[all-commits] [llvm/llvm-project] 374f65: [OpenMP] Fix passing target id features to AMDGPU ...

[Joseph Huber via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 374f6554c3e409e4b9b5fd0ec90c5272768f5ab9
      https://github.com/llvm/llvm-project/commit/374f6554c3e409e4b9b5fd0ec90c5272768f5ab9
  Author: Joseph Huber <huberjn at outlook.com>
  Date:   2024-06-07 (Fri, 07 Jun 2024)

  Changed paths:
    M clang/lib/Driver/ToolChains/AMDGPU.cpp
    M clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp
    M clang/test/Driver/amdgpu-openmp-toolchain.c

  Log Message:
  -----------
  [OpenMP] Fix passing target id features to AMDGPU offloading (#94765)

Summary:
AMDGPU supports a `target-id` feature which is used to qualify targets
with different incompatible features. These are both rules and target
features. Currently, we pass `-target-cpu` twice when offloading to
OpenMP, and do not pass the target-id features at all. The effect was
that passing something like `--offload-arch=gfx90a:xnack+` would show up
as `-target-cpu=gfx90a:xnack+ -target-cpu=gfx90a`. Thus ignoring the
xnack completely and passing it twice. This patch fixes that to pass it
once and then separate it like how HIP does.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications