alpaka-group · SimeonEhrig · Nov 15, 2023 · Sep 13, 2023 · Sep 13, 2023 · Sep 13, 2023
diff --git a/.zenodo.json b/.zenodo.json
@@ -7,6 +7,11 @@
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
       "orcid": "0000-0003-3396-6154"
     },
+    {
+      "name": "Bastrakova, Kseniia",
+      "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
+      "orcid": "0000-0001-8970-5098"
+    },
     {
       "name": "Bocci, Andrea",
       "affiliation": "CERN",
@@ -22,44 +27,48 @@
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
       "orcid": "0000-0002-8218-3116"
     },
+    {
+      "name": "Ferragina, Luca",
+      "affiliation": "CERN"
+    },
     {
       "name": "Gruber, Bernhard Manfred",
       "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf, CERN",
       "orcid": "0000-0001-7848-1690"
     },
     {
-      "name": "Huebl, Axel",
-      "affiliation": "Lawrence Berkeley National Laboratory",
-      "orcid": "0000-0003-1943-7141"
+      "name": "Kaever, Christian",
+      "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf"
     },
     {
       "name": "Kelling, Jeffrey",
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
       "orcid": "0000-0003-1761-2591"
     },
     {
-      "name": "Pantaleo, Felice",
+      "name": "Martin-Haugh, Stewart",
+      "affiliation": "STFC Rutherford Appleton Laboratory",
+      "orcid": "0000-0001-9457-1928"
+    },
+    {
+      "name": "Perego, Aurora",
       "affiliation": "CERN",
-      "orcid": "0000-0003-3266-4357"
+      "orcid": "0000-0003-1576-6757"
     },
     {
       "name": "Stephan, Jan",
       "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf",
       "orcid": "0000-0001-7839-4386"
     },
-    {
-      "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": "Worpitz, Benjamin",
-      "affiliation": "LogMeIn Inc."
+      "name": "Young, Jeffrey",
+      "affiliation": "Georgia Institute of Technology",
+      "orcid": "0000-0001-9841-4057"
     }
   ],
   "contributors": [
@@ -73,6 +82,12 @@
       "affiliation": "TU Dresden",
       "type": "Other"
     },
+    {
+      "name": "Hübl, Axel",
+      "affiliation": "Lawrence Berkeley National Laboratory",
+      "orcid": "0000-0003-1943-7141",
+      "type": "Other"
+    },
     {
       "name": "Knespel, Maximilian",
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
@@ -99,6 +114,12 @@
       "affiliation": "JetBrains",
       "type": "Other"
     },
+    {
+      "name": "Pantaleo, Felice",
+      "affiliation": "CERN",
+      "orcid": "0000-0003-3266-4357",
+      "type": "Other"
+    },
     {
       "name": "Rogers, David M.",
       "affiliation": "Oak Ridge National Laboratory",
@@ -120,6 +141,12 @@
       "affiliation": "Deutsches Zentrum für Luft- und Raumfahrt e.V.",
       "type": "Other"
     },
+    {
+      "name": "Vyskočil, Jiří",
+      "affiliation": "CASUS, Helmholtz-Zentrum Dresden-Rossendorf",
+      "orcid": "0000-0001-8822-0929",
+      "type": "Other"
+    },
     {
       "name": "Werner, Matthias",
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",
@@ -130,6 +157,11 @@
       "affiliation":"TU Dresden",
       "type": "Other"
     },
+    {
+      "name": "Worpitz, Benjamin",
+      "affiliation": "LogMeIn Inc.",
+      "type": "Other"
+    },
     {
       "name": "Zacharias, Malte",
       "affiliation": "Helmholtz-Zentrum Dresden-Rossendorf",

diff --git a/CHANGELOG.md b/CHANGELOG.md
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
@@ -2,19 +2,19 @@
 
 ## Formatting
 
-Please format your code before before opening pull requests using clang-format 14 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.
 
 ### Visual Studio and CLion
 Suport 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
-First install clang-format-14. Instructions therefore can be found on the web.
+First install clang-format-16. Instructions therefore can be found on the web.
 To format your changes since branching off develop, you can run this command in bash:
 ```
-git clang-format-14 develop
+git clang-format-16 develop
 ```
 To format all code in your working copy, you can run this command in bash:
 ```
-find -iname '*.cpp' -o -iname '*.hpp' | xargs clang-format-14 -i
+find -iname '*.cpp' -o -iname '*.hpp' | xargs clang-format-16 -i
 ```
diff --git a/README.md b/README.md
diff --git a/cmake/alpakaCommon.cmake b/cmake/alpakaCommon.cmake
@@ -614,7 +614,6 @@ if(alpaka_ACC_SYCL_ENABLE)
 
         if(alpaka_SYCL_ONEAPI_FPGA)
             target_compile_definitions(alpaka INTERFACE "ALPAKA_SYCL_ONEAPI_FPGA")
-            alpaka_set_compiler_options(DEVICE target alpaka "-fintelfpga")
 
             if(alpaka_SYCL_ONEAPI_FPGA_MODE STREQUAL "emulation")
                 target_compile_definitions(alpaka INTERFACE "ALPAKA_FPGA_EMULATION")

diff --git a/docs/source/conf.py b/docs/source/conf.py
@@ -13,7 +13,7 @@
 # The short X.Y version.
 version = u'1.0.0'
 # The full version, including alpha/beta/rc tags.
-release = u'1.0.0-develop'
+release = u'1.0.0-rc1'
 
 # The master toctree document.
 master_doc = 'index'

diff --git a/docs/source/dev/style.rst b/docs/source/dev/style.rst
@@ -14,13 +14,13 @@ whitespace and braces automatically. Usage:
 
 .. code-block:: bash
 
-  clang-format-14 -i <sourcefile>
+  clang-format-16 -i <sourcefile>
 
 * If you want to format the entire code base execute the following command from alpaka's top-level directory:
 
 .. code-block:: bash
 
-  find example include test -name '*.hpp' -o -name '*.cpp' | xargs clang-format-14 -i
+  find example include test -name '*.hpp' -o -name '*.cpp' | xargs clang-format-16 -i
 
 Windows users should use `Visual Studio's native clang-format integration
 <https://devblogs.microsoft.com/cppblog/clangformat-support-in-visual-studio-2017-15-7-preview-1/>`.

diff --git a/include/alpaka/atomic/AtomicOmpBuiltIn.hpp b/include/alpaka/atomic/AtomicOmpBuiltIn.hpp
@@ -187,7 +187,8 @@ namespace alpaka
 #        pragma omp atomic capture compare
                 {
                     old = ref;
-                    ref = (ref <= value) ? ref : value;
+                    if(value < ref)
+                        ref = value;
                 }
                 return old;
             }
@@ -205,7 +206,8 @@ namespace alpaka
 #        pragma omp atomic capture compare
                 {
                     old = ref;
-                    ref = (ref >= value) ? ref : value;
+                    if(value > ref)
+                        ref = value;
                 }
                 return old;
             }
@@ -217,21 +219,12 @@ namespace alpaka
         {
             ALPAKA_FN_HOST static auto atomicOp(AtomicOmpBuiltIn const&, T* const addr, T const& value) -> T
             {
+                // TODO(bgruber): atomic increment with wrap around is not implementable in OpenMP 5.1
                 T old;
-                auto& ref(*addr);
-// atomically update ref, but capture the original value in old
-#        if BOOST_COMP_GNUC
-#            pragma GCC diagnostic push
-#            pragma GCC diagnostic ignored "-Wconversion"
-#        endif
-#        pragma omp atomic capture compare
+#        pragma omp critical(AlpakaOmpAtomicOp)
                 {
-                    old = ref;
-                    ref = ((ref >= value) ? 0 : (ref + 1));
+                    old = AtomicInc{}(addr, value);
                 }
-#        if BOOST_COMP_GNUC
-#            pragma GCC diagnostic pop
-#        endif
                 return old;
             }
         };
@@ -242,21 +235,12 @@ namespace alpaka
         {
             ALPAKA_FN_HOST static auto atomicOp(AtomicOmpBuiltIn const&, T* const addr, T const& value) -> T
             {
+                // TODO(bgruber): atomic decrement with wrap around is not implementable in OpenMP 5.1
                 T old;
-                auto& ref(*addr);
-// atomically update ref, but capture the original value in old
-#        if BOOST_COMP_GNUC
-#            pragma GCC diagnostic push
-#            pragma GCC diagnostic ignored "-Wconversion"
-#        endif
-#        pragma omp atomic capture compare
+#        pragma omp critical(AlpakaOmpAtomicOp)
                 {
-                    old = ref;
-                    ref = ((ref == 0) || (ref > value)) ? value : (ref - 1);
+                    old = AtomicDec{}(addr, value);
                 }
-#        if BOOST_COMP_GNUC
-#            pragma GCC diagnostic pop
-#        endif
                 return old;
             }
         };
@@ -293,8 +277,8 @@ namespace alpaka
             ALPAKA_FN_HOST static auto atomicOp(AtomicOmpBuiltIn const&, T* const addr, T const& value) -> T
             {
                 T old;
-// \TODO: Currently not only the access to the same memory location is protected by a mutex but all atomic ops on all
-// threads.
+                // \TODO: Currently not only the access to the same memory location is protected by a mutex but all
+                // atomic ops on all threads.
 #        pragma omp critical(AlpakaOmpAtomicOp)
                 {
                     old = TOp()(addr, value);
@@ -309,8 +293,8 @@ namespace alpaka
                 T const& value) -> T
             {
                 T old;
-// \TODO: Currently not only the access to the same memory location is protected by a mutex but all atomic ops on all
-// threads.
+                // \TODO: Currently not only the access to the same memory location is protected by a mutex but all
+                // atomic ops on all threads.
 #        pragma omp critical(AlpakaOmpAtomicOp2)
                 {
                     old = TOp()(addr, compare, value);

diff --git a/include/alpaka/idx/MapIdx.hpp b/include/alpaka/idx/MapIdx.hpp
@@ -5,6 +5,7 @@
 #pragma once
 
 #include "alpaka/core/Common.hpp"
+#include "alpaka/core/Unreachable.hpp"
 #include "alpaka/vec/Traits.hpp"
 #include "alpaka/vec/Vec.hpp"
 
@@ -50,6 +51,8 @@ namespace alpaka
         }
         else
             static_assert(!sizeof(TElem), "Not implemented");
+
+        ALPAKA_UNREACHABLE({});
     }
 
     //! Maps an N dimensional index to a N dimensional position based on the pitches of a view without padding or a
@@ -89,5 +92,7 @@ namespace alpaka
         }
         else
             static_assert(!sizeof(TElem), "Not implemented");
+
+        ALPAKA_UNREACHABLE({});
     }
 } // namespace alpaka
diff --git a/include/alpaka/intrinsic/IntrinsicCpu.hpp b/include/alpaka/intrinsic/IntrinsicCpu.hpp
@@ -5,6 +5,7 @@
 #pragma once
 
 #include "alpaka/core/BoostPredef.hpp"
+#include "alpaka/core/Unreachable.hpp"
 #include "alpaka/intrinsic/IntrinsicFallback.hpp"
 #include "alpaka/intrinsic/Traits.hpp"
 
@@ -52,6 +53,7 @@ namespace alpaka
                 // Fallback to standard library
                 return static_cast<std::int32_t>(std::bitset<sizeof(UnsignedIntegral) * CHAR_BIT>(value).count());
 #endif
+                ALPAKA_UNREACHABLE(0);
             }
         };
 
@@ -79,6 +81,7 @@ namespace alpaka
 #else
                 return alpaka::detail::ffsFallback(value);
 #endif
+                ALPAKA_UNREACHABLE(0);
             }
         };
     } // namespace trait

diff --git a/include/alpaka/kernel/TaskKernelGpuUniformCudaHipRt.hpp b/include/alpaka/kernel/TaskKernelGpuUniformCudaHipRt.hpp
@@ -50,6 +50,10 @@ namespace alpaka
 {
     namespace detail
     {
+#        if BOOST_COMP_CLANG
+#            pragma clang diagnostic push
+#            pragma clang diagnostic ignored "-Wunused-template"
+#        endif
         //! The GPU CUDA/HIP kernel entry point.
         // \NOTE: 'A __global__ function or function template cannot have a trailing return type.'
         // We have put the function into a shallow namespace and gave it a short name, so the mangled name in the
@@ -74,6 +78,9 @@ namespace alpaka
 #        endif
             kernelFnObj(const_cast<TAcc const&>(acc), args...);
         }
+#        if BOOST_COMP_CLANG
+#            pragma clang diagnostic pop
+#        endif
     } // namespace detail
 
     namespace uniform_cuda_hip

diff --git a/include/alpaka/mem/buf/BufUniformCudaHipRt.hpp b/include/alpaka/mem/buf/BufUniformCudaHipRt.hpp
@@ -310,8 +310,8 @@ namespace alpaka
                 return {
                     dev,
                     reinterpret_cast<TElem*>(memPtr),
-                    [queue = std::move(queue)](TElem* ptr)
-                    { ALPAKA_UNIFORM_CUDA_HIP_RT_CHECK_NOEXCEPT(TApi::freeAsync(ptr, queue.getNativeHandle())); },
+                    [q = std::move(queue)](TElem* ptr)
+                    { ALPAKA_UNIFORM_CUDA_HIP_RT_CHECK_NOEXCEPT(TApi::freeAsync(ptr, q.getNativeHandle())); },
                     extent,
                     static_cast<std::size_t>(width) * sizeof(TElem)};
             }

diff --git a/include/alpaka/mem/buf/Traits.hpp b/include/alpaka/mem/buf/Traits.hpp
@@ -164,14 +164,14 @@ namespace alpaka
     //! this function is provided for convenience in the cases where the difference is not relevant,
     //! and the pinned/mapped memory is only used as a performance optimisation.
     //!
-    //! \tparam TPlatform The platform from which the buffer is accessible.
     //! \tparam TElem The element type of the returned buffer.
     //! \tparam TIdx The linear index type of the buffer.
     //! \tparam TExtent The extent type of the buffer.
+    //! \tparam TPlatform The platform from which the buffer is accessible.
     //! \param host The host device to allocate the buffer on.
     //! \param extent The extent of the buffer.
     //! \return The newly allocated buffer.
-    template<typename TPlatform, typename TElem, typename TIdx, typename TExtent>
+    template<typename TElem, typename TIdx, typename TExtent, typename TPlatform>
     ALPAKA_FN_HOST auto allocMappedBufIfSupported(
         DevCpu const& host,
         TPlatform const& platform,

diff --git a/include/alpaka/platform/PlatformCpu.hpp b/include/alpaka/platform/PlatformCpu.hpp
@@ -16,6 +16,12 @@ namespace alpaka
     //! The CPU device platform.
     struct PlatformCpu : concepts::Implements<ConceptPlatform, PlatformCpu>
     {
+#if defined(BOOST_COMP_GNUC) && BOOST_COMP_GNUC >= BOOST_VERSION_NUMBER(11, 0, 0)                                     \
+    && BOOST_COMP_GNUC < BOOST_VERSION_NUMBER(12, 0, 0)
+        // This is a workaround for g++-11 bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96295
+        // g++-11 complains in *all* places where a PlatformCpu is used, that it "may be used uninitialized"
+        char c = {};
+#endif
     };
 
     namespace trait

diff --git a/include/alpaka/platform/PlatformUniformCudaHipRt.hpp b/include/alpaka/platform/PlatformUniformCudaHipRt.hpp
@@ -28,6 +28,12 @@ namespace alpaka
     template<typename TApi>
     struct PlatformUniformCudaHipRt : concepts::Implements<ConceptPlatform, PlatformUniformCudaHipRt<TApi>>
     {
+#    if defined(BOOST_COMP_GNUC) && BOOST_COMP_GNUC >= BOOST_VERSION_NUMBER(11, 0, 0)                                 \
+        && BOOST_COMP_GNUC < BOOST_VERSION_NUMBER(12, 0, 0)
+        // This is a workaround for g++-11 bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96295
+        // g++-11 complains in *all* places where a PlatformCpu is used, that it "may be used uninitialized"
+        char c = {};
+#    endif
     };
 
     namespace trait