Age | Commit message (Collapse) | Author | Files | Lines |
|
* py-pytest-html: add 3.2.0
* py-pytest-html: Add py-py version requirement
See https://github.com/pytest-dev/pytest-html/blob/v3.2.0/setup.py#L16
* py-pytest-html: Add dependencies from setup.py and pyproject.toml
* py-pytest-html: Add git url
* py-pytest-html: Add conflict with py-pytest@7.2: pending py-pytest-html@4
|
|
|
|
|
|
|
|
* Adding optional hip test
* Modifications to run every samples test
* Skipping test directories without a Makefile
* fix styling and cleaning code
* fix styling and changed method of itterating through sample folders
* changed to new syntax for standalone tests
* Updates for changes in syntax
|
|
|
|
* molgw package.py
* molgw package.py almost ready
* bug fix in molgw package.py
* MOLGW v3.2
* black pass
* duplicated line eliminated
* remove FIXME in the header
* add me as a maintainer
* fix previous commit
* sort the imports in the order spack wants
* chop the too-long lines
* many fixes
- variants before dependences
- eliminate useless build and install stages
- no openmp with intel-mkl was broken
* after blackization
* cleaning
* Update var/spack/repos/builtin/packages/molgw/package.py
Co-authored-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* Update var/spack/repos/builtin/packages/molgw/package.py
Co-authored-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* Update var/spack/repos/builtin/packages/molgw/package.py
Co-authored-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* Update var/spack/repos/builtin/packages/molgw/package.py
Co-authored-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* Update var/spack/repos/builtin/packages/molgw/package.py
Co-authored-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* use threaded mkl when openmp is triggered
---------
Co-authored-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
|
|
I noticed this as we run a `pip check` on all modules we generate. This
dependeny seems to have been around for a long time:
https://github.com/dask/distributed/blob/2020.12.0/requirements.txt#L8
https://github.com/dask/distributed/blob/2023.4.1/pyproject.toml#L39
While it does not seem essential, it would be nice to have consistent
installations by including the missing dependency.
|
|
|
|
|
|
|
|
|
|
Currently lm-sensors defaults to the compiler specified in the Makefile
(gcc) rather than the one specified in the spec. This patch appends the
CC flag to the make invocation with the spec compiler to fix this
behavior.
|
|
|
|
* initial commit for rocm-5.5.0 release
* fix the hipsparse build error for 5.5.0
* fix build error for amrex .add hiprand as a dependency
* modify the patch for rocprofiler-dev
* add hiprand for +rocm build
* initial commit for rocm-5.5.1 release
* bump up the version for rocm-5.5.1 release.
* bump up the version for rocmlir.miopen to use this backend only till 5.5
* add new recipe py-barectf and add it as dependency for rocprofiler-dev
* revert the changes for rocprofiler-dev for 5.5.0/1 for now as it depends
on hsa-amdaqlprofile.so which is a closed source and no spack recipe is
available for now.
* add rocm-core as dependency for rocm packages from 5.5.0 onwards
* avoid download of the gtest for building unit tests
|
|
After #36679, the default build type is Release, so there is no need to
explicitly set the build type as a custom variant.
|
|
|
|
|
|
|
|
Co-authored-by: koparasy <parasyris1@llnl.com>
Co-authored-by: Loic Pottier <pottier1@llnl.gov>, Tom Stitt <stitt4@llnl.gov>
|
|
* mruby: added mruby package
* mruby: replaced / with os.path.join
* mruby: improving style
* mruby: improving style again
|
|
|
|
|
|
* feature (packages): implement `odgi` package
This commit re-implements odgi package (superseded by #38741)
* fix (packages): remove redundant `requires()` from odgi package
This commit removes redundant use of `requires()` for gcc version in the `odgi` package
|
|
Instead of inferring whether to sign binaries, make it explicit,
and fail rebuild jobs early if signing is required but cannot be
accomplished.
|
|
libelf fails to build with clang16+ due to Wimplicit-int and
Wimplicit-function-declarations becoming errors by default. This breaks
the configuration stage, so no build takes place. This patch fixes this
by passing -Wno-error=implicit-int and
-Wno-error=implicit-function-declarations as cflags.
|
|
The message is now outdated after
https://github.com/spack/spack/pull/36679 which could lead to some
confusion.
|
|
gcoff uses the register keyword in a couple different places which
causes errors when building with c++17, which is the default in clang
16. This patch adds the -Wno-register flag to ignore these errors when
when building with clang 16.
|
|
Clang 16's change to erroring out by default on implicit function
declarations and implicit integers causes the build script for unzip to
break. Since this project hasn't had a release since 2010, we need to
patch it downstream/pass additional flags to get the build to succeed.
|
|
With the release of clang 16, clang now treats implicit function
declarations and implicit integers as errors rather than warnings,
causing the build to fail. This patch adds flags to prevent build
failures.
|
|
|
|
|
|
|
|
|
|
|
|
* Add Kokkos 4.1.0 release
* Add -fp-model=precise flag when using SYCL
* Fix sycl flag
Co-authored-by: Robert Cohn <rscohn2@gmail.com>
---------
Co-authored-by: Robert Cohn <rscohn2@gmail.com>
Co-authored-by: Tamara Dahlgren <35777542+tldahlgren@users.noreply.github.com>
|
|
(#39000)
|
|
* netcdf-fortran: add version 4.6.1
* netcdf-fortran: fix the 'check' method
|
|
* bfs: add new package
* Remove redundant set to PREFIX during build
|
|
* quickjs: make Makefile use correct compiler
* added lto variant and better handling of compilers
* quickjs: correct style
* quickjs: correct style for flake8
|
|
|
|
Co-authored-by: Thomas Briffard <thomas.briffard@michelin.com>
|
|
|
|
|
|
|
|
|
|
|
|
re-distributing (#39042)
Previously, spack would list the ./spack/binary_distribution file twice when pushing
a package that was installed from a binary tarball itself.
|
|
The sanitization function is completely bogus as it tries to replace /
on unix after ... splitting on it. The way it's implemented is very
questionable: the input is a file name, not a path. It doesn't make
sense to interpret the input as a path and then make the components
valid -- you'll interpret / in a filename as a dir separator.
It also fails to deal with path components that contain just unsupported
characters (resulting in empty component).
The correct way to deal with this is to have a function that takes a
potential file name and replaces unsupported characters.
I'm not going to fix the other issues on Windows, such as reserved file
names, but left a note, and hope that @johnwparent can fix that
separately.
(Obviously we wouldn't have this problem at all if we just fixed the
filename in a safe way instead of trying to derive something from
the url; we could use the content digest when available for example)
|
|
(#39058)
|