[llvm] [LoopVectorize][X86][AMDLibm] Add Missing AMD LibM trig vector intrinsics (PR #101125)
Farzon Lotfi via llvm-commits
llvm-commits at lists.llvm.org
Mon Jul 29 21:48:37 PDT 2024
================
@@ -1371,7 +1371,9 @@ TLI_DEFINE_VECFUNC("llvm.asin.f32", "amd_vrs16_asinf", FIXED(16), NOMASK, "_ZGV_
TLI_DEFINE_VECFUNC("acosf", "amd_vrs4_acosf", FIXED(4), NOMASK, "_ZGV_LLVM_N4v")
TLI_DEFINE_VECFUNC("acosf", "amd_vrs8_acosf", FIXED(8), NOMASK, "_ZGV_LLVM_N8v")
+TLI_DEFINE_VECFUNC("acosf", "amd_vrs16_acosf", FIXED(16), NOMASK, "_ZGV_LLVM_N16v")
----------------
farzonl wrote:
@rohitaggarwal007 git blame shows you added the amd vector lib support: https://github.com/llvm/llvm-project/pull/78560. Can you review this change?
https://github.com/llvm/llvm-project/pull/101125
More information about the llvm-commits
mailing list