Merge pull request #263 from chillenzer/alpaka-1.2.0

Alpaka 1.2.0

INSTALL.md

@@ -4,7 +4,7 @@ Install
  - C++14 compiler (clang, gcc, hipcc, icc, nvcc)
   - *Debian/Ubuntu:* `sudo apt-get install gcc build-essential`
   - *Arch Linux:* `sudo pacman -S base-devel`
- - `alpaka` 1.1.0
+ - `alpaka` 1.2.0
   - included as git submodule
  - `boost` >= 1.65.1
    - dependency of alpaka

alpaka/.github/workflows/gh-pages.yml

@@ -6,6 +6,8 @@ on:
   push:
     branches:
       - develop
+env:
+  ALPAKA_CI_OS_NAME: "Linux"
 
 jobs:
   gh-pages:

alpaka/.gitignore

@@ -8,6 +8,7 @@
 
 # tmp files
 *~
+.*.swp
 
 # netbeans project files
 /nbproject/

alpaka/.gitlab-ci.yml

@@ -17,7 +17,7 @@ variables:
   # container version of the generated jobs
   # should be merged with ALPAKA_GITLAB_CI_CONTAINER_VERSION
   # see: script/job_generator/generate_job_yaml.py
-  ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION: "3.1"
+  ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION: "3.2"
 
 generate:
   stage: generator
@@ -27,7 +27,9 @@ generate:
     - apk update && apk add python3~=3.11 py3-pip
     - pip3 install -r script/job_generator/requirements.txt
     # it is sufficient to verify once, as the same job matrix is generated, verified and then filtered each time
-    - python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --verify --wave compile_only_job -o compile_only.yml
+    # disable verify because we know that the generator is broken: https://github.com/thombashi/allpairspy/pull/10
+    #- python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --verify --wave compile_only_job -o compile_only.yml
+    - python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --wave compile_only_job -o compile_only.yml
     - python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --wave runtime_job_cpu -o runtime_cpu.yml
     - python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --wave runtime_job_gpu -o runtime_gpu.yml
     - cat compile_only.yml

alpaka/.pre-commit-config.yaml

@@ -0,0 +1,45 @@
+minimum_pre_commit_version: 3.2.0  # necessitated by Lucas-C's hooks
+default_install_hook_types: [pre-commit, pre-push]
+exclude: 'thirdParty'
+repos:
+- repo: https://github.com/pre-commit/mirrors-clang-format
+  rev: v16.0.6
+  hooks:
+  - id: clang-format
+    files: \.(cpp|hpp)
+- repo: https://github.com/pre-commit/pre-commit-hooks
+  rev: v4.5.0
+  hooks:
+  - id: no-commit-to-branch
+    args: [-b, develop]
+  - id: check-merge-conflict
+  - id: trailing-whitespace
+    exclude_types: [markdown, rst]
+  - id: end-of-file-fixer
+  - id: check-toml
+  - id: check-yaml
+  - id: mixed-line-ending
+  - id: check-executables-have-shebangs
+  - id: check-shebang-scripts-are-executable
+- repo: https://github.com/Lucas-C/pre-commit-hooks
+  rev: v1.5.4
+  hooks:
+    - id: forbid-tabs
+      types_or: [file]
+      exclude_types: [rst]
+    - id: remove-tabs
+      types_or: [file]
+      exclude_types: [rst]
+    - id: forbid-crlf
+    - id: remove-crlf
+- repo: https://github.com/astral-sh/ruff-pre-commit
+  rev: v0.2.1
+  hooks:
+    - id: ruff
+      types_or: [ python, pyi, jupyter ]
+      # The ignores in here are chosen to conform with the currently
+      # existing code and not motivated any other way.
+      args: [ --fix, --ignore, "F403,F405,E731"]
+    - id: ruff-format
+      types_or: [ python, pyi, jupyter ]
+      args: ["--line-length", "120"]

alpaka/.zenodo.json

@@ -12,29 +12,46 @@
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
       "orcid": "0000-0002-8218-3116"
     },
+    {
+      "name": "Erdem, Sven",
+      "affiliation": "Helmholtz-Zentrum Berlin"
+    },
+    {
+      "name": "Fila, Mateusz Jakub",
+      "affiliation": "CERN"
+    },
     {
       "name": "Gruber, Bernhard Manfred",
       "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf, CERN",
       "orcid": "0000-0001-7848-1690"
     },
     {
-      "name": "Martin-Haugh, Stewart",
-      "affiliation": "STFC Rutherford Appleton Laboratory",
-      "orcid": "0000-0001-9457-1928"
+      "name": "Lenz, Julian",
+      "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf",
+      "orcid": "0000-0001-5250-0005"
     },
     {
       "name": "Perego, Aurora",
       "affiliation": "CERN",
-      "orcid": "0000-0003-1576-6757"
+      "orcid": "0009-0002-5210-6213"
     },
     {
-      "name": "Tascon, Andres Rios",
-      "affiliation": "Princeton University"
+      "name": "Varvarin, Michael",
+      "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf"
+    },
+    {
+      "name": "Vyskočil, Jiří",
+      "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf",
+      "orcid": "0000-0001-8822-0929"
     },
     {
       "name": "Widera, René",
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
       "orcid": "0000-0003-1642-0459"
+    },
+    {
+      "name": "Yusufoglu, Mehmet",
+      "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf"
     }
   ],
   "contributors": [
@@ -98,6 +115,12 @@
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
       "type": "Other"
     },
+    {
+      "name": "Martin-Haugh, Stewart",
+      "affiliation": "STFC Rutherford Appleton Laboratory",
+      "orcid": "0000-0001-9457-1928",
+      "type": "Other"
+    },
     {
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf, TU Dresden",
       "name": "Matthes, Alexander",
@@ -143,14 +166,13 @@
       "type": "Other"
     },
     {
-      "name": "Vollmer, Daniel",
-      "affiliation": "Deutsches Zentrum für Luft- und Raumfahrt e.V.",
+      "name": "Tascon, Andres Rios",
+      "affiliation": "Princeton University",
       "type": "Other"
     },
     {
-      "name": "Vyskočil, Jiří",
-      "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf",
-      "orcid": "0000-0001-8822-0929",
+      "name": "Vollmer, Daniel",
+      "affiliation": "Deutsches Zentrum für Luft- und Raumfahrt e.V.",
       "type": "Other"
     },
     {

alpaka/CHANGELOG.md

@@ -3,6 +3,65 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
+## [1.2.0] - 2024-10-02
+
+### Added
+
+- device functions to simplify writing kernel code #2337 #2369 #2383
+- support Clang 18 and 19 #2387
+- support oneAPI 2024.2 #2368
+- support for mapped memory allocation for the SYCL backend #2375
+- support for pre-commit #2253
+- support for device and constant global variables in the SYCL backend #2242
+- alpaka::meta::isList, alpaka::meta::ToList and alpaka::meta::toTuple #2269
+- accelerator trait to check for single and multi-threads per block #2263
+- trait IsKernelTriviallyCopyable #2302
+- trait AccIsEnabled #2267
+- documentation: cmake flag to build alpaka benchmarks #2272
+- benchmark: babelstream support for different Accs #2299
+- example: using MdSpan to pass 2D data #2293
+- example: 2D heat equation #2365 #2383
+- example: Convolution #2228 #2220
+
+### Changed
+
+- update cheatsheet.rst  #2398 #2386 #2241
+- signature of `[get|is]ValidWorkDiv*` #2349
+- use shared CUDA libraries by default #2348 #2342
+- add thread count to CPU blocks accelerators #2338
+- link libcudart even when libcurand is not used #2329
+- ctest: display only output of tests, which failed #2322
+- example: Matrix Multiplication use MdSpan #2317
+- move the Complex class to internal namespace #2301
+- run examples with all enabled accelerators #2280
+- template order allocMappedBuf #2270
+- slow getWarpSize problem #2246
+- simplification of workdiv creation #2240
+- benchmarks: move from examples into own directory #2237
+
+### Fixed
+
+- `[get|is]ValidWorkDiv*` #2349 #2335
+- cray clang compiler errors #2392
+- fix and update SYCL targets #2390 #2361
+- single thread acc throw for invalid workdiv fix #2391
+- explicitly call alpaka::detail to achieve SYCL compilation #2385
+- deduction guide for vector #2376
+- issue with device global variables with CUDA 12.4 #2303
+- clang9/nvcc11.2 boost bug #2294
+- HIP: fix CMake relocatable device code option #2290
+- Re-enable AtomicAtomicRef #2288
+- alpaka_add_library relocatable device code #2273
+- forwarding of msvc compiler flag '/Zo' #2266
+- Windows: usage of Idx to alpaka::Idx #2265
+- compiler detection for clang 17 and 18 as CUDA compiler with libstdc++ (gcc) #2256
+- support for non-integral types in Vec generator constructor #2236
+- memcpy warning #2295
+
+### Removed
+
+- support for nvcc11.0 and nvcc11.1 #2310
+
 ## [1.1.0] - 2024-01-18
 
 ### Added

alpaka/CMakeLists.txt

@@ -35,6 +35,7 @@ set_property(GLOBAL PROPERTY USE_FOLDERS ON)
 # Options and Variants
 
 option(alpaka_BUILD_EXAMPLES "Build the examples" OFF)
+option(alpaka_BUILD_BENCHMARKS "Build the benchmarks" OFF)
 
 # Enable the test infrastructure only if alpaka is the top-level project
 if(CMAKE_PROJECT_NAME STREQUAL PROJECT_NAME)
@@ -46,8 +47,9 @@ endif()
 option(alpaka_INSTALL_TEST_HEADER "Install headers of the namespace alpaka::test. Attention, headers are not designed for production code, see documentation." OFF)
 
 include(CMakeDependentOption)
+
 cmake_dependent_option(alpaka_CHECK_HEADERS "Check all alpaka headers as part of the tests whether they can be compiled standalone." OFF BUILD_TESTING OFF)
-cmake_dependent_option(alpaka_USE_INTERNAL_CATCH2 "Use internally shipped Catch2" ON BUILD_TESTING OFF)
+cmake_dependent_option(alpaka_USE_INTERNAL_CATCH2 "Use internally shipped Catch2" ON "BUILD_TESTING OR alpaka_BUILD_BENCHMARKS" OFF)
 
 ################################################################################
 # Internal variables.
@@ -80,7 +82,7 @@ set(_alpaka_INCLUDE_DIRECTORY "${_alpaka_ROOT_DIR}/include")
 set(_alpaka_SUFFIXED_INCLUDE_DIR "${_alpaka_INCLUDE_DIRECTORY}/alpaka")
 
 # the sequential accelerator is required for the tests and examples
-if(alpaka_BUILD_EXAMPLES OR BUILD_TESTING)
+if(alpaka_BUILD_EXAMPLES OR alpaka_BUILD_BENCHMARKS OR BUILD_TESTING)
   if (NOT (alpaka_ACC_GPU_CUDA_ONLY_MODE OR alpaka_ACC_GPU_HIP_ONLY_MODE))
     if (NOT DEFINED alpaka_ACC_CPU_B_SEQ_T_SEQ_ENABLE)
       option(alpaka_ACC_CPU_B_SEQ_T_SEQ_ENABLE "enable alpaka serial accelerator" ON)
@@ -148,6 +150,10 @@ if(alpaka_BUILD_EXAMPLES)
     add_subdirectory("example/")
 endif()
 
+if(alpaka_BUILD_BENCHMARKS)
+    add_subdirectory("benchmarks/")
+endif()
+
 # Only build the tests if alpaka is the top-level project and BUILD_TESTING is ON
 if(CMAKE_PROJECT_NAME STREQUAL PROJECT_NAME AND BUILD_TESTING)
     add_subdirectory("test/")
@@ -184,10 +190,10 @@ if(CMAKE_PROJECT_NAME STREQUAL PROJECT_NAME)
 
     if(alpaka_INSTALL_TEST_HEADER)
         install(DIRECTORY "${_alpaka_SUFFIXED_INCLUDE_DIR}"
-	        DESTINATION "${CMAKE_INSTALL_INCLUDEDIR}")
+            DESTINATION "${CMAKE_INSTALL_INCLUDEDIR}")
     else()
-	install(DIRECTORY "${_alpaka_SUFFIXED_INCLUDE_DIR}"
-	        DESTINATION "${CMAKE_INSTALL_INCLUDEDIR}"
+    install(DIRECTORY "${_alpaka_SUFFIXED_INCLUDE_DIR}"
+            DESTINATION "${CMAKE_INSTALL_INCLUDEDIR}"
                 PATTERN "test" EXCLUDE)
     endif()
 

alpaka/CONTRIBUTING.md

@@ -1,11 +1,34 @@
 # Contributing
 
+Please review our more detailed [Coding Guidelines](https://alpaka.readthedocs.io/en/latest/dev/style.html) as well.
+
+## Pre-commit
+
+This project is set up for use with [pre-commit](https://pre-commit.com). Using it will make your code conform with most
+of our (easily automatable) code style guidelines automatically.
+In very short (for anything further see [pre-commit](https://pre-commit.com)), after running the following in your
+working clone of alpaka
+```bash
+# if not yet done, install the pre-commit executable following https://pre-commit.com
+cd /path/to/alpaka-working-clone
+pre-commit install
+```
+`git` will run a number of checks prior to every commit and push and will refuse to perform the
+pertinent action if they fail. Most of them (like e.g. the formatter) will have automatically altered your working tree
+with the necessary changes such that
+```bash
+git add -u
+```
+will make the next commit pass.
+
 ## Formatting
 
-Please format your code before before opening pull requests using clang-format 16 and the .clang-format file placed in the repository root.
+Please format your code before before opening pull requests using clang-format 16 and the .clang-format file placed in 
+the repository root. If you were using `pre-commit` during your changes, this has happened automatically already. If
+not, find further instructions below.
 
 ### Visual Studio and CLion
-Suport for clang-format is built-in since Visual Studio 2017 15.7 and CLion 2019.1.
+Support for clang-format is built-in since Visual Studio 2017 15.7 and CLion 2019.1.
 The .clang-format file in the repository will be automatically detected and formatting is done as you type, or triggered when pressing the format hotkey.
 
 ### Bash