Artem Belevich e4c3c6c873 Reverted the "[CUDA/OpenMP] Define only some host macros during device compilation" ...

The changes were breaking CUDA compilation.

Reverted revisions:
  r340681 D50845
   [CUDA/OpenMP] Define only some host macros during device compilation
  r340772 D51312
    [OpenMP][NVPTX] Use appropriate _CALL_ELF macro when offloading
  r340967 D51441
     Add predefined macro __gnu_linux__ for proper aux-triple

llvm-svn: 341115

2018-08-30 20:43:39 +00:00

50 KiB

Raw Blame History

View Raw