* add new flag when compiling mumps with %gcc@10.
* Fix style
* Try to fix formatting
* Use flag_handler approach suggested by @michaelkuhn
in the PR review.
* Delete former approach
* Another style issue
* Add another space
* More fixes
We still need mesa18 for some of our builds.
Those builds require python@2, normal mesa only works with
python@3.
* Remove the deprecation tag
* Add myself as a maintainer: I volunteer to help with this
package for the time being.
* There is only one version, no need to prefer it.
Modifications:
- Make use of SpackCommand objects wherever possible
- Deduplicated code when possible
- Moved cleaning of mirrors to fixtures
- Ensure mock configuration has a clear initialization order
* fixed install with ver 3 and python 3.0
* replaced @3 with @2.999
* [py-pyspark] added version requirements for py-py4j
* [py-pyspark] all versions require at least version 2.7 of python
* [py-pyspark] fixed comma syntax
Co-authored-by: Sid Pendelberry <sid@rit.edu>
The GROMACS package embeds references to its build tool chain.
Use the Spack utilities to make sure these references are correct
outside of the isolated Spack build environment.
`query()` calls `datetime.datetime.fromtimestamp` regardless of whether a
date query is being done. Guard this with an if statement to avoid the
unnecessary work.
Constructing a spec from a name instead of setting name directly forces
from_node_dict to call Spec.parse(), which is slow. Avoid this by using a
zero-arg constructor and setting name directly.
cmake was added as a runtime dependency to meson in #20449. This
introduces an unnecessary implicit cmake dependency, which increases
build time for meson considerably. cmake is only one of many methods for
finding dependencies (pkg-config, qmake etc.), which are also not
runtime dependencies of meson. Add cmake as a build dependency to mesa
instead.
This solves a few FIXMEs in conftest.py, where
we were manipulating globals and seeing side
effects prior to registering fixtures.
This commit solves the FIXMEs, but introduces
a performance regression on tests that may need
to be investigated
The method is now called "use_repositories" and
makes it clear in the docstring that it accepts
as arguments either Repo objects or paths.
Since there was some duplication between this
contextmanager and "use_repo" in the testing framework,
remove the latter and use spack.repo.use_repositories
across the entire code base.
Make a few adjustment to MockPackageMultiRepo, since it was
stating in the docstring that it was supposed to mock
spack.repo.Repo and was instead mocking spack.repo.RepoPath.
Some compilers, such as the NV compilers, do not recognize -isystem
dir when specified without a space.
Works: -isystem ../include
Does not work: -isystem../include
This PR updates the compiler wrapper to include the space with -isystem.
Environment views fail when the tmpdir used for view generation is
on a separate mount from the install_tree because the files cannot
by symlinked between the two. The fix is to use an alternative
tmpdir located alongside the view.
* [py-moviepy] created template
* [py-moviepy] added dependencies
* [py-moviepy] removed fixmes, added homepage and description
* [py-moviepy] updated to pypi and updated checksum
* [py-moviepy] added setuptools dependency
* [py-moviepy] more specific version limit
* [py-moviepy] added checksum for version 1.0.1
* [py-moviepy] numpy restriction not nesessary here
* 3DTK: add new package
* Add missing opencv variants
Co-authored-by: Michael Kuhn <michael@ikkoku.de>
* Fix cmake version req, add eigen dep
* Prefer trunk version
* Tell 3dtk where to find eigen
* Fix installation
* Fix installation
Co-authored-by: Michael Kuhn <michael@ikkoku.de>
This PR fixes the case where groff fails to build if the spack install
path is really long. There are a couple of perl scripts that get built,
and used, during the build phase that will fail when the perl
interpreter line is too long. Filtering the lines will not work because
the files don not exist after the configure phase and patching after the
build phase is too late. This PR runs the scripts explicitly with the
spack perl via the $(PERL) variable in the call to the script.
