90d3b0fb18
* Updates for 3.2.1 release. * Minor fix in gemm op profiler for raster order. * Add scheduler mapping for raster order in the kernels.
1.6 KiB
README > CUTLASS 3: Building with Clang as host compiler
Building with Clang as host compiler
CUTLASS 3.2(.1) reintroduces support for building with Clang as host compiler, and NVCC as device compiler. This is NOT the same as building with Clang as both host and device compiler ("CUDA Clang").
Software prerequisites
-
Clang (tested with Clang 14)
-
CUDA Toolkit (tested with 12.2; other versions likely work)
-
CMake (at least 3.18)
-
git
-
Python (at least 3.6)
Experience with Ubuntu 22.04 LTS is that clang requires the following packages to be installed.
$ sudo apt-get install clang cmake ninja-build pkg-config libgtk-3-dev liblzma-dev libstdc++-12-dev
A symptom of not installing all needed dependencies
is the following error when attempting to use clang:
"/usr/bin/ld: cannot find -lstdc++: No such file or directory".
Running CMake
The Clang build requires specifying the following three CMake options.
-
CMAKE_CXX_COMPILER=clang++ -
CMAKE_CUDA_HOST_COMPILER=clang++ -
CMAKE_C_COMPILER=clang
This assumes that clang++ and clang are in the user's PATH.
Please note that both CMAKE_CXX_COMPILER and CMAKE_C_COMPILER
must be set, even though CUTLASS is a C++ project, not a C project.
Users can also specify a particular CUDA Toolkit version
by setting the CMake option CMAKE_CUDA_COMPILER
to the path to the nvcc executable
that lives in the CUDA Toolkit's directory. For example,
if ${PATH_TO_CUDA_TOOLKIT} is the CUDA Toolkit directory,
then one can set CMAKE_CUDA_COMPILER as follows.
CMAKE_CUDA_COMPILER=${PATH_TO_CUDA_TOOLKIT}/bin/nvcc