[petsc-dev] issue with v3.21.5 build
Jeffrey Hyman
hymanjd at gmail.com
Tue Apr 29 09:27:42 CDT 2025
Oh, I remember that now. conda used to be so helpful. It always gets me
when I try to rebuild.
I removed it and got stuck in hdf5 again. Anaconda is gone, but it looks
lille mpif90 was not found.
configure:7444: mpif90 -c -ffree-line-length-none -ffree-line-length-0
-Wno-lto-type-mismatch -O3 -Wno-unused-function -fallow-argument-mismatch
-fallow-argument-mismatch conftest.f90 >&5
./configure: line 2482: mpif90: command not found
should I try --download-mpich=
https://urldefense.us/v3/__https://www.mpich.org/static/downloads/4.3.0/mpich-4.3.0.tar.gz__;!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGOy1WhdS2$ . that
Satish suggested?
J
On Tue, Apr 29, 2025 at 10:12 AM Matthew Knepley <knepley at gmail.com> wrote:
> On Tue, Apr 29, 2025 at 10:09 AM Jeffrey Hyman <hymanjd at gmail.com> wrote:
>
>> did 1 - 4, now I'm getting an error configuring hdf5. configure.log
>> attached,
>>
>
> I forgot to look at your environment. This is what was causing your MPI
> issues. Always turn off all Anaconda stuff in the env when building PETSc.
>
> configure:7444: mpif90 -c -ffree-line-length-none -ffree-line-length-0
> -Wno-lto-type-mismatch -O3 -Wno-unused-function -fallow-argument-mismatch
> -fallow-argument-mismatch -x f95 conftest.f90 >&5
>
> /Users/hymanjd/opt/anaconda3/bin/mpif90: line 392:
> x86_64-apple-darwin13.4.0-gfortran: command not found
>
> configure:7444: $? = 127
>
> configure: failed program was:
>
> | program main
>
> |
>
> | end
>
> configure:7454: result: unknown
>
> configure:7457: error: Fortran could not compile .f90 files
>
> It has ceased to be a force for good.
>
> Thanks,
>
> Matt
>
> On Tue, Apr 29, 2025 at 9:41 AM Matthew Knepley <knepley at gmail.com> wrote:
>>
>>> On Tue, Apr 29, 2025 at 9:28 AM Jeffrey Hyman <hymanjd at gmail.com> wrote:
>>>
>>>> I know he and Richard have it on the to do list. But I'll encourage
>>>> them at a meeting tomorrow. : )
>>>>
>>>> That file for the make to work, but the check failed. Looks like MPI
>>>> might have gotten crossed up in all the installs?
>>>>
>>>
>>> Yes, I would
>>>
>>> 1) Preserve the reconfigure-***.py and gmakegen.py
>>>
>>> 2) Remove the whole $PETSC_DIR
>>>
>>> 3) Get rid of the preinstalled OpenMPI you have. It is usually somewhat
>>> broken and not helping.
>>>
>>> 4) Reclone and rebuild
>>>
>>> THanks,
>>>
>>> Matt
>>>
>>>
>>>> Thanks, Matt! Feels close.
>>>> J
>>>>
>>>>
>>>> On Tue, Apr 29, 2025 at 9:20 AM Matthew Knepley <knepley at gmail.com>
>>>> wrote:
>>>>
>>>>> On Tue, Apr 29, 2025 at 9:17 AM Jeffrey Hyman <hymanjd at gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Unfortunately, Pflotran hasn't been updated to use 3.22 yet. So I'm
>>>>>> stuck with 3.21.5. Can I hack the python path?
>>>>>>
>>>>>
>>>>> Tell Glenn to get in gear :)
>>>>>
>>>>> I think you can just get this file:
>>>>>
>>>>> https://urldefense.us/v3/__https://gitlab.com/petsc/petsc/-__;!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO3a5JDt1$
>>>>> /blob/release-3.22/config/gmakegen.py?ref_type=heads
>>>>>
>>>>> and put it in $PETSC_DIR/config/gmakegen.py
>>>>>
>>>>> Finger crossed,
>>>>>
>>>>> Matt
>>>>>
>>>>>
>>>>>> J
>>>>>>
>>>>>>
>>>>>> On Tue, Apr 29, 2025 at 9:14 AM Matthew Knepley <knepley at gmail.com>
>>>>>> wrote:
>>>>>>
>>>>>>> On Tue, Apr 29, 2025 at 9:08 AM Jeffrey Hyman <hymanjd at gmail.com>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> This configure worked
>>>>>>>>
>>>>>>>> ./configure --COPTFLAGS='-O3' --CXXOPTFLAGS='-O3' --FOPTFLAGS='-O3
>>>>>>>> -Wno-unused-function -fallow-argument-mismatch' --with-debugging=no
>>>>>>>> --download-mpich=yes --download-hdf5=yes
>>>>>>>> --download-hdf5-fortran-bindings=yes --download-fblaslapack=yes
>>>>>>>> --download-metis=yes --download-parmetis=yes --download-cmake=
>>>>>>>> https://urldefense.us/v3/__https://gitlab.kitware.com/cmake/cmake/-/archive/v3.31.7/cmake-v3.31.7.tar.gz__;!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO6cDqHoB$
>>>>>>>> \
>>>>>>>> --download-cmake-cc=clang --download-cmake-cxx=clang++
>>>>>>>>
>>>>>>>> but then it failed on the make. configure and make attached.
>>>>>>>>
>>>>>>>
>>>>>>> We could backport a check for this. The Python is newer than the
>>>>>>> release 3.21.5, so we are using something that went away. If you use 3.22
>>>>>>> or higher this is fixed.
>>>>>>>
>>>>>>> Thanks,
>>>>>>>
>>>>>>> Matt
>>>>>>>
>>>>>>>
>>>>>>>> Thanks again for the help. Much appreciated.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Tue, Apr 29, 2025 at 8:53 AM Jeffrey Hyman <hymanjd at gmail.com>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> I updated brew llvm to the latest clang
>>>>>>>>>
>>>>>>>>> Homebrew clang version 20.1.3
>>>>>>>>> Target: arm64-apple-darwin24.3.0
>>>>>>>>> Thread model: posix
>>>>>>>>> InstalledDir: /opt/homebrew/Cellar/llvm/20.1.3/bin
>>>>>>>>> Configuration file:
>>>>>>>>> /opt/homebrew/etc/clang/arm64-apple-darwin24.cfg
>>>>>>>>>
>>>>>>>>> and that seems happier with gfortran. I'm guessing that updating
>>>>>>>>> my mac OS to sequoia mixed stuff up.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Tue, Apr 29, 2025 at 8:24 AM Jeffrey Hyman <hymanjd at gmail.com>
>>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>>> I had tried with clang as well, but was getting this error. Looks
>>>>>>>>>> like something with the arm64. Maybe remove and reinstall gfortran?
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> =============================================================================================
>>>>>>>>>> TESTING: checkFortranNameMangling from
>>>>>>>>>> config.compilers(config/BuildSystem/config/compilers.py:78
>>>>>>>>>>
>>>>>>>>>> *********************************************************************************************
>>>>>>>>>> UNABLE to CONFIGURE with GIVEN OPTIONS (see
>>>>>>>>>> configure.log for details):
>>>>>>>>>>
>>>>>>>>>> ---------------------------------------------------------------------------------------------
>>>>>>>>>> Unknown Fortran name mangling: Are you sure the C and Fortran
>>>>>>>>>> compilers are compatible?
>>>>>>>>>> Perhaps one is 64-bit and one is 32-bit?
>>>>>>>>>> See https://urldefense.us/v3/__https://petsc.org/release/faq/*macos-gfortran__;Iw!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO3A4hOO3$
>>>>>>>>>>
>>>>>>>>>> *********************************************************************************************
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> On Tue, Apr 29, 2025 at 7:56 AM Matthew Knepley <
>>>>>>>>>> knepley at gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> On Mon, Apr 28, 2025 at 10:02 PM Satish Balay <
>>>>>>>>>>> balay.anl at fastmail.org> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> What do you have for:
>>>>>>>>>>>>
>>>>>>>>>>>> clang --version
>>>>>>>>>>>>
>>>>>>>>>>>> Try:
>>>>>>>>>>>>
>>>>>>>>>>>> ./configure COPTFLAGS=-O3 CXXOPTFLAGS=-O3 FOPTFLAGS="-O3
>>>>>>>>>>>> -Wno-unused-function -fallow-argument-mismatch" \
>>>>>>>>>>>> --with-debugging=no --download-mpich=
>>>>>>>>>>>> https://urldefense.us/v3/__https://www.mpich.org/static/downloads/4.3.0/mpich-4.3.0.tar.gz__;!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGOy1WhdS2$
>>>>>>>>>>>> \
>>>>>>>>>>>> --download-hdf5=yes --with-hdf5-fortran-bindings=yes
>>>>>>>>>>>> --download-fblaslapack=yes \
>>>>>>>>>>>> --download-metis=yes --download-parmetis=yes \
>>>>>>>>>>>> CC=gcc-14 CXX=g++-14 FC=gfortran-14 --download-make=yes \
>>>>>>>>>>>> --download-cmake=
>>>>>>>>>>>> https://urldefense.us/v3/__https://gitlab.kitware.com/cmake/cmake/-/archive/v3.31.7/cmake-v3.31.7.tar.gz__;!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO6cDqHoB$
>>>>>>>>>>>> \
>>>>>>>>>>>> --download-cmake-cc=clang --download-cmake-cxx=clang++
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> You could use clang. Your g++ is broken:
>>>>>>>>>>>
>>>>>>>>>>> Error when bootstrapping CMake:
>>>>>>>>>>>
>>>>>>>>>>> Problem while running make
>>>>>>>>>>>
>>>>>>>>>>> ---------------------------------------------
>>>>>>>>>>>
>>>>>>>>>>> Log of errors:
>>>>>>>>>>> /Users/hymanjd/src/petsc/arch-darwin-c-debug/externalpackages/cmake-3.29.0/Bootstrap.cmk/cmake_bootstrap.log
>>>>>>>>>>>
>>>>>>>>>>> ---------------------------------------------
>>>>>>>>>>>
>>>>>>>>>>> stdout:
>>>>>>>>>>>
>>>>>>>>>>> In file included from
>>>>>>>>>>> /opt/homebrew/Cellar/gcc/14.2.0_1/include/c++/14/ext/string_conversions.h:43,
>>>>>>>>>>>
>>>>>>>>>>> from
>>>>>>>>>>> /opt/homebrew/Cellar/gcc/14.2.0_1/include/c++/14/bits/basic_string.h:4154,
>>>>>>>>>>>
>>>>>>>>>>> from
>>>>>>>>>>> /opt/homebrew/Cellar/gcc/14.2.0_1/include/c++/14/string:54,
>>>>>>>>>>>
>>>>>>>>>>> from
>>>>>>>>>>> /Users/hymanjd/src/petsc/arch-darwin-c-debug/externalpackages/cmake-3.29.0/Source/cmTimestamp.h:9,
>>>>>>>>>>>
>>>>>>>>>>> from
>>>>>>>>>>> /Users/hymanjd/src/petsc/arch-darwin-c-debug/externalpackages/cmake-3.29.0/Source/cmTimestamp.cxx:15:
>>>>>>>>>>>
>>>>>>>>>>> /opt/homebrew/Cellar/gcc/14.2.0_1/include/c++/14/cstdlib:141:11:
>>>>>>>>>>> error: 'at_quick_exit' has not been declared in '::'
>>>>>>>>>>>
>>>>>>>>>>> 141 | using ::at_quick_exit;
>>>>>>>>>>>
>>>>>>>>>>> | ^~~~~~~~~~~~~
>>>>>>>>>>>
>>>>>>>>>>> /opt/homebrew/Cellar/gcc/14.2.0_1/include/c++/14/cstdlib:164:11:
>>>>>>>>>>> error: 'quick_exit' has not been declared in '::'
>>>>>>>>>>>
>>>>>>>>>>> 164 | using ::quick_exit;
>>>>>>>>>>>
>>>>>>>>>>> | ^~~~~~~~~~
>>>>>>>>>>>
>>>>>>>>>>> make: *** [cmTimestamp.o] Error 1
>>>>>>>>>>>
>>>>>>>>>>> make: *** Waiting for unfinished jobs....
>>>>>>>>>>>
>>>>>>>>>>> Error running configure on CMAKE: Could
>>>>>>>>>>> not execute "['./configure
>>>>>>>>>>> --prefix=/Users/hymanjd/src/petsc/arch-darwin-c-debug --parallel=8
>>>>>>>>>>> CC="/opt/homebrew/bin/gcc-14" CXX="/opt/homebrew/bin/g++-14" --
>>>>>>>>>>> -DCMAKE_USE_OPENSSL=OFF']":
>>>>>>>>>>>
>>>>>>>>>>> Anything that includes <string> looks like it will break.
>>>>>>>>>>>
>>>>>>>>>>> Thanks,
>>>>>>>>>>>
>>>>>>>>>>> Matt
>>>>>>>>>>>
>>>>>>>>>>> Satish
>>>>>>>>>>>>
>>>>>>>>>>>> On Mon, 28 Apr 2025, Jeffrey Hyman wrote:
>>>>>>>>>>>>
>>>>>>>>>>>> > Thanks, I'll give it a spin and let you know how it goes.
>>>>>>>>>>>> >
>>>>>>>>>>>> > JDH
>>>>>>>>>>>> >
>>>>>>>>>>>> > On Mon, Apr 28, 2025 at 8:43 PM Matthew Knepley <
>>>>>>>>>>>> knepley at gmail.com> wrote:
>>>>>>>>>>>> >
>>>>>>>>>>>> > > On Mon, Apr 28, 2025 at 8:34 PM Jeffrey Hyman <
>>>>>>>>>>>> hymanjd at gmail.com> wrote:
>>>>>>>>>>>> > >
>>>>>>>>>>>> > >> Hi,
>>>>>>>>>>>> > >>
>>>>>>>>>>>> > >> I am trying to build v3.21.5 to use with pflotran. I'm
>>>>>>>>>>>> running into
>>>>>>>>>>>> > >> issues with cmake. configure log attached. Any
>>>>>>>>>>>> suggestions?
>>>>>>>>>>>> > >>
>>>>>>>>>>>> > >
>>>>>>>>>>>> > > The documentation could be clearer here. We cannot use the
>>>>>>>>>>>> PETSc compilers
>>>>>>>>>>>> > > to build CMake since it explicitly does not want MPI. Thus
>>>>>>>>>>>> you need to
>>>>>>>>>>>> > > specify them yourself. That means you would need
>>>>>>>>>>>> > >
>>>>>>>>>>>> > > --download-cmake-cc=/opt/homebrew/bin/gcc-14
>>>>>>>>>>>> > > --download-cmake-cxx=/opt/homebrew/bin/g++-14
>>>>>>>>>>>> > >
>>>>>>>>>>>> > > Thanks,
>>>>>>>>>>>> > >
>>>>>>>>>>>> > > Matt
>>>>>>>>>>>> > >
>>>>>>>>>>>> > >
>>>>>>>>>>>> > >> Thanks for the help,
>>>>>>>>>>>> > >> Jeffrey
>>>>>>>>>>>> > >>
>>>>>>>>>>>> > >>
>>>>>>>>>>>> > >
>>>>>>>>>>>> > > --
>>>>>>>>>>>> > > What most experimenters take for granted before they begin
>>>>>>>>>>>> their
>>>>>>>>>>>> > > experiments is infinitely more interesting than any results
>>>>>>>>>>>> to which their
>>>>>>>>>>>> > > experiments lead.
>>>>>>>>>>>> > > -- Norbert Wiener
>>>>>>>>>>>> > >
>>>>>>>>>>>> > >
>>>>>>>>>>>> https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!bZa6EMKsRCbKBAnduL8Vva4vRPvbjUsz_v4OKJW8-c8D8gQcN4fZPfrAaQfZ37j1kQzigfg-TVDeQNOczHyI$
>>>>>>>>>>>> > > <
>>>>>>>>>>>> https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!bZa6EMKsRCbKBAnduL8Vva4vRPvbjUsz_v4OKJW8-c8D8gQcN4fZPfrAaQfZ37j1kQzigfg-TVDeQKP18rKz$
>>>>>>>>>>>> >
>>>>>>>>>>>> > >
>>>>>>>>>>>> >
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> --
>>>>>>>>>>> What most experimenters take for granted before they begin their
>>>>>>>>>>> experiments is infinitely more interesting than any results to which their
>>>>>>>>>>> experiments lead.
>>>>>>>>>>> -- Norbert Wiener
>>>>>>>>>>>
>>>>>>>>>>> https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO7A0NIjs$
>>>>>>>>>>> <https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO1EGG5Xq$ >
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> What most experimenters take for granted before they begin their
>>>>>>> experiments is infinitely more interesting than any results to which their
>>>>>>> experiments lead.
>>>>>>> -- Norbert Wiener
>>>>>>>
>>>>>>> https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO7A0NIjs$
>>>>>>> <https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO1EGG5Xq$ >
>>>>>>>
>>>>>>
>>>>>
>>>>> --
>>>>> What most experimenters take for granted before they begin their
>>>>> experiments is infinitely more interesting than any results to which their
>>>>> experiments lead.
>>>>> -- Norbert Wiener
>>>>>
>>>>> https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO7A0NIjs$
>>>>> <https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO1EGG5Xq$ >
>>>>>
>>>>
>>>
>>> --
>>> What most experimenters take for granted before they begin their
>>> experiments is infinitely more interesting than any results to which their
>>> experiments lead.
>>> -- Norbert Wiener
>>>
>>> https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO7A0NIjs$
>>> <https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO1EGG5Xq$ >
>>>
>>
>
> --
> What most experimenters take for granted before they begin their
> experiments is infinitely more interesting than any results to which their
> experiments lead.
> -- Norbert Wiener
>
> https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO7A0NIjs$
> <https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!YMeq7HPz3Z9rUv4sDKNYziJ-XIL88Zv9NLkmcUrYIbbynYVVLNCKL0HHfMnAJT8bbyhi6DrS5nQGO1EGG5Xq$ >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20250429/6a6e5173/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: configure.log
Type: application/octet-stream
Size: 6632312 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20250429/6a6e5173/attachment-0001.obj>
More information about the petsc-dev
mailing list