From 7791529dac64a3cc67a8dc982c04563021a69117 Mon Sep 17 00:00:00 2001 From: Evgeny Mankov Date: Wed, 3 Jan 2024 14:14:28 +0000 Subject: [PATCH] [HIPIFY][doc] CUDA `12.3.2` is the latest supported release + CUDA 12.3.2 is supported by LLVM >= 17.0.0 + Updated the `README.md` accordingly + Tested on Windows 10 and Ubuntu 21.10 --- docs/hipify-clang.md | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/docs/hipify-clang.md b/docs/hipify-clang.md index faccf2fe..fd74b3bd 100644 --- a/docs/hipify-clang.md +++ b/docs/hipify-clang.md @@ -23,7 +23,7 @@ After applying all the matchers, the output HIP source is produced. 1. [**LLVM+CLANG**](http://releases.llvm.org) of at least version [4.0.0](http://releases.llvm.org/download.html#4.0.0); the latest stable and recommended release: [**17.0.6**](https://github.com/llvm/llvm-project/releases/tag/llvmorg-17.0.6). -2. [**CUDA**](https://developer.nvidia.com/cuda-downloads) of at least version [8.0](https://developer.nvidia.com/cuda-80-ga2-download-archive), the latest supported version is [**12.3.1**](https://developer.nvidia.com/cuda-downloads). +2. [**CUDA**](https://developer.nvidia.com/cuda-downloads) of at least version [8.0](https://developer.nvidia.com/cuda-80-ga2-download-archive), the latest supported version is [**12.3.2**](https://developer.nvidia.com/cuda-downloads). @@ -180,12 +180,12 @@ After applying all the matchers, the output HIP source is produced. 17.0.4, 17.0.5, 17.0.6 - + - + @@ -416,9 +416,9 @@ Ubuntu 14: LLVM 4.0.0 - 7.1.0, CUDA 7.0 - 9.0, cuDNN 5.0.5 - 7.6.5 Ubuntu 16-18: LLVM 8.0.0 - 14.0.6, CUDA 8.0 - 10.2, cuDNN 5.1.10 - 8.0.5 -Ubuntu 20-21: LLVM 9.0.0 - 17.0.6, CUDA 8.0 - 12.3.1, cuDNN 5.1.10 - 8.9.7 +Ubuntu 20-21: LLVM 9.0.0 - 17.0.6, CUDA 8.0 - 12.3.2, cuDNN 5.1.10 - 8.9.7 -Ubuntu 22: LLVM 13.0.0 - 17.0.6, CUDA 10.0 - 12.3.1, cuDNN 8.0.5 - 8.9.7 +Ubuntu 22: LLVM 13.0.0 - 17.0.6, CUDA 10.0 - 12.3.2, cuDNN 8.0.5 - 8.9.7 Minimum build system requirements for the above configurations: @@ -487,7 +487,7 @@ make test-hipify ```shell Running HIPify regression tests ======================================== -CUDA 12.3.103 - will be used for testing +CUDA 12.3.107 - will be used for testing LLVM 17.0.6 - will be used for testing x86_64 - Platform architecture Linux 5.13.0-21-generic - Platform OS @@ -608,8 +608,8 @@ Testing Time: 7.90s | 14.0.0 - 14.0.6 | 7.0 - 11.7.1 | 8.0.5 - 8.4.1 | 2017.15.9.57*, 2019.16.11.17, 2022.17.2.6 | 3.24.0 | 3.10.6 | | 15.0.0 - 15.0.7 | 7.0 - 11.8.0 | 8.0.5 - 8.8.1 | 2019.16.11.25, 2022.17.5.2 | 3.26.0 | 3.11.2 | | 16.0.0 - 16.0.6 | 7.0 - 12.2.2 | 8.0.5 - 8.9.5 | 2019.16.11.29, 2022.17.7.1 | 3.27.3 | 3.11.4 | -| 17.0.1** 17.0.6 | 7.0 - 12.3.1 | 8.0.5 - 8.9.7 | 2019.16.11.32, 2022.17.8.3 | 3.28.1 | 3.12.1 | -| 18.0.0git | 7.0 - 12.3.1 | 8.0.5 - 8.9.7 | 2019.16.11.32, 2022.17.8.3 | 3.28.1 | 3.12.1 | +| 17.0.1** 17.0.6 | 7.0 - 12.3.2 | 8.0.5 - 8.9.7 | 2019.16.11.32, 2022.17.8.3 | 3.28.1 | 3.12.1 | +| 18.0.0git | 7.0 - 12.3.2 | 8.0.5 - 8.9.7 | 2019.16.11.32, 2022.17.8.3 | 3.28.1 | 3.12.1 | `*` LLVM 14.x.x is the latest major release supporting Visual Studio 2017. To build LLVM 14.x.x correctly by Visual Studio 2017, `-DLLVM_FORCE_USE_OLD_TOOLCHAIN=ON` should be added to a corresponding cmake command line.
12.3.112.3.2 LATEST STABLE CONFIG
18.0.0git12.3.112.3.2 + +