Remove HVX extension support

.github/workflows/cmake.yml

@@ -178,49 +178,6 @@ jobs:
       working-directory: ${{github.workspace}}/build
       run: python3 ./Test.py "-r=.."
 
-  build_and_test_hexagon_cross:
-    runs-on: ubuntu-latest
-
-    strategy:
-      matrix:
-        build_type: [Debug, Release]
-
-    env:
-      HEXAGON_TOOLCHAIN: ${{github.workspace}}/clang+llvm-22.1.0-cross-hexagon-unknown-linux-musl/x86_64-linux-gnu
-
-    steps:
-    - uses: actions/checkout@v6
-
-    - name: Host properties
-      run: lscpu
-
-    - name: Install host dependencies
-      run: sudo apt-get update && sudo apt-get install -y libc++1 libc++abi1 liburing2 libglib2.0-0t64
-
-    - name: Install Hexagon toolchain
-      run: |
-        wget -q https://artifacts.codelinaro.org/artifactory/codelinaro-toolchain-for-hexagon/22.1.0_/clang+llvm-22.1.0-cross-hexagon-unknown-linux-musl.tar.zst
-        tar --zstd -xf clang+llvm-22.1.0-cross-hexagon-unknown-linux-musl.tar.zst
-
-    - name: Configure CMake
-      run: >
-        cmake ./prj/cmake -B ${{github.workspace}}/build
-        -DCMAKE_BUILD_TYPE=${{matrix.build_type}}
-        -DSIMD_TOOLCHAIN="${{env.HEXAGON_TOOLCHAIN}}/bin/hexagon-linux-musl-clang"
-        -DSIMD_TARGET="hexagon"
-        -DSIMD_SYNET=OFF
-
-    - name: Build
-      run: cmake --build ${{github.workspace}}/build --config ${{matrix.build_type}} --parallel$(nproc)
-
-    - name: Test
-      working-directory: ${{github.workspace}}/build
-      run: >
-        ${{env.HEXAGON_TOOLCHAIN}}/bin/qemu-hexagon -cpu any
-        -L ${{env.HEXAGON_TOOLCHAIN}}/target/hexagon-unknown-linux-musl
-        ./Test "-r=.." -m=a -tt=0 "-ot=log_${{matrix.build_type}}.txt" -ts=5 -mt=1
-        -fi=SquareS -fi=ValueSum -fi=y8u -fi=amCo -fi=GetStat -fi=ColSums -fi=RowSums -fi=FillPixel -fi=FillBgra -fi=CorrelationSum -fi=BgrToRgb -fi=BgrToGray -fi=Abs -fi=AddFeat
-
   build_and_test_mingw:
     runs-on: windows-latest
 

README.md

@@ -8,7 +8,7 @@ neural network.
 
 The algorithms are optimized with using of different SIMD CPU extensions. 
 In particular the library supports following CPU extensions: 
-SSE, AVX, AVX-512 and AMX for x86/x64, NEON, SVE, SVE2 for ARM, HVX for Hexagon.
+SSE, AVX, AVX-512 and AMX for x86/x64, NEON, SVE, SVE2 for ARM.
 
 The Simd Library has C API and also contains useful C++ classes and functions to facilitate access to C API. 
 The library supports dynamic and static linking, 32-bit and 64-bit Windows and Linux, 
@@ -59,7 +59,7 @@ Files of CMake build systems are placed in the directory:
 
 `simd/prj/cmake/`
 
-The library can be built for x86/x64, ARM(32/64), and Hexagon platforms using the G++ or Clang compilers.
+The library can be built for x86/x64 and ARM(32/64) platforms using the G++ or Clang compilers.
 Using the native compiler (g++) for the current platform is simple:
 
 	mkdir build
@@ -82,13 +82,6 @@ And for ARM (64 bit):
 	cmake ../prj/cmake -DSIMD_TOOLCHAIN="/your_toolchain/usr/bin/aarch64-linux-gnu-g++" -DSIMD_TARGET="aarch64" -DCMAKE_BUILD_TYPE="Release"
 	make
 
-And for Hexagon with HVX (cross-compilation using the Hexagon Clang toolchain):
-
-	mkdir build
-	cd build
-	cmake ../prj/cmake -DSIMD_TOOLCHAIN="/your_toolchain/bin/hexagon-linux-musl-clang" -DSIMD_TARGET="hexagon" -DCMAKE_BUILD_TYPE="Release"
-	make
-
 As result the library and the test application will be built in the current directory.
 
 There are addition build parameters:
@@ -195,7 +188,7 @@ Also you can use parameters:
 * `-tr=2` a number of test execution repeats.
 * `-ts=1` to print statistics of time of tests execution.
 * `-cc=1` to check c++ API.
-* `-de=2` a flags of SIMD extensions which testing are disabled. Base - 1, 2 - SSE4.1/NEON/HVX, 4 - AVX2/SVE2, 8 - AVX-512BW, 16 - AVX-512VNNI, 32 - AMX-BF16.
+* `-de=2` a flags of SIMD extensions which testing are disabled. Base - 1, 2 - SSE4.1/NEON, 4 - AVX2/SVE2, 8 - AVX-512BW, 16 - AVX-512VNNI, 32 - AMX-BF16.
 * `-wu=100` a time to warm up CPU before testing (in milliseconds).
 * `-pt=1` a boolean flag to pin threads to cpu cores.
 

docs/2026.html

@@ -176,12 +176,15 @@ <h5>Removing</h5>
  <li>HVX optimization of function GetColSums.</li>
  <li>HVX optimization of function GetRowSums.</li>
  <li>HVX optimization of function GetStatistic.</li>
+ <li>HVX optimization of function HistogramConditional.</li>
  <li>HVX optimization of function HistogramMasked.</li>
  <li>HVX optimization of function OperationBinary8u.</li>
  <li>HVX optimization of function SquareSum.</li>
  <li>HVX optimization of function ValueSquareSum.</li>
  <li>HVX optimization of function ValueSquareSums.</li>
  <li>HVX optimization of function ValueSum.</li>
+ <li>Enum value SimdCpuInfoHvx.</li>
+ <li>Stopping of support of HVX extension (Hexagon platform).</li>
 </ul>
 
 <h4>Python wrapper</h4>
@@ -198,6 +201,7 @@ <h5>Improving</h5>
 <h5>Removing</h5>
 <ul>
  <li>Function Simd.Lib.AlphaBlending2x.</li>
+ <li>Enum value Simd.CpuInfo.HVX.</li>
 </ul>
 
 <h4>Test framework</h4>
@@ -284,6 +288,9 @@ <h5>Removing</h5>
 <ul>
  <li>SIMD_INT8_DEBUG cmake option.</li>
  <li>SIMD_INT8_DEBUG_DISABLE macro.</li>
+ <li>SIMD_HVX_DISABLE macro.</li>
+ <li>CMake build for Hexagon platform.</li>
+ <li>Job build_and_test_hexagon_cross in Github actions script for CMake.</li>
 </ul>
 
 <h4>Documentation</h4>

docs/index.html

@@ -26,7 +26,7 @@ <h3>Description</h3>
 pixel format conversion, image scaling and filtration, extraction of statistic information from images, motion detection,
 object detection and classification, neural network.</p>
 <p>The algorithms are optimized with using of different SIMD CPU extensions. 
-In particular the library supports following CPU extensions: SSE, AVX, AVX-512 and AMX for x86/x64, NEON, SVE, SVE2 for ARM, HVX for Hexagon.</p>
+In particular the library supports following CPU extensions: SSE, AVX, AVX-512 and AMX for x86/x64, NEON, SVE, SVE2 for ARM.</p>
 <p>The Simd Library has C API and also contains useful C++ and Python wrapper classes and functions to facilitate access to C API. 
 The library supports dynamic and static linking, 32-bit and 64-bit Windows and Linux, 
 MSVS, G++ and Clang compilers, MSVS project and CMake build systems.</p>

prj/cmake/CMakeLists.txt

@@ -185,8 +185,6 @@ else()
 		include(x86.cmake)
 	elseif((CMAKE_SYSTEM_PROCESSOR MATCHES "arm") OR (CMAKE_SYSTEM_PROCESSOR STREQUAL "aarch64"))
 		include(arm.cmake)
-	elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "hexagon")
-		include(hexagon.cmake)
 	else()
 	    message(FATAL_ERROR "Unknown value of CMAKE_SYSTEM_PROCESSOR!")
 	endif()

prj/txt/DoxygenOverview.txt

@@ -8,7 +8,7 @@
     extraction of statistic information from images, motion detection, neural network.
 
     The algorithms are optimized with using of different SIMD CPU extensions. 
-    In particular the library supports following CPU extensions: SSE, AVX, AVX-512 and AMX for x86/x64, NEON, SVE, SVE2 for ARM, HVX for Hexagon.
+    In particular the library supports following CPU extensions: SSE, AVX, AVX-512 and AMX for x86/x64, NEON, SVE, SVE2 for ARM.
 
     The %Simd Library has C API and also contains useful C++ classes and functions to facilitate access to C API.
     The library supports dynamic and static linking, 32-bit and 64-bit Windows and Linux, MSVS, G++ and Clang compilers, MSVS project and CMake build systems.
@@ -163,7 +163,7 @@
      - `-tr=2` a number of test execution repeats.
      - `-ts=1` to print statistics of time of tests execution.
      - `-cc=1` to check c++ API.
-     - `-de=2` a flags of SIMD extensions which testing are disabled. Base - 1, 2 - SSE4.1/NEON/HVX, 4 - AVX2/SVE2, 8 - AVX-512BW, 16 - AVX-512VNNI, 32 - AMX-BF16.
+     - `-de=2` a flags of SIMD extensions which testing are disabled. Base - 1, 2 - SSE4.1/NEON, 4 - AVX2/SVE2, 8 - AVX-512BW, 16 - AVX-512VNNI, 32 - AMX-BF16.
      - `-wu=100` a time to warm up CPU before testing (in milliseconds).
      - `-pt=1` a boolean flag to pin threads to cpu cores.
 */