Merge pull request #37 from AMD-Lightning-Internal/HIPIFY_rel_merge_20250721

[HIPIFY] Promote `amd-mainline` into `release/rocm-rel-7.0`

Author: vamovsik

CHANGELOG.md

@@ -3,6 +3,23 @@
 Documentation for HIPIFY is available at
 [https://rocmdocs.amd.com/projects/HIPIFY/en/latest/](https://rocmdocs.amd.com/projects/HIPIFY/en/latest/).
 
+## HIPIFY for ROCm 7.0.0
+
+### Added
+
+* CUDA 12.9.1 support
+* cuDNN 9.11.0 support
+* cuTENSOR 2.2.0.0 support
+* LLVM 20.1.8 support
+
+### Resolved issues
+
+* `hipDNN` support is removed by default
+* [#1859](https://github.com/ROCm/HIPIFY/issues/1859)[hipify-perl] Fix warnings on unsupported Driver or Runtime APIs which were erroneously not reported
+* [#1930](https://github.com/ROCm/HIPIFY/issues/1930) Revise `JIT API`
+* [#1962](https://github.com/ROCm/HIPIFY/issues/1962) Support for cuda-samples helper headers
+* [#2035](https://github.com/ROCm/HIPIFY/issues/2035) Remove `const_cast<const char**>` in `hiprtcCreateProgram` and `hiprtcCompileProgram`
+
 ## HIPIFY for ROCm 6.4.0
 
 ### Added

CMakeLists.txt

@@ -267,12 +267,6 @@ install(
     # Set RPATH for the binary.
     set_target_properties(hipify-clang PROPERTIES LINK_FLAGS "-Wl,--enable-new-dtags -Wl,--rpath,\\$ORIGIN/../lib" )
 
-    option(FILE_REORG_BACKWARD_COMPATIBILITY "Enable File Reorg with backward compatibility" OFF)
-
-    if(FILE_REORG_BACKWARD_COMPATIBILITY)
-      include(hipify-backward-compat.cmake)
-    endif()
-
     set(CPACK_PACKAGING_INSTALL_PREFIX "/opt/rocm" CACHE PATH "HIP Package Installation Path")
     set(BUILD_DIR ${CMAKE_CURRENT_BINARY_DIR}/packages/hipify-clang)
 

bin/hipify-perl

@@ -27,7 +27,7 @@ To ensure LLVM is found, or in case of multiple LLVM instances, specify the path
 
 .. code-block:: bash
 
-  -DCMAKE_PREFIX_PATH=/usr/llvm/20.1.6/dist
+  -DCMAKE_PREFIX_PATH=/usr/llvm/20.1.8/dist
 
 On Windows, specify the following option for CMake:
 ``-G "Visual Studio 17 2022"``
@@ -58,7 +58,7 @@ belongs to an appropriate version.
 LLVM >= 10.0.0
 -----------------
 
-1. Download `LLVM project <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.6>`_ sources.
+1. Download `LLVM project <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.8>`_ sources.
 
 2. Build `LLVM project <http://llvm.org/docs/CMake.html>`_:
 
@@ -148,7 +148,7 @@ LLVM >= 10.0.0
 
      .. code-block:: shell
 
-      -DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.10.2
+      -DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.11.0
 
 6. [Optional] Install `CUB 1.9.8 <https://github.com/NVIDIA/cub/releases/tag/1.9.8>`_ for ``CUDA < 11.0`` only;
    for ``CUDA >= 11.0``, the CUB shipped with CUDA will be used for testing.
@@ -177,13 +177,13 @@ LLVM >= 10.0.0
 
      .. code-block:: bash
 
-      python /usr/llvm/20.1.6/llvm-project/llvm/utils/lit/setup.py install
+      python /usr/llvm/20.1.8/llvm-project/llvm/utils/lit/setup.py install
       
      **Windows**:
 
      .. code-block:: shell
 
-      python D:/LLVM/20.1.6/llvm-project/llvm/utils/lit/setup.py install
+      python D:/LLVM/20.1.8/llvm-project/llvm/utils/lit/setup.py install
 
      In case of errors similar to ``ModuleNotFoundError: No module named 'setuptools'``, upgrade the ``setuptools`` package:
 
@@ -197,23 +197,23 @@ LLVM >= 10.0.0
 
      .. code-block:: bash
 
-      -DLLVM_EXTERNAL_LIT=/usr/llvm/20.1.6/build/bin/llvm-lit
+      -DLLVM_EXTERNAL_LIT=/usr/llvm/20.1.8/build/bin/llvm-lit
 
      **Windows**:
 
      .. code-block:: shell
 
-      -DLLVM_EXTERNAL_LIT=D:/LLVM/20.1.6/build/Release/bin/llvm-lit.py
+      -DLLVM_EXTERNAL_LIT=D:/LLVM/20.1.8/build/Release/bin/llvm-lit.py
 
    * ``FileCheck``:
 
      **Linux**:
 
-     Copy from ``/usr/llvm/20.1.6/build/bin/`` to ``CMAKE_INSTALL_PREFIX/dist/bin``.
+     Copy from ``/usr/llvm/20.1.8/build/bin/`` to ``CMAKE_INSTALL_PREFIX/dist/bin``.
 
      **Windows**:
 
-     Copy from ``D:/LLVM/20.1.6/build/Release/bin`` to ``CMAKE_INSTALL_PREFIX/dist/bin``.
+     Copy from ``D:/LLVM/20.1.8/build/Release/bin`` to ``CMAKE_INSTALL_PREFIX/dist/bin``.
 
      Alternatively, specify the path to ``FileCheck`` in the ``CMAKE_INSTALL_PREFIX`` option.
 
@@ -281,8 +281,8 @@ Linux testing
 
 On Linux, the following configurations are tested:
 
-* Ubuntu 22-23: LLVM 13.0.0 - 20.1.6, CUDA 7.0 - 12.8.1, cuDNN 8.0.5 - 9.10.2, cuTensor 1.0.1.0 - 2.2.0.0
-* Ubuntu 20-21: LLVM 9.0.0 - 20.1.6, CUDA 7.0 - 12.8.1, cuDNN 5.1.10 - 9.10.2, cuTensor 1.0.1.0 - 2.2.0.0
+* Ubuntu 22-23: LLVM 13.0.0 - 20.1.8, CUDA 7.0 - 12.8.1, cuDNN 8.0.5 - 9.11.0, cuTensor 1.0.1.0 - 2.2.0.0
+* Ubuntu 20-21: LLVM 9.0.0 - 20.1.8, CUDA 7.0 - 12.8.1, cuDNN 5.1.10 - 9.11.0, cuTensor 1.0.1.0 - 2.2.0.0
 * Ubuntu 16-19: LLVM 8.0.0 - 14.0.6, CUDA 7.0 - 10.2, cuDNN 5.1.10 - 8.0.5
 * Ubuntu 14: LLVM 4.0.0 - 7.1.0, CUDA 7.0 - 9.0, cuDNN 5.0.5 - 7.6.5
 
@@ -292,7 +292,7 @@ Minimum build system requirements for the above configurations:
 
 Recommended build system requirements:
 
-* CMake 4.0.2, GNU C/C++ 13.2, Python 3.13.3.
+* CMake 4.0.3, GNU C/C++ 13.2, Python 3.13.5.
 
 Here's how to build ``hipify-clang`` with testing support on ``Ubuntu 23.10.01``:
 
@@ -302,11 +302,11 @@ Here's how to build ``hipify-clang`` with testing support on ``Ubuntu 23.10.01``
   -DHIPIFY_CLANG_TESTS=ON \
   -DCMAKE_BUILD_TYPE=Release \
   -DCMAKE_INSTALL_PREFIX=../dist \
-  -DCMAKE_PREFIX_PATH=/usr/llvm/20.1.6/dist \
+  -DCMAKE_PREFIX_PATH=/usr/llvm/20.1.8/dist \
   -DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda-12.8.1 \
-  -DCUDA_DNN_ROOT_DIR=/usr/local/cudnn-9.10.2 \
+  -DCUDA_DNN_ROOT_DIR=/usr/local/cudnn-9.11.0 \
   -DCUDA_TENSOR_ROOT_DIR=/usr/local/cutensor-2.2.0.0 \
-  -DLLVM_EXTERNAL_LIT=/usr/llvm/20.1.6/build/bin/llvm-lit \
+  -DLLVM_EXTERNAL_LIT=/usr/llvm/20.1.8/build/bin/llvm-lit \
   ../hipify
 
 The corresponding successful output is:
@@ -331,20 +331,20 @@ The corresponding successful output is:
   --    - Is part of HIP SDK    : OFF
   --    - Install clang headers : ON
   -- Found ZLIB: /usr/lib/x86_64-linux-gnu/libz.so (found version "1.2.13")
-  -- Found LLVM 20.1.6:
-  --    - CMake module path     : /usr/llvm/20.1.6/dist/lib/cmake/llvm
-  --    - Clang include path    : /usr/llvm/20.1.6/dist/include
-  --    - LLVM Include path     : /usr/llvm/20.1.6/dist/include
-  --    - Binary path           : /usr/llvm/20.1.6/dist/bin
+  -- Found LLVM 20.1.8:
+  --    - CMake module path     : /usr/llvm/20.1.8/dist/lib/cmake/llvm
+  --    - Clang include path    : /usr/llvm/20.1.8/dist/include
+  --    - LLVM Include path     : /usr/llvm/20.1.8/dist/include
+  --    - Binary path           : /usr/llvm/20.1.8/dist/bin
   -- Linker detection: GNU ld
   -- ---- The below configuring for hipify-clang testing only ----
-  -- Found Python: /usr/bin/python3.13 (found suitable version "3.13.3", required range is "3.0...3.14") found components: Interpreter
+  -- Found Python: /usr/bin/python3.13 (found suitable version "3.13.5", required range is "3.0...3.14") found components: Interpreter
   -- Found lit: /usr/local/bin/lit
   -- Found FileCheck: /GIT/LLVM/trunk/dist/FileCheck
   -- Initial CUDA to configure:
   --    - CUDA Toolkit path     : /usr/local/cuda-12.8.1
   --    - CUDA Samples path     :
-  --    - cuDNN path            : /usr/local/cudnn-9.10.2
+  --    - cuDNN path            : /usr/local/cudnn-9.11.0
   --    - cuTENSOR path         : /usr/local/cuTensor/2.2.0.0
   --    - CUB path              :
   -- Found CUDAToolkit: /usr/local/cuda-12.8.1/targets/x86_64-linux/include (found version "12.8.93")
@@ -354,7 +354,7 @@ The corresponding successful output is:
   -- Found CUDA config:
   --    - CUDA Toolkit path     : /usr/local/cuda-12.8.1
   --    - CUDA Samples path     : OFF
-  --    - cuDNN path            : /usr/local/cudnn-9.10.2
+  --    - cuDNN path            : /usr/local/cudnn-9.11.0
   --    - CUB path              : /usr/local/cuda-12.8.1/include/cub
   --    - cuTENSOR path         : /usr/local/cuTensor/2.2.0.0
   -- Configuring done (0.6s)
@@ -372,11 +372,11 @@ The corresponding successful output is:
   Running HIPify regression tests
   ===============================================================
   CUDA 12.8.93 - will be used for testing
-  LLVM 20.1.6 - will be used for testing
+  LLVM 20.1.8 - will be used for testing
   x86_64 - Platform architecture
   Linux 6.5.0-15-generic - Platform OS
   64 - hipify-clang binary bitness
-  64 - python 3.13.3 binary bitness
+  64 - python 3.13.5 binary bitness
   ===============================================================
   -- Testing: 106 tests, 12 threads --
   Testing Time: 6.91s
@@ -468,16 +468,16 @@ Tested configurations:
     - ``3.11.4``
   * - ``17.0.1`` :sup:`6` - ``18.1.8`` :sup:`7`
     - ``7.0 - 12.3.2``
-    - ``8.0.5  - 9.10.2``
-    - ``2019.16.11.46, 2022.17.13.6``
-    - ``4.0.2``
-    - ``3.13.3``
-  * - ``19.1.0 - 20.1.6``
+    - ``8.0.5  - 9.11.0``
+    - ``2019.16.11.48, 2022.17.14.5``
+    - ``4.0.3``
+    - ``3.13.5``
+  * - ``19.1.0 - 20.1.8``
     - ``7.0 - 12.8.1``
-    - ``8.0.5  - 9.10.2``
-    - ``2019.16.11.46, 2022.17.13.6``
-    - ``4.0.2``
-    - ``3.13.3``
+    - ``8.0.5  - 9.11.0``
+    - ``2019.16.11.48, 2022.17.14.5``
+    - ``4.0.3``
+    - ``3.13.5``
 
 :sup:`5` LLVM 14.x.x is the latest major release supporting Visual Studio 2017.
 
@@ -502,12 +502,12 @@ Building with testing support using ``Visual Studio 17 2022`` on ``Windows 11``:
   -DHIPIFY_CLANG_TESTS=ON \
   -DCMAKE_BUILD_TYPE=Release \
   -DCMAKE_INSTALL_PREFIX=../dist \
-  -DCMAKE_PREFIX_PATH=D:/LLVM/20.1.6/dist \
+  -DCMAKE_PREFIX_PATH=D:/LLVM/20.1.8/dist \
   -DCUDA_TOOLKIT_ROOT_DIR="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.8" \
   -DCUDA_SDK_ROOT_DIR="C:/ProgramData/NVIDIA Corporation/CUDA Samples/v12.8" \
-  -DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.10.2 \
+  -DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.11.0 \
   -DCUDA_TENSOR_ROOT_DIR=D:/CUDA/cuTensor/2.2.0.0 \
-  -DLLVM_EXTERNAL_LIT=D:/LLVM/20.1.6/build/Release/bin/llvm-lit.py \
+  -DLLVM_EXTERNAL_LIT=D:/LLVM/20.1.8/build/Release/bin/llvm-lit.py \
   ../hipify
 
 The corresponding successful output is:
@@ -519,39 +519,39 @@ The corresponding successful output is:
   -- The CXX compiler identification is MSVC 19.42.34435.0
   -- Detecting C compiler ABI info
   -- Detecting C compiler ABI info - done
-  -- Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.42.34433/bin/Hostx64/x64/cl.exe - skipped
+  -- Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped
   -- Detecting C compile features
   -- Detecting C compile features - done
   -- Detecting CXX compiler ABI info
   -- Detecting CXX compiler ABI info - done
-  -- Check for working CXX compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.42.34433/bin/Hostx64/x64/cl.exe - skipped
+  -- Check for working CXX compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped
   -- Detecting CXX compile features
   -- Detecting CXX compile features - done
   -- HIPIFY config:
   --    - Build hipify-clang    : ON
   --    - Test hipify-clang     : ON
   --    - Is part of HIP SDK    : OFF
   --    - Install clang headers : ON
-  -- Found LLVM 20.1.6:
-  --    - CMake module path     : D:/LLVM/20.1.6/dist/lib/cmake/llvm
-  --    - Clang include path    : D:/LLVM/20.1.6/dist/include
-  --    - LLVM Include path     : D:/LLVM/20.1.6/dist/include
-  --    - Binary path           : D:/LLVM/20.1.6/dist/bin
+  -- Found LLVM 20.1.8:
+  --    - CMake module path     : D:/LLVM/20.1.8/dist/lib/cmake/llvm
+  --    - Clang include path    : D:/LLVM/20.1.8/dist/include
+  --    - LLVM Include path     : D:/LLVM/20.1.8/dist/include
+  --    - Binary path           : D:/LLVM/20.1.8/dist/bin
   -- ---- The below configuring for hipify-clang testing only ----
-  -- Found Python: C:/Users/TT/AppData/Local/Programs/Python/Python313/python.exe (found suitable version "3.13.3", required range is "3.0...3.14") found components: Interpreter
+  -- Found Python: C:/Users/TT/AppData/Local/Programs/Python/Python313/python.exe (found suitable version "3.13.5", required range is "3.0...3.14") found components: Interpreter
   -- Found lit: C:/Users/TT/AppData/Local/Programs/Python/Python313/Scripts/lit.exe
-  -- Found FileCheck: D:/LLVM/20.1.6/dist/bin/FileCheck.exe
+  -- Found FileCheck: D:/LLVM/20.1.8/dist/bin/FileCheck.exe
   -- Initial CUDA to configure:
   --    - CUDA Toolkit path     : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.8
   --    - CUDA Samples path     : C:/ProgramData/NVIDIA Corporation/CUDA Samples/v12.8
-  --    - cuDNN path            : D:/CUDA/cuDNN/9.10.2
+  --    - cuDNN path            : D:/CUDA/cuDNN/9.11.0
   --    - cuTENSOR path         : D:/CUDA/cuTensor/2.2.0.0
   --    - CUB path              :
   -- Found CUDAToolkit: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.8/include (found version "12.8.93")
   -- Found CUDA config:
   --    - CUDA Toolkit path     : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.8
   --    - CUDA Samples path     : C:/ProgramData/NVIDIA Corporation/CUDA Samples/v12.8
-  --    - cuDNN path            : D:/CUDA/cuDNN/9.10.2
+  --    - cuDNN path            : D:/CUDA/cuDNN/9.11.0
   --    - cuTENSOR path         : D:/CUDA/cuTensor/2.2.0.0
   --    - CUB path              : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.8/include/cub
   -- Configuring done (4.4s)

docs/building/building-hipify.rst

@@ -41,7 +41,7 @@ Release Dependencies
 
 * `LLVM+Clang <http://releases.llvm.org>`_ version is determined at least partially by 
   the CUDA version you are using, as shown in the table below. The recommended Clang release 
-  is the latest stable release `20.1.6 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.6>`_, 
+  is the latest stable release `20.1.8 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.8>`_, 
   or at least version `4.0.0 <http://releases.llvm.org/download.html#4.0.0>`_.
 
 .. list-table::
@@ -61,7 +61,9 @@ Release Dependencies
       `20.1.3 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.3>`_,
       `20.1.4 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.4>`_,
       `20.1.5 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.5>`_,
-      `20.1.6 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.6>`_:sup:`1`
+      `20.1.6 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.6>`_,
+      `20.1.7 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.7>`_,
+      `20.1.8 <https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.8>`_:sup:`1`
     - ✅
     - ✅
   * - `12.6.3 <https://developer.nvidia.com/cuda-12-6-3-download-archive>`_
@@ -249,7 +251,7 @@ Release Dependencies
 In most cases, you can get a suitable version of ``LLVM+Clang`` with your package manager. However, you can also
 `download a release archive <http://releases.llvm.org/>`_ and build or install it. In case of multiple versions of ``LLVM`` installed, set
 `CMAKE_PREFIX_PATH <https://cmake.org/cmake/help/latest/variable/CMAKE_PREFIX_PATH.html>`_ so that
-``CMake`` can find the desired version of ``LLVM``. For example, ``-DCMAKE_PREFIX_PATH=D:\LLVM\20.1.6\dist``.
+``CMake`` can find the desired version of ``LLVM``. For example, ``-DCMAKE_PREFIX_PATH=D:\LLVM\20.1.8\dist``.
 
 Usage
 =====
@@ -285,7 +287,7 @@ header files used during the hipification process:
 
 .. code:: shell
 
-  ./hipify-clang square.cu --cuda-path=/usr/local/cuda-12.8 --clang-resource-directory=/usr/llvm/20.1.6/dist/lib/clang/20
+  ./hipify-clang square.cu --cuda-path=/usr/local/cuda-12.8 --clang-resource-directory=/usr/llvm/20.1.8/dist/lib/clang/20
 
 For more information, refer to the `Clang manual for compiling CUDA <https://llvm.org/docs/CompileCudaWithLLVM.html#compiling-cuda-code>`_.
 

docs/how-to/hipify-clang.rst

@@ -19,7 +19,7 @@
 | CUSOLVER API     | [HIP SOLVER API](tables/CUSOLVER_API_supported_by_HIP.md)         |                                                           |                                                                         |
 | CURAND API       | [HIP RAND API](tables/CURAND_API_supported_by_HIP.md)             | [ROC RAND API](tables/CURAND_API_supported_by_ROC.md)     | [HIP + ROC RAND API](tables/CURAND_API_supported_by_HIP_and_ROC.md)     |
 | CUFFT API        | [HIP FFT API](tables/CUFFT_API_supported_by_HIP.md)               |                                                           |                                                                         |
-| CUDNN API        | [HIP DNN API](tables/CUDNN_API_supported_by_HIP.md)               | [MIOPEN API](tables/CUDNN_API_supported_by_MIOPEN.md)     | [HIP + MIOPEN API](tables/CUDNN_API_supported_by_HIP_and_MIOPEN.md)     |
+| CUDNN API        |                                                                   | [MIOPEN API](tables/CUDNN_API_supported_by_MIOPEN.md)     |                                                                         |
 | CUTENSOR API     | [HIP TENSOR API](tables/CUTENSOR_API_supported_by_HIP.md)         |                                                           |                                                                         |
 | CUB API          | [HIP CUB API](tables/CUB_API_supported_by_HIP.md)                 |                                                           |                                                                         |