[PATCH] D88929: [OpenMP] Change CMake Configuration to Build for Highest CUDA Architecture by Default

[Joseph Huber via Phabricator via cfe-commits]

jhuber6 added a comment.

In D88929#2315402 <https://reviews.llvm.org/D88929#2315402>, @ye-luo wrote:

> The link I posted indicated that independent feature is merged since 3.12. Better to avoid deprecated stuff when introducing new cmake lines even though some existing lines may still rely on deprecated cmake.

This requires adding CUDA as a language, is that alright inside of Clang? Nothing is using CUDA, we're just checking the architecture.

In D88929#2315404 <https://reviews.llvm.org/D88929#2315404>, @ye-luo wrote:

> 3.18 introduces CMAKE_CUDA_ARCHITECTURES. Does 3.18 supports detection? If we know a new way works since 3.18, I think putting both with if-else makes sense.
>
> Regarding doing it manually. I think it is more risky than using existing schemes shipped with cmake.

I'm not sure how that would affect the other Libomptarget stuff that uses FindCUDA since language enabling is done at the highest level, does that conflict?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D88929/new/

https://reviews.llvm.org/D88929