summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2022-09-20stat: add v4.2.1 (#32691)Gregory Lee1-2/+8
2022-09-19Bump up the version for rocm-5.2.1 release (#32195)Sreenivasa Murthy Kolam49-16/+294
* Bump up the version for rocm-5.2.1-initial commit * Bump up the version for rocm-5.2.1 release * Bump up the version for rocm-5.2.1 release * correct the PROF_API_HEADER_PATH to include * Bump up the version of rocm-openmp-extras for rocm-5.2.1 release * bump up the version of rocwmma for 5.2.1
2022-09-19add pio2_5_9 to spack package (#32716)Jim Edwards1-0/+1
2022-09-19py-pandas: add v1.5.0 (#32714)Adam J. Stewart1-1/+4
2022-09-19Added variant to ensure Arm Forge license acceptance (#32635)kent-cheung-arm1-2/+23
* Added variant to ensure Arm Forge license acceptance * Update var/spack/repos/builtin/packages/arm-forge/package.py Co-authored-by: Tamara Dahlgren <35777542+tldahlgren@users.noreply.github.com>
2022-09-19imagemagick: new variant ghostscript (#32450)Wouter Deconinck1-4/+12
* imagemagick: new variant ghostscript Ghostscript adds about 75 dependencies to an installation of imagemagick (97 without, 172 with Ghostscript). This adds a variant (defaulting to true for backward compatibility) that allows users to turn off Ghostscript support. * imagemagick: be explicit when `--with-gslib` * imagemagick: fix suggestion fail * imagemagick: use spec.prefix.share.font * imagemagick: default ghostscript false * imagemagick: no need for join_path anymore Co-authored-by: Tamara Dahlgren <35777542+tldahlgren@users.noreply.github.com> Co-authored-by: Tamara Dahlgren <35777542+tldahlgren@users.noreply.github.com>
2022-09-19Deprecate Python 2 support (#32615)Adam J. Stewart1-43/+211
This PR deprecates using Spack to install [EOL Python versions](https://endoflife.date/python), including Python 2.7, 3.1–3.6. It also deprecates running Spack with Python 2.7. Going forward, we expect Spack to have the following Python support timeline. ### Running Spack * Spack 0.18 (spring 2022): Python 2.7, 3.5–3.10 * Spack 0.19 (fall 2022): Python 2.7 (deprecated), 3.6–3.11 * Spack 0.20 (spring 2023): Python 3.6–3.11 ### Building with Spack * Spack 0.18 (spring 2022): Python 2.7, 3.1–3.10 * Spack 0.19 (fall 2022): Python 2.7, 3.1–3.6 (deprecated), 3.7–3.11 * Spack 0.20 (spring 2023): Python 3.7–3.11 This is a reboot of #28003. See #31824 for a detailed discussion of the motivation for this PR. If you have concerns about this change, please comment on #31824.
2022-09-19py-datalad: add 0.17.5 (#32676)Manuela Kuhn6-35/+115
* py-datalad: add 0.17.5 * Fix description of py-types-urllib3 * Update var/spack/repos/builtin/packages/py-datalad/package.py Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com> * Update var/spack/repos/builtin/packages/py-datalad/package.py Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com> * Add colorama dependency for windows * Fix importlib-metadata dependency * Update var/spack/repos/builtin/packages/py-pytest/package.py Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com> * Update var/spack/repos/builtin/packages/py-pytest/package.py Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com> * Update var/spack/repos/builtin/packages/py-pytest/package.py Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com> * Use conflict to avoid dependency duplication Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
2022-09-19NCCL should not be dependent on ROCm builds since RCCL is used instead. (#32461)rsanthanam-amd1-3/+3
Allow the nccl flag to be specified even for ROCm builds so that NCCL kernels are included in the build. In this case the NCCL kernels will use RCCL as the backend implementation.
2022-09-19Adding intel-oneapi-itac package (#32658)Stephen Sachs1-0/+42
* Adding intel-oneapi-itac package * Make black happy * add rscohn2 as maintainer * black prefers double quotes Co-authored-by: Stephen Sachs <stesachs@amazon.com> Co-authored-by: Robert Cohn <rscohn2@gmail.com>
2022-09-19Ginkgo: improve smoke testing (#32647)Terry Cojean1-14/+52
* Ginkgo: improve smoke testing * Fix style issues * Pass `self.compiler.{cc,cxx}` to the smoke tests
2022-09-19patch etsf-io tests (#32666)Ashwin Kumar3-0/+59
(Patches from meisam.tabriz@mpcdf.mpg.de and smoke tests from @fangohr)
2022-09-19root: determine_variants cxxstd=20 support (#32673)Wouter Deconinck1-0/+2
This is somewhat ahead of reality, but once/when external ROOT installs with cxxstd=20 are common, we will need this to pick up the correct variant.
2022-09-19fasttree: updating to 2.1.11 (#32683)snehring1-0/+7
2022-09-19Don't run bootstrap on package only PRs (#32689)Massimiliano Culpo1-2/+1
* Don't run bootstrap on package only PRs * Run bootstrap tests when ci.yaml is modified * Test a package only PR * Revert "Test a package only PR" This reverts commit af96b1af60b0c31efcc9a2875ffb1a809ef97241.
2022-09-19py-sip: add v6.6.2 (#32695)Adam J. Stewart1-2/+7
* py-sip: add v6.6.2 * Fix python deptype
2022-09-19edm4hep: allow variant cxxstd 20, sync with podio (#32698)Wouter Deconinck1-1/+5
Now that `podio` can support `cxxstd` variants 17 and 20, we can allow `edm4hep` to use `cxxstd=20` as well, but must ensure that `edm4hep` uses the same `cxxstd` variant as `podio`. Solution as in `celeritas`.
2022-09-19geant4-vmc: add version 6-1-p1 (#32704)Laurent Aphecetche1-0/+1
2022-09-19Fix benchmark recipe after branch name was changed (#32708)iarspider1-1/+1
2022-09-19geant4: new bugfix version 11.0.3 (#32700)Wouter Deconinck1-0/+1
Geant4 has a new bugfix version 11.0 patch-03, [release notes](https://geant4-data.web.cern.ch/ReleaseNotes/Patch.11.0-3.txt). No build system changes.
2022-09-19py-neurora: add 1.1.6.8 (#32684)Manuela Kuhn1-0/+1
2022-09-19py-pydicom: add 2.3.0 (#32681)Manuela Kuhn1-0/+1
2022-09-18Attempt at patching phist find mpi heuristic (#32688)John W. Parent1-1/+1
for updated phist version
2022-09-17Add skip_import to PythonPackage and use it in py-nilearn (#32664)Manuela Kuhn3-36/+58
* Add skip_import to PythonPackage and use it in py-nilearn * Fix dependencies
2022-09-17xyce: Change preferred to last numeric version (#32679)Paul Kuberry1-2/+3
- Also handles new constraints on Trilinos version
2022-09-16[intel-oneapi-compilers] fix CC/CXX (#32677)Robert Cohn1-4/+4
2022-09-16CI: add coverage on Windows (#32610)Massimiliano Culpo1-6/+21
Co-authored-by: Tom Scogland <scogland1@llnl.gov>
2022-09-16bugfix: package hash should affect process, dag, and dunder hashes (#32516)Todd Gamblin5-1/+40
This fixes a bug where two installations that differ only by package hash will not show up in `spack find`. The bug arose because `_cmp_node` on `Spec` didn't include the package hash in its yielded fields. So, any two `Spec` objects that were only different by package hash would appear to be equal and would overwrite each other when inserted into the same `dict`. Note that we could still *install* specs with different package hashes, and they would appear in the database, but we code that needed to put them into data structures that use `__hash__` would have issues. This PR makes `Spec.__hash__` and `Spec.__eq__` include the `process_hash()`, and it makes `Spec._cmp_node` include the package hash. All of these *should* include all information in a spec so that we don't end up in a situation where we are blind to particular field differences. Eventually, we should unify the `_cmp_*` methods with `to_node_dict` so there aren't two sources of truth, but this needs some thought, since the `_cmp_*` methods exist for speed. We should benchmark whether it's really worth having two types of hashing now that we use `json` instead of `yaml` for spec hashing. - [x] Add `package_hash` to `Spec._cmp_node` - [x] Add `package_hash` to `spack.solve.asp.spec_clauses` so that the `package_hash` will show up in `spack diff`. - [x] Add `package_hash` to the `process_hash` (which doesn't affect abstract specs but will make concrete specs correct) - [x] Make `_cmp_iter` report the dag_hash so that no two specs with different process hashes will be considered equal.
2022-09-15adding butterflypack 2.2.0 (#32670)liuyangzhuan1-1/+2
2022-09-15New package: xrdcl-record (#32663)iarspider1-0/+20
* New package: xrdcl-record * Flake-8
2022-09-15Fenicsx v0.5.0 (#32611)Matthew Archer7-7/+45
* initial commit of 0.5.0 changes * updated dependencies * updated ffcx sha * comment style * llvm compilers * introduce pugixml dependency for 0.5.0: * update compilers to support C++20 features * style fixes * xtensor and xtl not needed for basix 0.5.0 and above * Skip to Basix 0.5.1 The 0.5.1 release removes the C++ build dependency on Python that sneaked into the 0.5.0 build system. * Improve depends on version ranges * More dependency version improvements Co-authored-by: Chris Richardson <chris@bpi.cam.ac.uk> Co-authored-by: Garth N. Wells <gnw20@cam.ac.uk>
2022-09-15added missing dependency for py-msgpack that breaks neovim (#32631)Robert Underwood1-1/+2
* added missing dependency for py-msgpack that breaks neovim * Update var/spack/repos/builtin/packages/py-pynvim/package.py Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com> Co-authored-by: Robert Underwood <runderwood@anl.gov> Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
2022-09-15FPLO: add new package (#32580)Glenn Johnson6-0/+373
2022-09-14[py-pip] fix setup_dependent_package in py-pip (#32632)Matthieu Dorier1-1/+1
2022-09-14libcatalyst: use git instead of fixed urls (#32642)Vicente Bolea1-8/+2
The issue is that we are not not able to install (Fetch URL error) any version of catalyst other than the specified in the spack package.py. This very version is accessible only because it is cached by Spack. The real URL does not exist anymore, I believe the reason is that there used to be a tag in catalyst that does not exist anymore.
2022-09-14Bugfix: find_libraries (#32653)John W. Parent2-9/+32
53a7b49 created a reference error which broke `.libs` (and `find_libraries`) for many packages. This fixes the reference error and improves the testing for `find_libraries` by actually checking the extension types of libraries that are retrieved by the function.
2022-09-14Manifest parsing: skip invalid manifest files (#32467)Peter Scheibel3-4/+57
* catch json schema errors and reraise as property of SpackError * no need to catch subclass of given error * Builtin json library for Python 2 uses more generic type * Correct instantiation of SpackError (requires a string rather than an exception) * Use exception chaining (where possible)
2022-09-14llvm: fix detection of LLDB_PYTHON_EXE_RELATIVE_PATH (#32584)Sergey Kosukhin2-0/+35
2022-09-14Allow version to accept the '=' token without activating lexer switch (#32257)psakievich2-54/+126
2022-09-14py-awkward: allow newer, fixed rapidjson versions (#32650)Valentin Volkl1-1/+1
Rapidjson was constrained due to a bug (see https://github.com/spack/spack/pull/29889) , but newer (although untagged) versions of rapidjson have since been added to spack (https://github.com/spack/spack/pull/29869). Tested the build of py-awkward with the latest, works fine.
2022-09-14Add checksum for yoda 1.9.6 (#32648)iarspider1-0/+1
2022-09-14[py-pyprecice] add v2.5.0.0 and v2.5.0.1 and update checksum of v2.4.0.0 ↵Ishaan Desai1-11/+18
(#32551) * Update versions 2.5.0.0 and 2.5.0.1 * Applying review changes * Updating incorrect checksum for v2.4.0.0 * Add for loop to define depends_on for preCICE versions and bindings versions * Formatting * Missing comma
2022-09-14compadre: add v1.5.0, v1.4.1 (#32609)Paul Kuberry1-6/+14
2022-09-14Add hdf5-vol-async v1.3 (#32634)Houjun Tang1-0/+1
2022-09-14py-pytorch-lightning: add v1.7.6 (#32644)Adam J. Stewart1-0/+1
2022-09-14py-tokenize-rt: add missing setuptools dependency (#32639)Adam J. Stewart1-0/+1
2022-09-14phist: add v1.9.5 (#32621)Jonas Thies2-17/+9
2022-09-13berkeley-db: Correct TLS auto-detection (#32592)Erik Schnetter2-0/+29
2022-09-14Revert "e4s ci: restore power builds (#32296)" (#32633)Zack Galbreath3-463/+290
This reverts commit 0d18c32bca8398e2daa736a940154caa235cbacc.
2022-09-13json-glib: add pkgconfig dependency (#32638)Adam J. Stewart1-0/+1