Age | Commit message (Collapse) | Author | Files | Lines |
|
`get_platform()` is pretty expensive and can be called many times in a
spack invocation.
- [x] memoize `get_platform()`
|
|
|
|
Added checksum for latest NetCDF release 4.7.3
|
|
* Add dependencies for hpcrypt
* address review comments
* flake
* license-fix
* fix checksums
* Update py-hvace homepage
Co-Authored-By: Adam J. Stewart <ajstewart426@gmail.com>
* update py-hvac url
Co-Authored-By: Adam J. Stewart <ajstewart426@gmail.com>
|
|
|
|
pkgconfig is the virtual provider, pkg-config is an implementation.
|
|
Added checksum for recent 1.12.1 release.
|
|
Added checksum and tag name for last two SEACAS releases.
|
|
doesn't understand a custom, user-defined compiler version. However, if
the compiler's version check fails, you can't build anything with the
custom compiler.
- [x] Be more lenient: fall back to the custom compiler version and use
it verbatim if the version check fails.
|
|
`pgcc -V` was failing on power machines because it returns 2 (despite
correctly printing version information). On x86_64 machines the same
command returns 0 and doesn't cause an error.
- [x] Ignore return value of 2 for pgcc when doign a version check
|
|
|
|
|
|
* package py-zc-lockfile
* package py-zc-lockfile
* autopep8, docutils cleanup
* Update var/spack/repos/builtin/packages/py-zc-lockfile/package.py
Co-Authored-By: Adam J. Stewart <ajstewart426@gmail.com>
* Update var/spack/repos/builtin/packages/py-zc-lockfile/package.py
Co-Authored-By: Adam J. Stewart <ajstewart426@gmail.com>
* dependency at wrong level
* license bits
* Remove py-zc-buildout dependency
|
|
* add new package : filebench
* remove alpha version and duplicated autoheader cmd
* refine automake cmd in sh()
* refine filebench url as a stable tarball link
|
|
* add new package : hama
* mv hama/hama/package.py to hama/package.py
|
|
* add new package : hbase
* add hbase description
* refine hbase version and depends
|
|
(#14203)
* root: Rationalize and improve version, variant and ROOT option handling.
* Completely re-vamp CMake option handling for readability and maintainability:
* Three categories of option: control, builtin and feature, alphabetically sorted.
* Each option is described as a list: an option name followed by an optional value which is either Boolean or a string representing the name of a variant. If the value is omitted, it defaults to the option name.
* New functions `_process_opts()` and `_process_opt()` (nested) to turn all supplied option/value specifications into CMake arguments.
* Remove overly-terse per-option comments in favor of (much) more comprehensive notes in README.md.
* Variants and conflicts:
* Remove `test` variant in favor of pegging ROOT `testing` option to the value of `self.run_tests` since the install is unaffected, per ROOT developer.
* Remove commented-out and never-functional variants: `asimage`, `avahi`, `kerberos`, `ldap`, `libcxx`, `odbc`, `oracle`, `pythia8`, `xinetd`.
* New variant `vmc` (default `OFF`) to control the Virtual Monte Carlo interface.
* Conflict: `+opengl` is incompatible with `~x`.
* Conflict: `http` is now an unconditional conflict due to dependency issues (see README.md).
* Remove commented-out and non-existent dependencies `avahi`, `kerberos`, `ldap`, `libcxx`, `odbc`, `oracle`, `pythia`, `veccore` (per #13949).
* New and changed options:
* Option `pch` was inadvertently set to `OFF` due to its dependence on a nonexistent variant `pch`. As it happens its value is ignored in the ROOT configuration handling, so there was no deleterious effect. It has been fixed to `ON` to better reflect actual behavior pending enablement of tuntime C++ modules.
* Add new versions 6.18.0{0,2,4}:
* Require CMake 3.9 for 6.18.00+.
* Add conflicts for variants `qt4` and `table` representing ROOT build options for which support was discontinued. Remove redundant conflict on \@master.
* C++ standard is now specified with `-DCMAKE_CXX_STANDARD=X` rather than `-Dcxx=X`.
* Remove old version 5.34.38 (wrong build system).
See README.md for more details of option-related changes.
* Flake8
* `rpath` option is a control option rather than a feature.
|
|
|
|
* Add new DD4hep release and some forgotten build requirements
* PR review suggestions
Use master naming convention for development branch, and put versions in decreasing order.
|
|
|
|
* rose: Update boost dependency for rose
* rose: Updated rose to version 0.9.12.45
* rose: Updated jdk dependency
* rose: Updated rose to version 0.9.13.0
* rose: Fixed formatting
* rose: Added maintainer and switch dependency to java@8
|
|
Add the latest release of openPMD-api, v0.10.2
|
|
* add new package : openresty
* remove rc version
|
|
* add new package : nicstat
* refine nicstat edit operation
* refine CMODEL regex expression
|
|
* Added package for Half C++ header-only library.
Fixed an checksum for Hydrogen 1.3.2. Cleaned up the Clara package to
not create an empty bin directory.
* Fixed flake8
* Added maintainer
|
|
* add new package : haproxy
* add url_for_version() and refine haproxy build code
|
|
* add new package : cosbench
* add cosbench depends and remove unstable version
* Update var/spack/repos/builtin/packages/cosbench/package.py
Co-Authored-By: Adam J. Stewart <ajstewart426@gmail.com>
* Update var/spack/repos/builtin/packages/cosbench/package.py
Co-Authored-By: Adam J. Stewart <ajstewart426@gmail.com>
|
|
* tf verbosity
* Add more debug flags
modified: var/spack/repos/builtin/packages/py-tensorflow/package.py
* Fix later
* More debug/verbose options.
modified: var/spack/repos/builtin/packages/py-tensorflow/package.py
|
|
|
|
|
|
Thanks!
|
|
|
|
|
|
Vendors for ARM come out of `/proc/cpuinfo` as hex numbers instead of readable strings.
- Add support for associating vendor names with the hex numbers.
- Also move these mappings from Python code to `microarchitectures.json`
- Move darwin feature name mappings to `microarchitectures.json` as well
|
|
MRNet had an optional dependency on `cti`, which doesn't exist. Remove
it from the repo, at least until someone adds `cti` back.
|
|
|
|
|
|
* add new package : py-s3cmd
* add some depends and move depends behind the version
|
|
Before this commit we used to run the entire unit test suite
in the presence of a failure. Since we currently rely a lot
on the state of the filesystem etc. the end report was most
of the time showing spurious failures that were a consequence
of the first failing test.
This PR makes unit tests exit at the first failing test
Also, pin codecov at v4.5.4 (last one supporting Python 2.6)
|
|
|
|
* when constructing package hash, default to including a method in the content hash if we can't determine whether it would be included by examining the AST
* add a test for updated content-hash calculations
* refactor content hash tests to eliminate repeated lines
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|