575430?check_suite_focus=true
Pinned Comments
rmsc commented on 2025-05-23 21:37 (UTC)
I patched this as much as I could, but it's still not compiling due an error in one of the solvers, which is written in Fortran. Help is really welcome.
That said, this release is so old that it's probably best to just use the git version.