Hi
I think AmberTools supports python 3.11 too.
The fix is in update.3:
http://ambermd.org/ATPatches.php
There are also conda distribution for ambertools 23 with py311 here:
https://anaconda.org/conda-forge/ambertools/files
Hai
On Wed, Sep 27, 2023 at 10:33 AM Doležal Rafael via AMBER <amber.ambermd.org>
wrote:
> Dear All,
>
> I have also encountered errors related with Miniconda/Python3.11 when
> compiling Amber22/AmberTools23 source code on Centos 7.9.2009. The problem
> was solved after I disabled the automatic Miniconda download in the
> run_cmake script:
>
> # Assume this is Linux:
>
> cmake $AMBER_PREFIX/amber22_src \
> -DCMAKE_INSTALL_PREFIX=$AMBER_PREFIX/amber22 \
> -DCOMPILER=GNU \
> -DMPI=FALSE -DCUDA=TRUE -DINSTALL_TESTS=TRUE \
> -DDOWNLOAD_MINICONDA=FALSE \
> 2>&1 | tee cmake.log
>
> Then I used CUDA-12.0.0, GCC-12.2.0, Python-3.10.4, SciPy-2022.05,
> CMake-3.26.3, Matplotlib-3.5.2. and other related Python tools.
> Finally, I was successful. Pmemd.cuda works well with NVIDIA
> A100-SXM4-40GB.
>
> Have a nice day,
> Rafael Dolezal
>
> This message is a reaction to the e-mail bellow.
>
>
>
>
> From: Charo del Genio via AMBER <amber.ambermd.org<mailto:
> amber.ambermd.org
> ?Subject=Re%3A%20%5BAMBER%5D%20AmberTools23%20fails%20to%20build%20with%20python%203.11>>
> Date: Wed, 3 May 2023 21:59:21 +0100
>
> Dear all,
> I'm having serious problems building Amber with python 3.11.
> Specifically, I get a large number of errors when compiling pytraj, many of
> which are of the kind
>
> pytraj/datasets/c_datasetlist.cpp:26453:1: error: too many initializers
> for ‘PyTypeObject’ {aka ‘_typeobject’}
>
> Upon inspecting the file, it looks like the slots that are assumed for
> PyTypeObject structures are wrong. In fact, the file has
>
> #if PY_VERSION_HEX >= 0x030B00A2
> 0, /*tp_inline_values_offset*/
> #endif
>
> but there is no tp_inline_values_offset slot for python 3.11, as can be
> checked in the official documentation at
> https://docs.python.org/3.11/c-api/typeobj.html
>
> Incidentally, it seems that there may be similar problems for other,
> older, versions of python as well. The only ones I'm confident are OK are
> 3.10.11 and 3.9.16.
>
>
> Additionally, the same file causes the errors
>
> pytraj/datasets/c_datasetlist.cpp:30516:26: error: ‘_PyErr_StackItem’ {aka
> ‘struct _err_stackitem’} has no member named ‘exc_type’
> pytraj/datasets/c_datasetlist.cpp:30518:24: error: ‘_PyErr_StackItem’ {aka
> ‘struct _err_stackitem’} has no member named ‘exc_traceback’
>
> again, in numerous locations. I'm not sure, however, what these mean,
> since I'm by no means an expert of python.
>
>
> So, I suggest that a warning may be put up on the website to indicate that
> pythonn 3.11 is not (yet) supported.
>
>
> Cheers,
>
> Charo
>
>
>
>
>
> --
> Dr. Charo I. del Genio
> Senior Lecturer in Statistical Physics
> Applied Mathematics Research Centre (AMRC)
> Design Hub
> Coventry University Technology Park
> Coventry CV1 5FB
> UK
> https://charodelgenio.weebly.com<https://charodelgenio.weebly.com/>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Sep 27 2023 - 21:00:02 PDT
