================
@@ -337,9 +337,12 @@ void AMDGPUTargetInfo::getTargetDefines(const LangOptions
&Opts,
if (hasFastFMA())
Builder.defineMacro("FP_FAST_FMA");
- Builder.defineMacro("__AMDGCN_WAVEFRONT_SIZE__", Twine(WavefrontSize));
- // ToDo: deprecate this macro for naming consistency.
- Builder.defineMacro("__AMDGCN_WAVEFRONT_SIZE", Twine(WavefrontSize));
+ Builder.defineMacro("__AMDGCN_WAVEFRONT_SIZE__", Twine(WavefrontSize),
+ "compile-time-constant access to the wavefront size will
"
+ "be removed in a future release");
+ Builder.defineMacro("__AMDGCN_WAVEFRONT_SIZE", Twine(WavefrontSize),
+ "compile-time-constant access to the wavefront size will
"
+ "be removed in a future release");
----------------
jhuber6 wrote:
This never should've been defined on the host side for sure. We do this in the
OpenMP device runtime.
```c
return llvm::omp::getAMDGPUGridValues<__AMDGCN_WAVEFRONT_SIZE>();
```
That could probably be replaced with
```c
return __builtin_amdgcn_wavefrontsize();
```
https://github.com/llvm/llvm-project/pull/112849
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
