================ @@ -0,0 +1,32 @@ +// RUN: %clang_cc1 -triple amdgpu-amd-amdhsa -fsyntax-only -verify -x c %s +// RUN: %clang_cc1 -triple amdgpu-amd-amdhsa -fsyntax-only -verify -x c++ %s +// RUN: %clang_cc1 -triple amdgpu-amd-amdhsa -fcuda-is-device -x hip -fsyntax-only -verify %s +// RUN: %clang_cc1 -triple amdgpu-amd-amdhsa -cl-std=CL1.2 -fsyntax-only -verify -x cl %s +// RUN: %clang_cc1 -triple amdgpu-amd-amdhsa -cl-std=CL2.0 -fsyntax-only -verify -x cl %s +// RUN: %clang_cc1 -triple amdgpu-amd-amdhsa -cl-std=CL3.0 -fsyntax-only -verify -x cl %s + +// expected-no-diagnostics + +#define AS(N) __attribute__((address_space(N))) +#define OCL(Name) __attribute__((opencl_##Name)) ---------------- arsenm wrote:
I think these macros and the variable names just make this unnecessarily difficult to read https://github.com/llvm/llvm-project/pull/230677 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
