https://github.com/arsenm commented:
amdgpu should directly use llvm.log* and llvm.exp* for the f16 and f32 cases, still relying on the libraries for f64 https://github.com/llvm/llvm-project/pull/128540 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits