Hi all,
I am trying to compile the Einstein toolkit on the SuperMUC-NG cluster but I am having some issues with the HDF5 libraries and I don’t understand whether it is a problem of the cluster itself or if there is any workaround that maybe some of you know.
I first this warning message when the CST is running on the HDF5 section: ******************************************************************************** Running configuration script for thorn HDF5: Additional requested language support: Fortran WARNING in HDF5 configuration: None of H5pubconf.h H5pubconf-64.h H5pubconf-32.h found in Automatic detection of szip/zlib compression not possible Finished running configuration script for thorn HDF5. Checking consistency... Creating Thorn-Flesh bindings... Creating implementation bindings... Creating parameter bindings... Creating variable bindings... Creating schedule bindings... Creating function bindings... CST finished. ********************************************************************************
Then the compilation goes on until it fails when there’s the first call to the HDF5 library: ******************************************************************************** dss/dsshome1/00/di39wog2/EinsteinToolkit/Cactus/arrangements/EinsteinAnalysis/AHFinderDirect/src/patch/patch_system.cchttp://patch_system.cc(87): catastrophic error: cannot open source file "hdf5.h" # include <hdf5.h> ^
/dss/dsshome1/00/di39wog2/EinsteinToolkit/Cactus/arrangements/EinsteinAnalysis/AHFinderDirect/src/patch/patch_system.cchttp://patch_system.cc(87): catastrophic error: cannot open source file "hdf5.h" # include <hdf5.h> ^
compilation aborted for /dss/dsshome1/00/di39wog2/EinsteinToolkit/Cactus/configs/ET_SUPERMUC_NG/build/AHFinderDirect/patch/patch_system.cchttp://patch_system.cc (code 4) make[3]: *** [patch_system.cc.o] Error 4 make[2]: *** [patch/make.checked] Error 2 make[1]: *** [/dss/dsshome1/00/di39wog2/EinsteinToolkit/Cactus/configs/ET_SUPERMUC_NG/lib/libthorn_AHFinderDirect.a] Error 2 make: *** [ET_SUPERMUC_NG] Error 2
********************************************************************************
I attach here both log and error file of the compilation and my configuration script if you need it. I have already tried every possible hdf5 module they have on the cluster. They all lead to the same exact error. Please let me know if anything else is needed.
Best regards, Giuseppe
Hello Giuseppe,
HDF5_LIBS = /lrz/sys/io_tools/hdf5/1.8.15/serial_icc17/lib HDF5_INC = /lrz/sys/io_tools/hdf5/1.8.15/serial_icc17/include
these variables should be called HDF5_LIB_DIRS and HDF5_INC_DIRS (see eg https://bitbucket.org/simfactory/simfactory2/src/463a900955cd0a62ac4af064b20...) though I think it will also work if you remove those two (but keep HDF5_DIR).
Yours, Roland
Hi Roland,
of course you are right, thanks!! I have tried with just HDF5_DIR before but it complained. Now it works fine.
Cheers, Giuseppe
On 8 Oct 2019, at 18:24, Haas, Roland <rhaas@illinois.edumailto:rhaas@illinois.edu> wrote:
Hello Giuseppe,
HDF5_LIBS = /lrz/sys/io_tools/hdf5/1.8.15/serial_icc17/lib HDF5_INC = /lrz/sys/io_tools/hdf5/1.8.15/serial_icc17/include these variables should be called HDF5_LIB_DIRS and HDF5_INC_DIRS (see eg https://bitbucket.org/simfactory/simfactory2/src/463a900955cd0a62ac4af064b20...) though I think it will also work if you remove those two (but keep HDF5_DIR).
Yours, Roland
-- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://pgp.mit.edu .
users@lists.einsteintoolkit.org
-
Ficarra, Giuseppe -
Haas, Roland