Pull a free-standing `find_python_in_prefix` function out of `python`'s `command()`
property.
Originally done for #44382 but not ultimately used, I still think this is a good
refactor, so submitting as a separate pull request.
Signed-off-by: Todd Gamblin <tgamblin@llnl.gov>
* e4s oneapi: upgrade to latest compilers oneapi@2025.1
* update specs and package preferences
* enable some more dav packages
* enable additional specs
* e4s oneapi: packages: elfutils does not have bzip2 variant
* e4s oneapi: packages: elfutils does not have xz variant
* e4s oneapi: comment out heffte+sycl
* comment out e4s oneapi failures
* comment out more failures
* comment out failing spec
* Revert "paraview: add patch for Intel Classic compilers (#49116)"
This reverts commit 7a95e2beb5.
We'll mark Intel Classic compilers as conflicting with ParaView
versions 5.13.0-5.13.2 instead since 5.13.3 is available and can be
built with with those compilers.
* Add conflict for Intel Class compilers and ParaView 5.13.0-5.13.2.
* paraview: add new v5.13.3 release
This commit reorders ASP setup, so that rules from
possible compilers are collected first.
This allows us to know the dependencies that may be
injected before counting the possible dependencies,
so we can account for them too.
Proceeding this way makes it easier to inject
complex runtimes, like hip.
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
Deal with the "issue" that passing a str instance does not cause a
type check failure, because str is a subset of Sequence[str] and
Iterable[str]. Instead fix it by special casing the str instance.
* fix: move depends_on(c,cxx,fortran) with other dependencies, after variants
* treewide style: move depends_on(c,cxx,fortran) with other dependencies, after variants
* treewide style: move depends_on(c,cxx,fortran) with other dependencies, after variant
---------
Co-authored-by: Alec Scott <hi@alecbcs.com>
Compatibility with Python 3.6 is still tested by the
rhel8-platform-python job, and Ubuntu 20.04 will be
removed soon from the list of runners:
https://github.com/actions/runner-images/issues/11101
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
CMake 4.0.0 breaks compatibility with CMake projects
requiring a CMake < 3.5. However, many projects that
specify a minimum requirement for versions older
than 3.5 are actually compatible with newer CMake
and do not use CMake 3.4 or older features. This
allows those projects to use a newer CMake
Co-authored-by: John W. Parent <45471568+johnwparent@users.noreply.github.com>
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
The requirements being removed are redundant, or
even outdated (%cray-prgenv-* is not a compiler in v0.23).
When compilers turned into nodes, these constraints,
with the "one_of" policy, started being unsat under
certain conditions e.g. we can't compile anymore
with GCC and depend on LLVM as a library.
Remove the requirements to make the recipe solvable
again.
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* update packages to work with python 3.12+
* partd updates
* py-nc-time-axis updates for 3.12
* Add new py-cftime as required for py-nc-time-axis
* fix dropped python 3.9
* switch from when blocks to flat
* remove redundant requires
* protect version range for python@:3.11
* add new c-blosc to support newly added python 3.12 py-blosc version
* add scikit-build 0.18.1 for python 3.12 required for this set of commits
* add complete optional variant for py-partd to match pyproject.toml. Deprecate super old versions
* only set system blosc for the required case
* style
* Remove incorrect python bound
* improve python version reqs and move to more canonical depends_on
* move to depends from req
* add new python range limit, update comment
* remove @coreyjadams as maintainer as per their request https://github.com/spack/spack/pull/48830#issuecomment-2705062587
* Fix python bounds
Co-authored-by: Wouter Deconinck <wdconinc@gmail.com>
---------
Co-authored-by: Wouter Deconinck <wdconinc@gmail.com>
* Check for LSF, FLux, and Slurm when determing MPI exec
* Make scheduler/MPI exec helper functions methods of CachedCMakeBuilder
* Remove axom workaround for running mpi on machines with flux
* Clearly split old and new hip settings requirements
* Apply generic rocm handling to every project
* make default logic for hip support more robust
* GPU_TARGET is only necessary under certain project specific conditions, it should not be necessary in general
* Update logic to find amdclang++
