* New package: cxxopts
* Use +unicode instead of unicode=True
- Make the unicode option more explicit
* Add two new variants to spack for upcoming 1.5, stable and develop
* Add as maintainer
* Add depends_on on clauses
* Remove unrelated change
* qgis
* Update package.py
QGIS 3.12.1 can use PROJ >= 4.9.3. Therefore both version restrictions on PROJ were incorrect.
https://github.com/qgis/QGIS/blob/final-3_12_1/INSTALL
* Update package.py
Add explanation to (hopefully temporary) removal of hdf5 dependency.
* Remove overly restrictive GRASS version number.
* flake8
Co-authored-by: Elizabeth Fischer <elizabeth.fischer@alaska.edu>
Set location for dependencies specifying explicitly both
the include and lib path. This permits to handle cases where
the libraries are installed in lib64 instead of lib.
fixes#17556fixes#10842closes#18150
- add cuda variant, enabled by default, but conflicting with
strumpack@:3.9.999
- add zfp variant, enabled by default, but conflicting with
strumpack@:3.9.999
- update minimum CMake version to 3.11
- for version 4.0.0:, do not use mpi wrappers. v4.0.0 uses CMake
MPI targets
- for version 4.0.0, add dependency on butterflypack@1.2.0:
- remove versions 3.1.0 and older
- make parmetis variant True by default
- add TODO for slate variant (spack package not ready yet)
While I believe there must have been a reason to restrict libtool to <=
2.4.2, adios compiles just fine with libtool 2.4.6 for me.
In fact, without this change, I'm getting this error:
libtool: Version mismatch error. This is libtool 2.4.6, but the
libtool: definition of this LT_INIT comes from libtool 2.4.2.
libtool: You should recreate aclocal.m4 with macros from libtool 2.4.6
This doesn't make much sense, since spack did build libtool@2.4.2 as a
dependency, and was supposedly trying to use it. My guess is that on
this system (NERSC's cori) the system libtool in /usr/bin, which is
2.4.6 somehow got picked up partially.
Semi-recently the lua spackage was updated to explicitly add libtinfow
to the lua build line. Ncurses provides this but only when the +termlib
variant is enabled
* New interface reconstruction package
* forgot to put in CMake option for Jali
* cleanup whitespace
* fix lines with more than 79 chars
* more long line cleanup
* fix typo WONTON_ENABLE_Kokkos ---> TANGRAM_ENABLE_Kokkos
Co-authored-by: Rao Garimella <rao@abyzou.lanl.gov>
* New interface reconstruction package
* forgot to put in CMake option for Jali
* cleanup whitespace
* fix lines with more than 79 chars
* more long line cleanup
Co-authored-by: Rao Garimella <rao@abyzou.lanl.gov>
See #18033
libssh seemed to detect and link to system krb5 libraries if found
to provide gssapi support, causing issues/system dependencies/etc.
We add a boolean variant gssapi
If +gssapi, the spack krb5 package is added as a dependency.
If ~gssapi, the Cmake flags are adjusted to not use gssapi so that
does not link to any krb5 package.
xz-utils already builds a shared library. The +pic variant adds the
compiler pic flag to the static archive so that it can be linked into
another shared library.
* Add Collier and SysCalc recipes
* Remove extra syscalc version
* Build collier with -j1 for @:1.2.4
* Add recipe for gosam-contrib
* Update gosam-contrib recipe with 'provides'
* Madgraph recipe, first version
* Finalize madgraph recipe + flake8
* Make py2 version of madgraph default; fix hash for syscalc; fix patch
* Handle virtual packages (#3)
* Update package.py
* Update packages.yaml
* Remove virtual packages - pt. 1
* Remove virtual packages - pt. 2
* Changes from review - pt. 1
* Changes from code review - pt. 2
* Update var/spack/repos/builtin/packages/collier/package.py
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* Update var/spack/repos/builtin/packages/madgraph5amc/package.py
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* Add hash for version 2.7.2 (available in our private mirror)
* Fixes for 2.7.3 family
* Patches for 2.7.3{.py3,}{.atlas,}
* Fix hash of syscalc
* Hack to fix concretization (2.7.3 matches 2.7.3.py3)
* Add conflict statement (reported to devs)
* Apply suggestions from code review
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* Update package.py
* Delete madgraph5amc-2.7.2.atlas.patch
* Delete madgraph5amc-2.7.2.patch
* Update package.py
* Apply suggestions from code review
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
Co-authored-by: iarspider <iarpsider@gmail.com>
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* Adding new packages Mvapich2x and Mvapich2-GDR which can be installed only via binary mirrors
* Added docstring descriptions to both packages
* Removed variant wrapper for cuda dependencies
* Fixed multiple flake8 errors
* Updated APIs to pass unit tests
* Updated APIs for MVAPICH2-X package and fixed flake8 warnings for MVAPICH2-GDR
* Changed url back to single line
* Removed extra parantesis around URL string
Co-authored-by: nithintsk <nithintsk@github.com>
* [root] add dataframe cmake option
@chissg @HadrienG2 @drbenmorgan
This has been a separate cmake option starting v6-19 I believe: 31292b9082
It should default to true -- not sure why, but this recipe sets it to off.
I could add a variant too, but since it has become an integral part of root and doesn't introduce extra dependencies, I'd propose to just set it to true like I do here.
* Update package.py
* Update the change to add gomp compatibity to llvm-openmp.
* Update the change to add gomp compatibity to llvm-openmp using append instead of extend.
* Fix flake8 issue.
Co-authored-by: Jim Galarowicz <jgalarowicz@newmexicoconsortium.org>
* pFUnit: Added support for version 4
pFUnit v4 uses submodules, so we must fetch from the repo rather
than grabbing the tarball (see #11642).
* pFUnit: Added conflicts
pFUnit 4 causes an internal compiler error with gcc 7.2.0, and
several pFUnit versions are incompatible with shared libraries.
* pFUnit: Added conflicts for version 4
Verson 4 uses Fortran 2008 features and cannot be built with gcc
compilers prior to 8.4.
* pFUnit: Fixed conflicts/dependencies as suggested
* pFUnit: Version 4 no longer fetches from git
Checksummable files are fetched instead.
* pFUnit: Simplify major version check
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* pFUnit: Removed unnecessary patch for v4
The patch is still applied to v3.
* pFUnit: Modified MPI flag for v4
pFUnit v3 and v4 use different CMake flags to enable/disable MPI
support. Also added a conflict for v3 with MPI enabled using
gfortran 10, since newer gfortran is more finicky about datatypes.
* pFUnit: Rearranged mpi logic
* pFUnit: changed m4 to a build dependency
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* pFUnit: Added URL back
I did not realize it was needed by "spack versions" and
"spack checksum". Thanks @adamjstewart!
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
When the user explicitly sets ~fortran, mpich builds without fortran
support. This will make building C/C++ libraries using clang easier,
since clang does not offer a fortran compiler by default (yet).
Since the user has to disable Fortran support explicitly, this change
is not breaking.
* Loosen Axom's variants, add shared variant for axom, fix clang/xlf rpath'ing problem on blueos
* Fix flake8
* Add main branch to list of known git branches
fixes#18028
Since now external packages support multiple modules
the correct thing to do is to check if the name of the
*first* module to be loaded contains the string "cray"
`cmake @3.16.3` is the version provided by Ubuntu 20.04. Adding this version here avoids the warning
```
==> Warning: Missing a source id for cmake@3.16.3
```
when using the system `cmake`.
* Spack recipes for ROCm Stage 1 Build components
* fix flake8 errors
* fixes for flake8 errors
* Add a patch for cmake 3.x suport
* Fix rpath issue where hsa-rocr-dev does not allow it to be filled in by spack
* Remove inherited cmake args from comgr
* Make hsakmt-roct compile: no -Werror because with const cast in numa, and actually add the numa dependency
* Remove redundant cmake args which is inherited
* Fix some dependencies
* Fix some python 2.x compatibilities
* Add amd gpu targets to rocfft
* Make comgr a link dep of rocm-dbgapi and remove redundant cmake args
* Remove redundant cmake args
* Remove more redundant cmake args
* Final redundant args
* Use cmake 3.x instead of a fixed version
* Remove random variable
* Use installed rocclr instead of nonexisting directory
* Don't build outside the staging folder
* Deploy some missing cmake target file
* Formatting
* Fix target list
* Properly handle the rocclr dependency
* Formatting
* Fix vermin test
* Make all 3.5.0 package depend exactly on eachother
* Add a few missing link dependencies
* Fix flake8
* Remove some other redundant flags
* Add gcc install prefix for gcc builds of llvm-amdgpu
* review changes for the spack recipes
* Do not hard-code versions
* Fix atmi install
- no more relative rpaths outside of install directory (required patch)
- fix build -> link dependencies
- remove unused build dependency
* Fix flake8 errors
* Remove unused variable and make things python 2.x compatible
* Fix flake8
* Move compiler config from rocfft -> hipcc
* Remove redundant dependency on fftw-api
* Remove redundant import
* Avoid hitting the ROCM_PATH variable altogether with a patch; also just fill in all variables
* Add missing deps z3, zlib and ncurses+termlib to llvm-amdgpu
* Fix perl shebang and add dep
* Fix typo and patch HIP_CLANG_ROOT detection in hip's cmake files
* fixing build failure due z3 and adding zlib for rocgdb
* new changes to add z3,curses dependency for llvm-amdgpu
* fix flake8 error
Co-authored-by: root <root@localhost.localdomain>
Co-authored-by: Harmen Stoppels <harmenstoppels@gmail.com>
This is needed because libcuda is used by the driver,
whereas libcudart is used by the runtime. CMake searches
for cudart instead of cuda.
On LLNL LC systems, libcuda is only found in compat and
stubs directories, meaning that the lookup of libraries
fails.
* Libtool: add spack external find support
* Less specific regex
* match -> search
* Clarify that min returns first alphabetically, not shortest
* Simplify version determination
* add py-ufl package from fenics
* add py-fiat package from fenics
* add py-ffcx package from fenics
* add py-dijitso package from fenics
* add dolfinx library from fenics
* amend ffcx to use ufl and fiat master branches
* setup variants complex and int64 of dolfinx
* add dolfinx python library as package
* add test dependencies to py-dolfinx
* remove broken doc variant
* remove test dependencies from py-dolfinx
* flake8 fixes to dolfinx and py-dolfinx
* make sure dolfinx cmake picks up the correct python version
* list build phases in py-dolfinx package
* remove unnecessary package url
* make pkgconf a build dependency
* make all python dependencies build+run
* py-ffcx needs py-setuptools to be a build/run dependency to support ffcx executable
* remove unnecessary variants from dolfinx
* add missing dependencies to py-dijitso
* remove stray line from py-dolfinx
* simplify definition of build_directory in py-dolfinx
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* use depends_on("python") rather than extends("python") in py-ffcx
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* use depends_on("python") rather than extends("python") in py-fiat
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* use depends_on("python") rather than extends("python") in py-ufl
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* rename py-fiat to py-fenics-fiat
* rename py-ufl to py-fenics-ufl
* fix error in depends_on(petsc) definition
* add missing dep on numpy to py-fenics-fiat
* specify python@3.8: as requirement for all fenics components
* use tuples rather than list for depends_on type=
* specify eigen@3.3.7: as dependency for dolfinx
* add js947 and chrisrichardson as maintainers for the fenics packages
* remove scipy dependency from py-dolfinx
* rename package py-ffcx -> py-fenics-ffcx
* rename package dolfinx -> fenics-dolfinx
* rename package py-dolfinx -> py-fenics-dolfinx
* remove pointless URL from py-fenics-dolfinx package
* rename package py-dijitso -> py-fenics-dijitso
* formatting
* remove unecessary cmake args from fenics-dolfinx
* revert py-fenics-fiat python version to 3:
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* revert py-fenics-ufl python version to 3.5:
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* add conflict to fenics-dolfinx for C++17 support
* revert py-fenics-ffcx python version to 3.5:
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
* pbbam: fix build error
* Update var/spack/repos/builtin/packages/pbbam/package.py
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
Packages can implement “detect_version” to support detection
of external instances of a package. This is generally easier
than implementing “determine_spec_details”. The API for
determine_version is similar: for example you can return
“None” to indicate that an executable is not an instance
of a package.
Users may implement a “determine_variants” method for a package.
When doing external detection, executables are grouped by version
and each group results in a single invocation of “determine_variants”
for the associated spec. The method returns a string specifying
the variants for the package. The method may additionally return
a dictionary representing extra attributes for the package.
These will be stored in the spec yaml and can be retrieved
from self.spec.extra_attributes
The Spack GCC package has been updated with an implementation
of “determine_variants” which adds the following extra
attributes to the package: c, cxx, fortran
The YAML config for paths and modules of external packages has
changed: the new format allows a single spec to load multiple
modules. Spack will automatically convert from the old format
when reading the configs (the updates do not add new essential
properties, so this change in Spack is backwards-compatible).
With this update, Spack cannot modify existing configs/environments
without updating them (e.g. “spack config add” will fail if the
configuration is in a format that predates this PR). The user is
prompted to do this explicitly and commands are provided. All
config scopes can be updated at once. Each environment must be
updated one at a time.
* KNEM url updated
Signed-off-by: Alex Margolin <alex.margolin@huawei.com>
* Allow UCX to be built against KNEM and XPMEM
Signed-off-by: Alex Margolin <alex.margolin@huawei.com>
Older versions do not compile correctly. New users should use 2.004,
not any of the older versions.
Co-authored-by: Adam J. Stewart <ajstewart426@gmail.com>
FlatCC has been removed from UnifyFS as a dependency on the develop
branch and for future releases.
spath is now an optional dependency for UnifyFS to normalize relative
paths provided by the user.
The tcl module for r-dorng will fail to load due to the [] characters in
the description. This happens for Tcl formatted modules loaded by Lmod
at least.
```
module load r-dorng-1.7.1-gcc-9.2.0-wtq7bne
Lmod has detected the following error: .../spack/share/spack/modules/linux-centos7-broadwell/r-dorng-1.7.1-gcc-9.2.0-wtq7bne:(r-dorng-1.7.1-gcc-9.2.0-wtq7bne):
invalid command name "L'Ecuyer"
```
Split text for short and long descriptions.
