[all-commits] [llvm/llvm-project] 76bfbb: [libomptarget][amdgpu] Call into deviceRTL instead...

[Jon Chesterfield via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 76bfbb74d38b611f150e8e1a4becc11be95703da
      https://github.com/llvm/llvm-project/commit/76bfbb74d38b611f150e8e1a4becc11be95703da
  Author: Jon Chesterfield <jonathanchesterfield at gmail.com>
  Date:   2021-01-04 (Mon, 04 Jan 2021)

  Changed paths:
    M clang/lib/CodeGen/CGOpenMPRuntimeAMDGCN.cpp
    M clang/test/OpenMP/amdgcn_target_codegen.cpp
    M openmp/libomptarget/deviceRTLs/amdgcn/src/amdgcn_interface.h
    M openmp/libomptarget/deviceRTLs/amdgcn/src/target_impl.hip

  Log Message:
  -----------
  [libomptarget][amdgpu] Call into deviceRTL instead of ockl

[libomptarget][amdgpu] Call into deviceRTL instead of ockl

Amdgpu codegen presently emits a call into ockl. The same functionality
is already present in the deviceRTL. Adds an amdgpu specific entry point
to avoid the dependency. This lets simple openmp code (specifically, that
which doesn't use libm) run without rocm device libraries installed.

Reviewed By: ronlieb

Differential Revision: https://reviews.llvm.org/D93356