CMake/Help/release/dev/cuda-clang-separable-compilation.rst at 1d300ee2a9f51e08d4006191e2d50fd5771af202 - CMake - Gitea - ZeroTwo

mirror/CMake

mirror of https://github.com/Kitware/CMake.git synced 2026-05-04 13:19:51 -05:00

Files

T

Raul Tambre c63fe01835 CUDA: Clang separable compilation

For NVCC the compiler takes care of device linking when passed the "-dlink"
flag.
Clang doesn't support such magic and requires the buildsystem to do the work
that NVCC does behind the scenes.

The implementation is based on Bazel's device linking documentation:
https://github.com/tensorflow/tensorflow/blob/7cabcdf073abad8c46e9dda62bb8fa4682d2061e/third_party/nccl/build_defs.bzl.tpl#L259

Closes: #20726

2020-09-24 15:19:54 -04:00

5 lines

143 B

ReStructuredText

Raw Blame History

 cuda-clang-separable-compilation
 --------------------------------
 * :prop_tgt:`CUDA_SEPARABLE_COMPILATION` is now supported when using Clang.