I made no other changes so it was still with the updated h5xx 0.9.2; that compilation still failed in the same module but with a different error, more below...
However when I reverted to the h5xx shipped with Espresso the compilation succeeded, so thank you!
This with h5xx 0.92 and HDF5 12.2.2, purely for information:
<<
[ 44%] Building CXX object src/core/CMakeFiles/Espresso_core.dir/magnetostatics/dds_replica.cpp.o
[ 44%] Building CXX object src/core/CMakeFiles/Espresso_core.dir/magnetostatics/dlc.cpp.o
In file included from /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/h5xx.hpp:30,
from /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/h5md_specification.hpp:24,
from /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/h5md_core.hpp:27,
from /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/h5md_core.cpp:22:
/home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/group.hpp: In function 'herr_t h5xx::detail::find_name_of_type_impl(hid_t, const char*, const H5L_info2_t*, void*)':
/home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/group.hpp:564:40: error: too few arguments to function 'herr_t H5Oget_info_by_name3(hid_t, const char*, H5O_info2_t*, unsigned int, hid_t)'
564 | herr_t retval = H5Oget_info_by_name(g_id, name, &obj_info, H5P_DEFAULT);
| ^
In file included from /home/apps/easybuild/software/HDF5/1.12.2-gompi-2022b/include/H5Apublic.h:22,
from /home/apps/easybuild/software/HDF5/1.12.2-gompi-2022b/include/hdf5.h:23,
from /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/h5xx.hpp:14:
/home/apps/easybuild/software/HDF5/1.12.2-gompi-2022b/include/H5Opublic.h:521:15: note: declared here
521 | H5_DLL herr_t H5Oget_info_by_name3(hid_t loc_id, const char *name, H5O_info2_t *oinfo, unsigned fields,
| ^~~~~~~~~~~~~~~~~~~~
make[2]: *** [src/core/CMakeFiles/Espresso_core.dir/io/writer/h5md_core.cpp.o] Error 1
make[2]: *** Waiting for unfinished jobs....
In file included from /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/h5xx.hpp:30,
from /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/h5md_specification.hpp:24,
from /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/h5md_specification.cpp:22:
/home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/group.hpp: In function 'herr_t h5xx::detail::find_name_of_type_impl(hid_t, const char*, const H5L_info2_t*, void*)':
/home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/group.hpp:564:40: error: too few arguments to function 'herr_t H5Oget_info_by_name3(hid_t, const char*, H5O_info2_t*, unsigned int, hid_t)'
564 | herr_t retval = H5Oget_info_by_name(g_id, name, &obj_info, H5P_DEFAULT);
| ^
In file included from /home/apps/easybuild/software/HDF5/1.12.2-gompi-2022b/include/H5Apublic.h:22,
from /home/apps/easybuild/software/HDF5/1.12.2-gompi-2022b/include/hdf5.h:23,
from /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/h5xx.hpp:14:
/home/apps/easybuild/software/HDF5/1.12.2-gompi-2022b/include/H5Opublic.h:521:15: note: declared here
521 | H5_DLL herr_t H5Oget_info_by_name3(hid_t loc_id, const char *name, H5O_info2_t *oinfo, unsigned fields,
| ^~~~~~~~~~~~~~~~~~~~
make[2]: *** [src/core/CMakeFiles/Espresso_core.dir/io/writer/h5md_specification.cpp.o] Error 1
make[1]: *** [src/core/CMakeFiles/Espresso_core.dir/all] Error 2
make: *** [all] Error 2
>>
This might relate to a change in HDF5's H50public.h; there are comments in the file (this by OCR so may be some typos):
518 * \since 1.12.0
519 *
520 */
521 H5_DLL herr_t H50get_info_by_name3(hid_t loc_id, const char *name, H50_info2_t *oinfo, unsigned fields,
522 hid_t lapl_id);
523
524 /**
.
.
1773 *
1774 * \version 1.10.5 The macro #H50get_info_by_name was removed and the function
1775 * H50get_info_by_namel() was copied to H50get_info_by_name().
1776 * \version 1.10.3 Function H50get_info_by_name() was copied to H50get_info_by_namel()
1777 * and the macro #H50get_info_by_name was created.
.
.
1870 */
1871 H5_DLL herr_t H50get_info2(hid_t loc_id, H50_infol_t *oinfo, unsigned fields);
1872
.
.
1887 *
1888 * \deprecated As of HDF5-1.12 this function has been deprecated in favor of
1889 * the function H50get_info_by_name3() or the macro H50get_info_by_name().
1890 *
The macros are defined in H5version.h:
768 #if !defined(H50get_info_by_name_vers) || H50get_info_by_name_vers == 3
769 #ifndef H50get_info_by_narr.e_vers
770 #define H50get_info_by_narr,e_vers 3
771 #endif /* H50get_info_by_name_vers */
772 #define H50get_info_by_name H50get_info_by_name3
773 #elif H50get_info_by_name_vers == 2
774 #define H50get_info_by_name H50get_info_by_name2
775 #elif H50get_info_by_name_vers == 1
776 #define H50get_info_by_name H50get_info_by_namel
777 #else /* H50get_info_by_name_vers */
778 #error "H50get_info_by_name_vers set to invalid value"
779 #endif /* H50get_info_by_name_vers */
It's not your issue but worth noting; I see that on the h5xx page: https://github.com/fhoefling/h5xx it reports that 'in 2016' h5xx was tested with HDF5 1.8.xx; and although the cmake files were updated recently the source was not updated in 4 years, so is not tracking changes in HDF5.
Once again many thanks for your detailed and quick response. Now I shall add the Python dependencies to our installation and build the Lmod module file.
William
William Brown
Senior System Administrator (Research IT)
Information Technology
121/122 St Stephen's Green Dublin 2 Ireland
T: +353 1 402 5234
M: +353 894 746447
E: williambrown@rcsi.ie
W: www.rcsi.com
Discover why we are Top 50 in the World for 'International Outlook' - THE World University Ranking 2023
If you receive this email outside of your normal working hours, you are not expected to respond until you return to work.
Registered Charity Number: 20001957
-----Original Message-----
From: Jean-Noël Grad <jgrad@icp.uni-stuttgart.de>
Sent: 05 August 2023 21:18
To: William Brown <williambrown@rcsi.ie>
Cc: espressomd-users@nongnu.org
Subject: Re: Compilation Error: h5xx; template..cannot be overloaded...
[You don't often get email from jgrad@icp.uni-stuttgart.de. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
CAUTION: This email originated from outside of the organisation. Do not click links or open attachments unless you recognise the sender and know that the content is safe.
Dear William Brown,
Thank you for reporting this issue to us, and for providing these detailed information about your build environment!
After some investigation, it seems to me the h5md/h5xx library is currently not compatible with hdf5 >= 1.13.0. We never caught it in the ESPResSo automated tests, because the latest versions of Fedora and Ubuntu are still providing hdf5 <= 1.12.1 at the moment (hdf5 1.14.1 is currently being explored in Fedora scratch builds, though). From the looks of it, h5md/h5xx is missing a few ifdef guards to become compliant with hdf5 >= 1.13.0. I have opened a ticket on their main repo to explain the situation [10]. Please find below my observations.
The hdf5 1.14 specifications defines "hsize_t" as "uint64_t" in H5public.h [4]. On a 64-bit platform, "uint64_t" is typically equivalent to "long unsigned int", which just so happens to also be equivalent to "std::size_t". Hence the constructor overloads in the h5xx code [5] are declared twice with the same argument types, which is not allowed.
This was not an issue with previous hdf5 releases, because the definition of "hsize_t" was "unsigned long long" in 1.10 [6] and 1.12 [7]. That was safe because "unsigned long long" and "long unsigned int"
are both 8 bytes, and the constructor overload resolution considered them as different types, so there was no re-definition. The change to "uint64_t" happened in 1.13 [8], more specifically in HDFGroup/hdf5#709.
The h5md/h5xx project could probably resolve this issue by adding ifdef guards around all std::size_t overloads to disable them when the hdf5 version is >= 1.13.0, although I'm not 100% sure how to properly handle portability issues architectures other than 64-bit.
Do you have by any chance the option of loading an older version of hdf5 on your cluster?
Best regards,
Jean-Noël Grad
Links:
[4]
https://github.com/HDFGroup/hdf5/blob/hdf5-1_14_1/src/H5public.h#L291
[5]
https://github.com/fhoefling/h5xx/blob/459f07af/h5xx/dataspace/dataspace.hpp#L49C77-L53C53
[6]
https://github.com/HDFGroup/hdf5/blob/hdf5-1_10_0/src/H5public.h#L184
[7]
https://github.com/HDFGroup/hdf5/blob/hdf5-1_12_0/src/H5public.h#L186
[8]
https://github.com/HDFGroup/hdf5/blob/hdf5-1_13_0/src/H5public.h#L291
[9] https://github.com/HDFGroup/hdf5/pull/709
[10] https://github.com/fhoefling/h5xx/issues/11
On 2023-08-04 18:29, William Brown wrote:
> I am trying to compile 4.2.1 using GCC 12.2.0. cmake seems to go OK;
>
> . the notice about doxygen missing components: dot I read is related
> to graphviz, and is only needed to produce the documents.
>
> $ CMAKE ..
>
> -- CMAKE VERSION: 3.24.3
>
> -- THE CXX COMPILER IDENTIFICATION IS GNU 12.2.0
>
> -- DETECTING CXX COMPILER ABI INFO
>
> -- DETECTING CXX COMPILER ABI INFO - DONE
>
> -- CHECK FOR WORKING CXX COMPILER:
> /HOME/APPS/EASYBUILD/SOFTWARE/GCCCORE/12.2.0/BIN/C++ - SKIPPED
>
> -- DETECTING CXX COMPILE FEATURES
>
> -- DETECTING CXX COMPILE FEATURES - DONE
>
> -- THE C COMPILER IDENTIFICATION IS GNU 12.2.0
>
> -- DETECTING C COMPILER ABI INFO
>
> -- DETECTING C COMPILER ABI INFO - DONE
>
> -- CHECK FOR WORKING C COMPILER:
> /HOME/APPS/EASYBUILD/SOFTWARE/GCCCORE/12.2.0/BIN/CC - SKIPPED
>
> -- DETECTING C COMPILE FEATURES
>
> -- DETECTING C COMPILE FEATURES - DONE
>
> -- CONFIG FILE:
> /HOME/APPS/CUSTOM/SOFTWARE/ESPRESSO/4.2.1/ESPRESSO/SRC/CONFIG/MYCONFIG
> -DEFAULT.HPP
>
> -- PERFORMING TEST RESULT__PRETTY_FUNCTION__
>
> -- PERFORMING TEST RESULT__PRETTY_FUNCTION__ - SUCCESS
>
> -- FOUND PYTHONINTERP:
> /HOME/APPS/EASYBUILD/SOFTWARE/PYTHON/3.10.8-GCCCORE-12.2.0/BIN/PYTHON3
> (FOUND SUITABLE VERSION "3.10.8", MINIMUM REQUIRED IS "3.8")
>
> -- FOUND CYTHON:
> /HOME/APPS/EASYBUILD/SOFTWARE/PYTHON/3.10.8-GCCCORE-12.2.0/BIN/PYTHON3
> ;-M;CYTHON (FOUND SUITABLE VERSION "0.29.36", MINIMUM REQUIRED IS
> "0.29.14")
>
> -- FOUND PYTHONHEADERS:
> /HOME/APPS/EASYBUILD/SOFTWARE/PYTHON/3.10.8-GCCCORE-12.2.0/INCLUDE/PYT
> HON3.10
>
> -- FOUND NUMPY:
> /HOME/APPS/EASYBUILD/SOFTWARE/SCIPY-BUNDLE/2023.02-GFBF-2022B/LIB/PYTH
> ON3.10/SITE-PACKAGES/NUMPY/CORE/INCLUDE
> (FOUND VERSION "1.24.2")
>
> -- FOUND FFTW3:
> /HOME/APPS/EASYBUILD/SOFTWARE/FFTW.MPI/3.3.10-GOMPI-2022B/LIB/LIBFFTW3
> .SO
>
> -- FOUND HDF5:
> /HOME/APPS/EASYBUILD/SOFTWARE/HDF5/1.14.0-GOMPI-2022B/LIB/LIBHDF5.SO;/
> USR/LIB64/LIBPTHREAD.SO;/HOME/APPS/EASYBUILD/SOFTWARE/SZIP/2.1.1-GCCCO
> RE-12.2.0/LIB/LIBSZ.SO;/HOME/APPS/EASYBUILD/SOFTWARE/ZLIB/1.2.12-GCCCO
> RE-12.2.0/LIB/LIBZ.SO;/USR/LIB64/LIBDL.SO;/USR/LIB64/LIBM.SO;/USR/LIB6
> 4/LIBM.SO;/USR/LIB64/LIBPTHREAD.SO
> (FOUND SUITABLE VERSION "1.14.0", MINIMUM REQUIRED IS "1.8") FOUND
> COMPONENTS: C
>
> -- FOUND GSL: /HOME/APPS/EASYBUILD/SOFTWARE/GSL/2.7-GCC-12.2.0/INCLUDE
> (FOUND VERSION "2.7")
>
> -- FOUND MPI_C:
> /HOME/APPS/EASYBUILD/SOFTWARE/OPENMPI/4.1.4-GCC-12.2.0/LIB/LIBMPI.SO
> (FOUND SUITABLE VERSION "3.1", MINIMUM REQUIRED IS "3.0")
>
> -- FOUND MPI_CXX:
> /HOME/APPS/EASYBUILD/SOFTWARE/OPENMPI/4.1.4-GCC-12.2.0/LIB/LIBMPI.SO
> (FOUND SUITABLE VERSION "3.1", MINIMUM REQUIRED IS "3.0")
>
> -- FOUND MPI: TRUE (FOUND SUITABLE VERSION "3.1", MINIMUM REQUIRED IS
> "3.0")
>
> -- FOUND MPIEXECBACKEND:
> /HOME/APPS/EASYBUILD/SOFTWARE/OPENMPI/4.1.4-GCC-12.2.0/BIN/MPIEXEC
>
> -- FOUND BOOST:
> /HOME/APPS/EASYBUILD/SOFTWARE/BOOST.MPI/1.81.0-GOMPI-2022B/LIB64/CMAKE
> /BOOST-1.81.0/BOOSTCONFIG.CMAKE (FOUND SUITABLE VERSION "1.81.0",
> MINIMUM REQUIRED IS "1.69.0") FOUND
> COMPONENTS: MPI SERIALIZATION FILESYSTEM SYSTEM UNIT_TEST_FRAMEWORK
>
> -- FOUND DOXYGEN: /USR/BIN/DOXYGEN (FOUND VERSION "1.8.5") FOUND
> COMPONENTS: DOXYGEN MISSING COMPONENTS: DOT
>
> -- FOUND SPHINX:
> /HOME/APPS/EASYBUILD/SOFTWARE/PYTHON/3.10.8-GCCCORE-12.2.0/BIN/PYTHON3
> ;-M;SPHINX (FOUND SUITABLE VERSION "5.3.0", MINIMUM REQUIRED IS
> "2.3.0")
>
> -- FOUND GIT: /USR/BIN/GIT (FOUND VERSION "1.8.3.1")
>
> -- THE FOLLOWING FEATURES HAVE BEEN ENABLED:
>
> * HDF5 (REQUIRED VERSION >= 1.8), PARALLEL
>
> -- THE FOLLOWING OPTIONAL PACKAGES HAVE BEEN FOUND:
>
> * FFTW3
>
> * MPIEXECBACKEND
>
> * DOXYGEN
>
> * SPHINX (REQUIRED VERSION >= 2.3.0)
>
> * GIT
>
> -- THE FOLLOWING REQUIRED PACKAGES HAVE BEEN FOUND:
>
> * PYTHONINTERP (REQUIRED VERSION >= 3.8)
>
> * CYTHON (REQUIRED VERSION >= 0.29.14)
>
> * PYTHONHEADERS
>
> * NUMPY
>
> * HDF5 (REQUIRED VERSION >= 1.8), PARALLEL
>
> * GSL
>
> * MPI
>
> * BOOST_MPI (REQUIRED VERSION == 1.81.0)
>
> * BOOST_SERIALIZATION (REQUIRED VERSION == 1.81.0)
>
> * BOOST_ATOMIC (REQUIRED VERSION == 1.81.0)
>
> * BOOST_FILESYSTEM (REQUIRED VERSION == 1.81.0)
>
> * BOOST_SYSTEM (REQUIRED VERSION == 1.81.0)
>
> * BOOST_HEADERS (REQUIRED VERSION == 1.81.0)
>
> * BOOST_UNIT_TEST_FRAMEWORK (REQUIRED VERSION == 1.81.0)
>
> * BOOST (REQUIRED VERSION >= 1.69.0)
>
> -- CONFIGURING DONE
>
> -- GENERATING DONE
>
> -- BUILD FILES HAVE BEEN WRITTEN TO:
> /HOME/APPS/CUSTOM/SOFTWARE/ESPRESSO/4.2.1/ESPRESSO/BUILD
>
> However the compilation fails:
>
> .
>
> .
>
> [ 44%] Building CXX object
> src/core/CMakeFiles/Espresso_core.dir/magnetostatics/dds_gpu.cpp.o
>
> [ 44%] Building CXX object
> src/core/CMakeFiles/Espresso_core.dir/magnetostatics/dds_replica.cpp.o
>
>
> [ 44%] Building CXX object
> src/core/CMakeFiles/Espresso_core.dir/magnetostatics/dlc.cpp.o
>
> In file included from
> /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/data
> space.hpp:16,
>
>
> from
> /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/attr
> ibute/attribute.hpp:17,
>
>
> from
> /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/attr
> ibute.hpp:17,
>
>
> from
> /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/h5xx
> .hpp:23,
>
>
> from
> /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/
> h5md_specification.hpp:24,
>
>
> from
> /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/
> h5md_core.hpp:27,
>
>
> from
> /home/apps/custom/software/espresso/4.2.1/espresso/src/core/io/writer/h5md_core.cpp:22:
>
>
> /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/dataspace/dataspace.hpp:52:5:
> error: 'template<long unsigned int N> h5xx::dataspace::dataspace(const
> boost::array<long unsigned int, NumDims>&)' cannot be overloaded with
> 'template<long unsigned int N> h5xx::dataspace::dataspace(const
> boost::array<long unsigned int, NumDims>&)'
>
> 52 | dataspace(boost::array<hsize_t, N> const& dims);
>
> | ^~~~~~~~~
>
> /home/apps/custom/software/espresso/4.2.1/espresso/libs/h5xx/h5xx/dataspace/dataspace.hpp:50:5:
> note: previous declaration 'template<long unsigned int N>
> h5xx::dataspace::dataspace(const boost::array<long unsigned int,
> NumDims>&)'
>
> 50 | dataspace(boost::array<std::size_t, N> const& dims);
>
> | ^~~~~~~~~
>
> .
>
> .
>
> These are the lines from
> espresso/libs/h5xx/h5xx/dataspace/dataspace.hpp relating to the first
> two error messages:
>
> /** construct simple dataspace of given extents, boost::array
> version */
>
> template <std::size_t N>
>
> dataspace(boost::array<std::size_t, N> const& dims);
>
> template <std::size_t N>
>
> dataspace(boost::array<hsize_t, N> const& dims);
>
> There are more errors in a similar vein about overloading that follow.
>
>
> I see that h5xx ships with Espresso, so I am wondering if there is a
> compatibility issue with the versions of HDF5 or Boost, or GCC 12. I
> did not have this problem building Espresso 4.1.4 with Boost/1.74 and
> HDF5/1.10.7 and GCC 10. I see that h5xx is identical in 4.1.4 and
> 4.2.1 and dates to 2018 and is # 89d8d43 [1] - and I see that h5xx has
> more recently had a version 0.9.2 which specifically mentions C++ 11
> changes: https://github.com/fhoefling/h5xx/issues. I have tried
> building with h5xx 0.9.2 but I get the same errors.
>
> I am using Easybuild and Lmod and these are the loaded versions of the
> required software:
>
> Currently Loaded Modules:
>
> 1) GCCcore/12.2.0 25)
> FFTW.MPI/3.3.10-gompi-2022b (H)
>
> 2) zlib/1.2.12-GCCcore-12.2.0 (H) 26)
> OpenBLAS/0.3.21-GCC-12.2.0 (H)
>
> 3) binutils/2.39-GCCcore-12.2.0 (H) 27)
> FlexiBLAS/3.2.1-GCC-12.2.0 (H)
>
> 4) GCC/12.2.0 28)
> ScaLAPACK/2.2.0-gompi-2022b-fb (H)
>
> 5) numactl/2.0.16-GCCcore-12.2.0 (H) 29) foss/2022b
>
> 6) XZ/5.2.7-GCCcore-12.2.0 (H) 30)
> ncurses/6.3-GCCcore-12.2.0 (H)
>
> 7) libxml2/2.10.3-GCCcore-12.2.0 (H) 31)
> libreadline/8.2-GCCcore-12.2.0 (H)
>
> 8) libpciaccess/0.17-GCCcore-12.2.0 (H) 32)
> Tcl/8.6.12-GCCcore-12.2.0 (H)
>
> 9) hwloc/2.8.0-GCCcore-12.2.0 (H) 33)
> SQLite/3.39.4-GCCcore-12.2.0 (H)
>
> 10) OpenSSL/1.1 (H) 34) GMP/6.2.1-GCCcore-12.2.0
>
>
> 11) libevent/2.1.12-GCCcore-12.2.0 (H) 35)
> libffi/3.4.4-GCCcore-12.2.0 (H)
>
> 12) UCX/1.13.1-GCCcore-12.2.0 (H) 36)
> Python/3.10.8-GCCcore-12.2.0
>
> 13) libfabric/1.16.1-GCCcore-12.2.0 (H) 37) gfbf/2022b
> (H)
>
> 14) PMIx/4.2.2-GCCcore-12.2.0 (H) 38)
> pybind11/2.10.3-GCCcore-12.2.0 (H)
>
> 15) UCC/1.1.0-GCCcore-12.2.0 (H) 39)
> SciPy-bundle/2023.02-gfbf-2022b (H)
>
> 16) OpenMPI/4.1.4-GCC-12.2.0 (H) 40)
> Biopython/1.81-foss-2022b
>
> 17) gompi/2022b (H) 41)
> Cython/0.29.36-GCCcore-12.2.0
>
> 18) bzip2/1.0.8-GCCcore-12.2.0 (H) 42) GSL/2.7-GCC-12.2.0
>
> 19) gzip/1.12-GCCcore-12.2.0 43)
> Szip/2.1.1-GCCcore-12.2.0
>
> 20) lz4/1.9.4-GCCcore-12.2.0 44) HDF5/1.14.0-gompi-2022b
>
> 21) zstd/1.5.2-GCCcore-12.2.0 45)
> cURL/7.86.0-GCCcore-12.2.0
>
> 22) ICU/72.1-GCCcore-12.2.0 46)
> libarchive/3.6.1-GCCcore-12.2.0 (H)
>
> 23) Boost.MPI/1.81.0-gompi-2022b 47)
> CMake/3.24.3-GCCcore-12.2.0 (H)
>
> 24) FFTW/3.3.10-GCC-12.2.0 (H)
>
> Where:
>
> H: Hidden Module
>
> I have looked around but I have not find any report of a similar
> issue. I need to use GCC 12 as the user needs to use this with
> Biopython/1.81, which has been built to use GCC 12 under Easybuild.
>
> I would be grateful for any suggestions as to what to try next.
>
> William
>
> William Brown
> Senior System Administrator (Research IT)
>
> Information Technology
> 121/122 St Stephen's Green Dublin 2 Ireland
> T: +353 1 402 5234 M: +353 894 746447
> E: williambrown@rcsi.ie W: http://www.rcsi.com/ [2]
>
> Discover why we are Top 50 in the World for 'International Outlook' -
> THE World University Ranking 2023 [3]
>
> If you receive this email outside of your normal working hours, you
> are not expected to respond until you return to work.
>
> Registered Charity Number: 20001957
>
> Links:
> ------
> [1]
> https://gith/
> ub.com%2Fh5md%2Fh5xx%2Fcommit%2F89d8d435872ec77dd931e709df98eefac8cbc9
> 25&data=""> > 103d4%7C607041e7a8124670bd3030f9db210f06%7C0%7C0%7C638268634647886999%
> 7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik
> 1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=BwI3nWGNHF4xxTwzeOwsGebtam2
> RmBI6QURrM37%2BF28%3D&reserved=0
> [2] http://www.rcsi.com/
> [3] https://www.rcsi.com/about/international