CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-01 09:01:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

HIP: fix AMDGPU_TARGETS, update documentation (#16803)

Browse Source
This commit is contained in:
Johannes Gäßler
2025-10-27 21:39:49 +01:00
committed by GitHub
parent c55d53acec
commit 80d28f104c
2 changed files with 9 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
docs/build.md
View File

@@ -261,10 +261,12 @@ You can download it from your Linux distro's package manager or from here: [ROCm
- Using `CMake` for Linux (assuming a gfx1030-compatible AMD GPU): - Using `CMake` for Linux (assuming a gfx1030-compatible AMD GPU):
```bash ```bash
HIPCXX="$(hipconfig -l)/clang" HIP_PATH="$(hipconfig -R)" \ HIPCXX="$(hipconfig -l)/clang" HIP_PATH="$(hipconfig -R)" \
cmake -S . -B build -DGGML_HIP=ON -DAMDGPU_TARGETS=gfx1030 -DCMAKE_BUILD_TYPE=Release \ cmake -S . -B build -DGGML_HIP=ON -DGPU_TARGETS=gfx1030 -DCMAKE_BUILD_TYPE=Release \
&& cmake --build build --config Release -- -j 16 && cmake --build build --config Release -- -j 16
``` ```
Note: `GPU_TARGETS` is optional, omitting it will build the code for all GPUs in the current system.
To enhance flash attention performance on RDNA3+ or CDNA architectures, you can utilize the rocWMMA library by enabling the `-DGGML_HIP_ROCWMMA_FATTN=ON` option. This requires rocWMMA headers to be installed on the build system. To enhance flash attention performance on RDNA3+ or CDNA architectures, you can utilize the rocWMMA library by enabling the `-DGGML_HIP_ROCWMMA_FATTN=ON` option. This requires rocWMMA headers to be installed on the build system.
The rocWMMA library is included by default when installing the ROCm SDK using the `rocm` meta package provided by AMD. Alternatively, if you are not using the meta package, you can install the library using the `rocwmma-dev` or `rocwmma-devel` package, depending on your system's package manager. The rocWMMA library is included by default when installing the ROCm SDK using the `rocm` meta package provided by AMD. Alternatively, if you are not using the meta package, you can install the library using the `rocwmma-dev` or `rocwmma-devel` package, depending on your system's package manager.
@@ -282,17 +284,17 @@ You can download it from your Linux distro's package manager or from here: [ROCm
```bash ```bash
HIPCXX="$(hipconfig -l)/clang" HIP_PATH="$(hipconfig -p)" \ HIPCXX="$(hipconfig -l)/clang" HIP_PATH="$(hipconfig -p)" \
HIP_DEVICE_LIB_PATH=<directory-you-just-found> \ HIP_DEVICE_LIB_PATH=<directory-you-just-found> \
cmake -S . -B build -DGGML_HIP=ON -DAMDGPU_TARGETS=gfx1030 -DCMAKE_BUILD_TYPE=Release \ cmake -S . -B build -DGGML_HIP=ON -DGPU_TARGETS=gfx1030 -DCMAKE_BUILD_TYPE=Release \
&& cmake --build build -- -j 16 && cmake --build build -- -j 16
``` ```
- Using `CMake` for Windows (using x64 Native Tools Command Prompt for VS, and assuming a gfx1100-compatible AMD GPU): - Using `CMake` for Windows (using x64 Native Tools Command Prompt for VS, and assuming a gfx1100-compatible AMD GPU):
```bash ```bash
set PATH=%HIP_PATH%\bin;%PATH% set PATH=%HIP_PATH%\bin;%PATH%
cmake -S . -B build -G Ninja -DAMDGPU_TARGETS=gfx1100 -DGGML_HIP=ON -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_BUILD_TYPE=Release cmake -S . -B build -G Ninja -DGPU_TARGETS=gfx1100 -DGGML_HIP=ON -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_BUILD_TYPE=Release
cmake --build build cmake --build build
``` ```
Make sure that `AMDGPU_TARGETS` is set to the GPU arch you want to compile for. The above example uses `gfx1100` that corresponds to Radeon RX 7900XTX/XT/GRE. You can find a list of targets [here](https://llvm.org/docs/AMDGPUUsage.html#processors) If necessary, adapt `GPU_TARGETS` to the GPU arch you want to compile for. The above example uses `gfx1100` that corresponds to Radeon RX 7900XTX/XT/GRE. You can find a list of targets [here](https://llvm.org/docs/AMDGPUUsage.html#processors)
Find your gpu version string by matching the most significant version information from `rocminfo | grep gfx | head -1 | awk '{print $2}'` with the list of processors, e.g. `gfx1035` maps to `gfx1030`. Find your gpu version string by matching the most significant version information from `rocminfo | grep gfx | head -1 | awk '{print $2}'` with the list of processors, e.g. `gfx1035` maps to `gfx1030`.

5
ggml/src/ggml-hip/CMakeLists.txt
View File

@@ -29,10 +29,11 @@ if (CXX_IS_HIPCC)
endif() endif()
else() else()
# Forward (AMD)GPU_TARGETS to CMAKE_HIP_ARCHITECTURES. # Forward (AMD)GPU_TARGETS to CMAKE_HIP_ARCHITECTURES.
if(AMDGPU_TARGETS AND NOT GPU_TARGETS)
set(GPU_TARGETS ${AMDGPU_TARGETS})
endif()
if(GPU_TARGETS AND NOT CMAKE_HIP_ARCHITECTURES) if(GPU_TARGETS AND NOT CMAKE_HIP_ARCHITECTURES)
set(CMAKE_HIP_ARCHITECTURES ${GPU_TARGETS}) set(CMAKE_HIP_ARCHITECTURES ${GPU_TARGETS})
elseif(AMDGPU_TARGETS AND NOT CMAKE_HIP_ARCHITECTURES)
set(CMAKE_HIP_ARCHITECTURES ${AMDGPU_TARGETS})
endif() endif()
cmake_minimum_required(VERSION 3.21) cmake_minimum_required(VERSION 3.21)
enable_language(HIP) enable_language(HIP)