Merge branch 'master' into extend-clang

Author: Vahid Tavanashad

.github/workflows/build-with-clang.yml

@@ -66,14 +66,23 @@ jobs:
         run: ls ${{ env.ONEAPI_ROOT }}/compiler
 
       - name: Build mkl-service
+<<<<<<< HEAD
+=======
+        shell: bash -l {0}
+>>>>>>> master
         run: |
           source ${{ env.ONEAPI_ROOT }}/setvars.sh
           echo $CMPLR_ROOT
           export CC=$CMPLR_ROOT/bin/icx
           export CFLAGS="${CFLAGS} -fno-fast-math"
           python setup.py develop
 
+<<<<<<< HEAD
       - name: Run mkl-service tests    
+=======
+      - name: Run mkl-service tests
+        shell: bash -l {0}
+>>>>>>> master
         run: |
           source ${{ env.ONEAPI_ROOT }}/setvars.sh
           pytest -s -v --pyargs mkl

CHANGES.rst renamed to CHANGELOG.md

@@ -1,57 +1,56 @@
 =====================
-mkl-service changelog
+`mkl-service` changelog
 =====================
 
+2.3.0
+=====
 
-1.0.0
+Fixed CI to actually execute tests. Populated CBWR constants to match MKL headers.
+Added tests checking that `cbwr_set` and `cbwr_get` round-trip.
+
+2.2.0
 =====
 
-Initial release of `mkl-service` package.
+Closed issues #8, #7 and #5. 
 
-2.0.0
-====
+Extended `mkl.cbwr_set` to recognize `'avx512_e1'`, `'avx512_mic_e1'`, as as strict conditional numerical reproducibility, supported via `'avx2,strict'`, `'avx512,strict'` (see [issue/8](http://github.com/IntelPython/mkl-service/issues/8)).
 
-Rerelease of `mkl-service` package with version bumped to 2.0.0 to avoid version clash with `mkl-service` package from Anaconda.
+Extended `mkl.cbwrt_get()` to mean `mkl.cbwr('all')`.
 
-Improved argument checking, which raises an informative error.
+2.1.0
+=====
 
-Loading the package with `import mkl` initializes Intel(R) MKL library to use LP64 interface (i.e. use of environment variable `MKL_INTERFACE` will not have effect).
+Change in setup script to not use `numpy.distutils` thus removing numpy as build-time dependency.
 
-The choice of threading layer can be controlled with environment variable `MKL_THREADING_LAYER`. However the unset variable is interpreted differently that in Intel(R) MKL itself. If `mkl-service` detects that Gnu OpenMP has been loaded in Python space, the threading layer of Intle(R) MKL will be set to Gnu OpenMP, instead of Intel(R) OpenMP.
+Change in conda-recipe to allow conda build to build the recipe, but ignoring run export on mkl-service coming from mkl-devel metadata. 
 
-2.0.0
-====
+2.0.2
+=====
 
-Work-around for VS 9.0 not having `inline` keyword, allowing the package to build on Windows for Python 2.7
+Correction to `setup.py` to not require Cython at the installation time.
 
 2.0.1
 =====
 
 Re-release, with some changes necessary for public CI builds to work.
 
-2.0.2
+2.0.0
 =====
 
-Correction to `setup.py` to not require Cython at the installation time.
+Work-around for VS 9.0 not having `inline` keyword, allowing the package to build on Windows for Python 2.7
 
-2.1.0
+2.0.0
 =====
 
-Change in setup script to not use `numpy.distutils` thus removing numpy as build-time dependency.
-
-Change in conda-recipe to allow conda build to build the recipe, but ignoring run export on mkl-service coming from mkl-devel metadata. 
-
-2.2.0
-=====
+Rerelease of `mkl-service` package with version bumped to 2.0.0 to avoid version clash with `mkl-service` package from Anaconda.
 
-Closed issues #8, #7 and #5. 
+Improved argument checking, which raises an informative error.
 
-Extended `mkl.cbwr_set` to recognize `'avx512_e1'`, `'avx512_mic_e1'`, as as strict conditional numerical reproducibility, supported via `'avx2,strict'`, `'avx512,strict'` (see [issue/8](http://github.com/IntelPython/mkl-service/issues/8)).
+Loading the package with `import mkl` initializes Intel(R) MKL library to use LP64 interface (i.e. use of environment variable `MKL_INTERFACE` will not have effect).
 
-Extended `mkl.cbwrt_get()` to mean `mkl.cbwr('all')`.
+The choice of threading layer can be controlled with environment variable `MKL_THREADING_LAYER`. However the unset variable is interpreted differently that in Intel(R) MKL itself. If `mkl-service` detects that Gnu OpenMP has been loaded in Python space, the threading layer of Intle(R) MKL will be set to Gnu OpenMP, instead of Intel(R) OpenMP.
 
-2.3.0
+1.0.0
 =====
 
-Fixed CI to actually execute tests. Populated CBWR constants to match MKL headers.
-Added tests checking that `cbwr_set` and `cbwr_get` round-trip.
+Initial release of `mkl-service` package.

LICENSE.txt

@@ -1,4 +1,4 @@
-Copyright (c) 2018-2019, Intel Corporation
+Copyright (c) 2018, Intel Corporation
 
 Redistribution and use in source and binary forms, with or without
 modification, are permitted provided that the following conditions are met:

README.md

@@ -1,4 +1,4 @@
-# ``mkl-service`` - Python package for run-time control of Intel(R) Math Kernel Library.
+# `mkl-service` - Python package for run-time control of Intel® OneAPI Math Kernel Library (OneMKL).
 [![Conda package](https://github.com/IntelPython/mkl-service/actions/workflows/conda-package.yml/badge.svg)](https://github.com/IntelPython/mkl-service/actions/workflows/conda-package.yml)
 [![Build mkl-service with clang](https://github.com/IntelPython/mkl-service/actions/workflows/build-with-clang.yml/badge.svg)](https://github.com/IntelPython/mkl-service/actions/workflows/build-with-clang.yml)
 [![OpenSSF Scorecard](https://api.securityscorecards.dev/projects/github.com/IntelPython/mkl-service/badge)](https://securityscorecards.dev/viewer/?uri=github.com/IntelPython/mkl-service)
@@ -8,19 +8,19 @@
 
 To install conda package, use `conda install -c https://software.repos.intel.com/python/conda/ mkl-service`, or `conda install -c conda-forge mkl-service`.
 
-To install pypi package, use `python -m pip install mkl-service`.
+To install PyPI package, use `python -m pip install mkl-service`.
 
 ---
 
-Intel(R) Math Kernel Library support functions are subdivided into the following groups according to their purpose:
+Intel® OneMKL support functions are subdivided into the following groups according to their purpose:
  - Version Information
  - Threading Control
  - Timing
  - Memory Management
  - Conditional Numerical Reproducibility Control
  - Miscellaneous
 
-A short example, illustrating it use:
+A short example, illustrating its use:
 
 ```python
 import tomopy

examples/example.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2019, Intel Corporation
+# Copyright (c) 2018, Intel Corporation
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:

mkl/__init__.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2024, Intel Corporation
+# Copyright (c) 2018, Intel Corporation
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:

mkl/_mkl_service.pxd

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2024, Intel Corporation
+# Copyright (c) 2018, Intel Corporation
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:

mkl/_mkl_service.pyx

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2024, Intel Corporation
+# Copyright (c) 2018, Intel Corporation
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:

mkl/tests/test_mkl_service.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2024, Intel Corporation
+# Copyright (c) 2018, Intel Corporation
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:

setup.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python
-# Copyright (c) 2018-2024, Intel Corporation
+# Copyright (c) 2018, Intel Corporation
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met: