Upgrade to clang-format-16

.clang-format

@@ -6,15 +6,16 @@ DisableFormat: false
 
 AccessModifierOffset: -4
 AlignAfterOpenBracket: AlwaysBreak
+AlignArrayOfStructures: None
 AlignConsecutiveAssignments: false
 AlignConsecutiveBitFields: false
 AlignConsecutiveDeclarations: false
 AlignConsecutiveMacros: false
 AlignEscapedNewlines: Right
 AlignOperands: Align
-AlignTrailingComments: false
+AlignTrailingComments:
+  Kind: Never
 AllowAllArgumentsOnNextLine: false
-AllowAllConstructorInitializersOnNextLine: false
 AllowAllParametersOfDeclarationOnNextLine: false
 AllowShortBlocksOnASingleLine: Never
 AllowShortCaseLabelsOnASingleLine: false
@@ -29,78 +30,109 @@ AlwaysBreakTemplateDeclarations: Yes
 BinPackArguments: false
 BinPackParameters: false
 BitFieldColonSpacing: Both
+BreakAfterAttributes: Never
 BreakBeforeBinaryOperators: All
 BreakBeforeBraces: Allman
-BreakBeforeConceptDeclarations: true
+BreakBeforeConceptDeclarations: Always
+BreakBeforeInlineASMColon: OnlyMultiline
 BreakBeforeTernaryOperators: true
 BreakConstructorInitializers: BeforeComma
 BreakInheritanceList: BeforeComma
 BreakStringLiterals: true
 ColumnLimit: 119
 CommentPragmas:  '^ COMMENT pragma:'
 CompactNamespaces: false
-ConstructorInitializerAllOnOneLineOrOnePerLine: true
 ConstructorInitializerIndentWidth: 4
 ContinuationIndentWidth: 4
 Cpp11BracedListStyle: true
-DeriveLineEnding: true
 DerivePointerAlignment: false
+EmptyLineAfterAccessModifier: Never
+EmptyLineBeforeAccessModifier: Always
 ExperimentalAutoDetectBinPacking: false
 FixNamespaceComments: true
 IncludeBlocks: Regroup
 IncludeIsMainRegex: '(Test)?$'
 IncludeIsMainSourceRegex: ''
+IndentAccessModifiers: false
 IndentCaseBlocks: true
 IndentCaseLabels: false
 IndentExternBlock: AfterExternBlock
 IndentGotoLabels: true
 IndentPPDirectives: AfterHash
-IndentRequires: true
+IndentRequiresClause: false
 IndentWidth: 4
 IndentWrappedFunctionNames: false
+InsertBraces: false
+InsertNewlineAtEOF: true
+IntegerLiteralSeparator:
+  Binary: 4
+  Decimal: 3
+  DecimalMinDigits: 7
+  Hex: 4
 KeepEmptyLinesAtTheStartOfBlocks: false
+LambdaBodyIndentation: Signature
+LineEnding: DeriveLF
 MacroBlockBegin: ''
 MacroBlockEnd:   ''
 MaxEmptyLinesToKeep: 2
 NamespaceIndentation: All
+PackConstructorInitializers: CurrentLine
 PenaltyBreakAssignment: 2
 PenaltyBreakBeforeFirstCallParameter: 19
 PenaltyBreakComment: 300
 PenaltyBreakFirstLessLess: 120
+PenaltyBreakOpenParenthesis: 0 # default made explicit here
 PenaltyBreakString: 1000
 PenaltyBreakTemplateDeclaration: 10
 PenaltyExcessCharacter: 1000000
+PenaltyIndentedWhitespace: 0 # default made explicit here
 PenaltyReturnTypeOnItsOwnLine: 1000
 PointerAlignment: Left
-QualifierAlignment: Right
+PPIndentWidth: -1 # follow IndentWidth
+QualifierAlignment: Custom
+QualifierOrder: ['friend', 'static', 'inline', 'constexpr', 'type', 'const', 'volatile', 'restrict']
+ReferenceAlignment: Pointer # follow PointerAlignment
 ReflowComments: true
+RemoveBracesLLVM: false
+RemoveSemicolon: false
+RequiresClausePosition: WithPreceding
+RequiresExpressionIndentation: OuterScope
+ShortNamespaceLines: 0
 SortIncludes: true
-SortUsingDeclarations: true
+SortUsingDeclarations: Lexicographic
+SeparateDefinitionBlocks: Always
 SpaceAfterCStyleCast: true
 SpaceAfterLogicalNot: false
 SpaceAfterTemplateKeyword: false
+SpaceAroundPointerQualifiers: Default # follow PointerAlignment
 SpaceBeforeAssignmentOperators: true
 SpaceBeforeCaseColon: false
 SpaceBeforeCpp11BracedList: false
 SpaceBeforeCtorInitializerColon: true
 SpaceBeforeInheritanceColon: true
 SpaceBeforeParens: Never
 SpaceBeforeRangeBasedForLoopColon: true
+SpaceBeforeSquareBrackets: false
 SpaceInEmptyBlock: false
 SpaceInEmptyParentheses: false
 SpacesBeforeTrailingComments: 1
 SpacesInAngles:  false
 SpacesInConditionalStatement: false
 SpacesInContainerLiterals: false
 SpacesInCStyleCastParentheses: false
+SpacesInLineCommentPrefix:
+  Minimum: 1
+  Maximum: -1
 SpacesInParentheses: false
 SpacesInSquareBrackets: false
-SpaceBeforeSquareBrackets: false
 TabWidth: 4
 UseCRLF: false
 UseTab: Never
 
 # Project specific options
+#AttributeMacros: []
+#ForEachMacros: []
+#IfMacros: []
 IncludeCategories:
   # Local headers (in "") above all else
   - Regex: '"([A-Za-z0-9.\/-_])+"'
@@ -118,5 +150,21 @@ IncludeCategories:
   # below alpaka's includes
   - Regex: '<([A-Za-z0-9.\/-_])+>'
     Priority: 4
+# Macros: []
+# NamespaceMacros: []
+StatementAttributeLikeMacros:
+  - 'ALPAKA_DEVICE_VOLATILE'
+  - 'ALPAKA_FN_ACC'
+  - 'ALPAKA_FN_EXTERN'
+  - 'ALPAKA_FN_HOST'
+  - 'ALPAKA_FN_HOST_ACC'
+  - 'ALPAKA_FN_INLINE'
+  - 'ALPAKA_STATIC_ACC_MEM_CONSTANT'
+  - 'ALPAKA_STATIC_ACC_MEM_GLOBAL'
+  - 'ALPAKA_UNROLL'
+  - 'ALPAKA_VECTORIZE_HINT'
+#StatementMacros: []
+#TypenameMacros: []
+#WhitespaceSensitiveMacros: []
 
 ...

.github/workflows/ci.yml

@@ -81,9 +81,9 @@ jobs:
     runs-on: ubuntu-20.04
     steps:
     - uses: actions/checkout@v3
-    - uses: DoozyX/clang-format-lint-action@v0.14
+    - uses: DoozyX/clang-format-lint-action@v0.16.2
       with:
-        clangFormatVersion: 14
+        clangFormatVersion: 16
         exclude: './thirdParty'
         inplace: True
     - run: |

example/babelstream/src/Stream.h

@@ -38,7 +38,6 @@ class Stream
     virtual void read_arrays(std::vector<T>& a, std::vector<T>& b, std::vector<T>& c) = 0;
 };
 
-
 // Implementation specific device functions
 void listDevices(void);
 std::string getDeviceName(int const);

example/babelstream/src/main.cpp

@@ -54,7 +54,7 @@
 #endif
 
 // Default size of 2^25
-int ARRAY_SIZE = 33554432;
+int ARRAY_SIZE = 33'554'432;
 unsigned int num_times = 100;
 unsigned int deviceIndex = 0;
 bool use_float = false;
@@ -101,7 +101,6 @@ int main(int argc, char* argv[])
         run<double>();
 }
 
-
 // Run the 5 main kernels
 template<typename T>
 std::vector<std::vector<double>> run_all(Stream<T>* stream, T& sum)
@@ -194,7 +193,6 @@ std::vector<std::vector<double>> run_nstream(Stream<T>* stream)
     return timings;
 }
 
-
 // Generic run routine
 // Runs the kernel(s) and prints output.
 template<typename T>
@@ -426,7 +424,6 @@ void run()
     delete stream;
 }
 
-
 template<typename T>
 void check_solution(unsigned int const ntimes, std::vector<T>& a, std::vector<T>& b, std::vector<T>& c, T& sum)
 {

example/bufferCopy/src/bufferCopy.cpp

@@ -25,7 +25,6 @@ struct PrintBufferKernel
     }
 };
 
-
 //! Tests if the value of the buffer on index i is equal to i.
 struct TestBufferKernel
 {

example/complex/src/complex.cpp

@@ -8,7 +8,6 @@
 #include <cstdint>
 #include <iostream>
 
-
 //! Complex numbers demonstration kernel
 struct ComplexKernel
 {

example/heatEquation/src/heatEquation.cpp

@@ -10,7 +10,6 @@
 #include <iostream>
 #include <utility>
 
-
 //! alpaka version of explicit finite-difference 1d heat equation solver
 //!
 //! Solving equation u_t(x, t) = u_xx(x, t) using a simple explicit scheme with
@@ -45,7 +44,6 @@ struct HeatEquationKernel
     }
 };
 
-
 //! Exact solution to the test problem
 //! u_t(x, t) = u_xx(x, t), x in [0, 1], t in [0, T]
 //! u(0, t) = u(1, t) = 0
@@ -59,7 +57,6 @@ auto exactSolution(double const x, double const t) -> double
     return std::exp(-pi * pi * t) * std::sin(pi * x);
 }
 
-
 //! Each kernel computes the next step for one point.
 //! Therefore the number of threads should be equal to numNodesX.
 //! Every time step the kernel will be executed numNodesX-times

example/monteCarloIntegration/src/monteCarloIntegration.cpp

@@ -72,7 +72,6 @@ struct Kernel
     }
 };
 
-
 auto main() -> int
 {
     // Defines and setup.

example/randomCells2D/src/randomCells2D.cpp

@@ -10,9 +10,9 @@
 #include <iostream>
 #include <tuple>
 
-unsigned constexpr NUM_CALCULATIONS = 256;
-unsigned constexpr NUM_X = 127;
-unsigned constexpr NUM_Y = 211;
+constexpr unsigned NUM_CALCULATIONS = 256;
+constexpr unsigned NUM_X = 127;
+constexpr unsigned NUM_Y = 211;
 
 /// Selected PRNG engine for single-value operation
 template<typename TAcc>
@@ -121,7 +121,7 @@ struct RunTimestepKernelVector
             using DistributionResult =
                 typename RandomEngineVector<TAcc>::template ResultContainer<float>; // Container type which will store
                                                                                     // the distribution results
-            unsigned constexpr resultVectorSize = std::tuple_size_v<DistributionResult>; // Size of the result vector
+            constexpr unsigned resultVectorSize = std::tuple_size_v<DistributionResult>; // Size of the result vector
             alpaka::rand::UniformReal<DistributionResult> dist; // Vector-aware distribution function
 
 

example/randomStrategies/src/randomStrategies.cpp

@@ -12,13 +12,14 @@
 #include <iostream>
 
 // This example generates NUM_ROLLS of random events for each of NUM_POINTS points.
-unsigned constexpr NUM_POINTS = 2000; ///< Number of "points". Each will be  processed by a single thread.
-unsigned constexpr NUM_ROLLS = 2000; ///< Amount of random number "dice rolls" performed for each "point".
+constexpr unsigned NUM_POINTS = 2000; ///< Number of "points". Each will be  processed by a single thread.
+constexpr unsigned NUM_ROLLS = 2000; ///< Amount of random number "dice rolls" performed for each "point".
 
 /// Selected PRNG engine
 // Comment the current "using" line, and uncomment a different one to change the PRNG engine
 template<typename TAcc>
 using RandomEngine = alpaka::rand::Philox4x32x10<TAcc>;
+
 // using RandomEngine = alpaka::rand::engine::cpu::MersenneTwister;
 // using RandomEngine = alpaka::rand::engine::cpu::TinyMersenneTwister;
 // using RandomEngine = alpaka::rand::engine::uniform_cuda_hip::Xor;
@@ -159,7 +160,6 @@ struct InitRandomKernel<Strategy::offset>
     }
 };
 
-
 /// Fill the result buffer with random "dice rolls"
 struct FillKernel
 {
@@ -217,7 +217,7 @@ struct Writer;
 template<>
 struct Writer<Strategy::seed>
 {
-    void static save(float const* buffer, Box const& box)
+    static void save(float const* buffer, Box const& box)
     {
         saveDataAndShowAverage("out_seed.csv", buffer, box);
     }
@@ -226,7 +226,7 @@ struct Writer<Strategy::seed>
 template<>
 struct Writer<Strategy::subsequence>
 {
-    void static save(float const* buffer, Box const& box)
+    static void save(float const* buffer, Box const& box)
     {
         saveDataAndShowAverage("out_subsequence.csv", buffer, box);
     }
@@ -235,7 +235,7 @@ struct Writer<Strategy::subsequence>
 template<>
 struct Writer<Strategy::offset>
 {
-    void static save(float const* buffer, Box const& box)
+    static void save(float const* buffer, Box const& box)
     {
         saveDataAndShowAverage("out_offset.csv", buffer, box);
     }

example/reduce/src/iterator.hpp

@@ -101,7 +101,7 @@ class Iterator
     //! Returns the current element.
     //!
     //! Returns a reference to the current index.
-    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE auto operator*() -> const T&
+    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE auto operator*() -> T const&
     {
         return mData[mIndex];
     }
@@ -123,8 +123,12 @@ class IteratorCpu : public Iterator<T, TBuf>
     //! \param linearizedIndex The linearized index.
     //! \param gridSize The grid size.
     //! \param n The problem size.
-    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE
-    IteratorCpu(TAcc const& acc, TBuf const* data, uint32_t linearizedIndex, uint32_t gridSize, uint64_t n)
+    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE IteratorCpu(
+        TAcc const& acc,
+        TBuf const* data,
+        uint32_t linearizedIndex,
+        uint32_t gridSize,
+        uint64_t n)
         : Iterator<T, TBuf>(
             data,
             static_cast<uint32_t>((n * linearizedIndex) / alpaka::math::min(acc, static_cast<uint64_t>(gridSize), n)),
@@ -244,8 +248,12 @@ class IteratorGpu : public Iterator<T, TBuf>
     //! \param linearizedIndex The linearized index.
     //! \param gridSize The grid size.
     //! \param n The problem size.
-    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE
-    IteratorGpu(TAcc const&, TBuf const* data, uint32_t linearizedIndex, uint32_t gridSize, uint64_t n)
+    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE IteratorGpu(
+        TAcc const&,
+        TBuf const* data,
+        uint32_t linearizedIndex,
+        uint32_t gridSize,
+        uint64_t n)
         : Iterator<T, TBuf>(data, linearizedIndex, n)
         , mGridSize(gridSize)
     {

example/reduce/src/kernel.hpp

@@ -27,7 +27,7 @@ struct cheapArray
     //! \param index The index of the element to be accessed.
     //!
     //! Returns the requested element per constant reference.
-    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE auto operator[](uint64_t index) const -> const T&
+    ALPAKA_FN_HOST_ACC ALPAKA_FN_INLINE auto operator[](uint64_t index) const -> T const&
     {
         return data[index];
     }

include/alpaka/acc/AccCpuOmp2Blocks.hpp

@@ -129,6 +129,7 @@ namespace alpaka
         {
             using type = AccCpuOmp2Blocks<TDim, TIdx>;
         };
+
         //! The CPU OpenMP 2.0 block accelerator device properties get trait specialization.
         template<typename TDim, typename TIdx>
         struct GetAccDevProps<AccCpuOmp2Blocks<TDim, TIdx>>
@@ -153,6 +154,7 @@ namespace alpaka
                         static_cast<size_t>(AccCpuOmp2Blocks<TDim, TIdx>::staticAllocBytes())};
             }
         };
+
         //! The CPU OpenMP 2.0 block accelerator name trait specialization.
         template<typename TDim, typename TIdx>
         struct GetAccName<AccCpuOmp2Blocks<TDim, TIdx>>