Hi,
Yes, the underlying problem appears to be another python snag.
And community advice appears to be to use an older Python than 3.11.
See web search results on longintrepr.h.
It doesn't pay to be on the bleeding edge of Python.
Advice on bullet proofing Amber wrt Python is welcome.
scott
ps
why did the dhampir cross the road ?
it was stalking a chicken.
what did dracula tell the dhampir?
walk this was; chickens have fowl blood.
On Thu, Dec 07, 2023 at 09:47:58AM -0600, Todd Minehardt via AMBER wrote:
> Irene,
>
> I can tell you that a full compile/install of the latest AmberTools 23 on
> Ubuntu 22.04 LTS without Miniconda - i.e., with gcc 11.4 and CUDA 12.3 -
> works just fine.
>
> The issue you're having is the location of at least one header file that
> apparently has moved upon migrating from 3.10 to 3.11, so you need to copy
> the file from one place to another and/or amend some INCLUDE paths.
>
> My solution to problems like this - with AMBER and in life in general - is
> to find the thing that works and do that instead. This is the UNIX way.
>
> You can:
>
> (1) Amend run_cmake to mimic mine (you'll need to be doing a bunch of apt
> installs iteratively along the way) and dodge the issue involving
> Miniconda/cython/whatever:
>
> ---------------------------------------------------------------------------------------------
> # Assume this is Linux:
>
> cmake $AMBER_PREFIX/amber22_src \
> -DCMAKE_INSTALL_PREFIX=/usr/local/amber22 \
> -DCOMPILER=GNU \
> -DMPI=FALSE -DCUDA=TRUE -DINSTALL_TESTS=TRUE \
> -DDOWNLOAD_MINICONDA=FALSE \
> -DTRUST_SYSTEM_LIBS=TRUE \
> 2>&1 | tee cmake.log
> ---------------------------------------------------------------------------------------------
>
> and then:
>
> ./run_cmake
> sudo make install
> echo "source /usr/local/amber22/amber.sh" >> ~/.bashrc
>
> OR
>
> (2) Hit the Google search results for "longintrepr.h: No such file or
> directory" and go from there. It's rough sledding and I recommend #1.
>
> Cheers,
>
> Todd
>
>
>
> On Thu, Dec 7, 2023 at 9:22???AM Irene Maffucci via AMBER <amber.ambermd.org>
> wrote:
>
> > Dear Amber users,
> >
> > I've been using Amber since Amber12 so I had to compile it multiple
> > times and on multiple working stations without having any issue. I have
> > now a new working station with Ubuntu22 on it and I am trying to compile
> > Amber22, however when compiling the serial version I end up with the
> > following error during the "Building pytraj native library" step:
> >
> > building 'pytraj.datasets.c_datasetlist' extension
> > creating build/temp.linux-x86_64-cpython-311
> > creating build/temp.linux-x86_64-cpython-311/pytraj
> > creating build/temp.linux-x86_64-cpython-311/pytraj/datasets
> > /usr/bin/gcc -DNDEBUG -fwrapv -O2 -Wall -fPIC -O2 -isystem
> > /usr/local/amber22_src/build/CMakeFiles/miniconda/install/include -fPIC
> > -O2 -isystem
> > /usr/local/amber22_src/build/CMakeFiles/miniconda/install/include -fPIC
> > -I/usr/local/amber22_src/AmberTools/src/pytraj/../cpptraj/src
> > -I/usr/local/amber22_src/AmberTools/src/pytraj/pytraj
> > -I/usr/local/amber22_src/build/CMakeFiles/miniconda/install/include/python3.11
> >
> > -c pytraj/datasets/c_datasetlist.cpp -o
> > build/temp.linux-x86_64-cpython-311/pytraj/datasets/c_datasetlist.o -O0
> > -ggdb -Wl,-rpath=/usr/local/amber22_src/build/AmberTools/src/cpptraj/src
> > pytraj/datasets/c_datasetlist.cpp:212:12: fatal error: longintrepr.h: No
> > such file or directory
> > 212 | #include "longintrepr.h"
> > | ^~~~~~~~~~~~~~~
> > compilation terminated.
> > error: command '/usr/bin/gcc' failed with exit code 1
> > make[2]: *** [AmberTools/src/pytraj/CMakeFiles/pytraj.dir/build.make:533
> > : AmberTools/src/pytraj/CMakeFiles/python-build/pytraj-build.stamp]
> > Error 1
> > make[1]: *** [CMakeFiles/Makefile2:7784 :
> > AmberTools/src/pytraj/CMakeFiles/pytraj.dir/all] Error 2
> > make: *** [Makefile:156 : all] Error 2
> >
> > I tried to find a solution by myself, it might come from a problem with
> > python 3.11 installed during the process, but I am no sure about it.
> > Did anyone encounter the same error or might be able to help me to solve
> > it?
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Dec 07 2023 - 13:30:05 PST
