[PATCH] D81924: [PowerPC] Allow constrained FP intrinsics in mightUseCTR

[Qiu Chaofan via Phabricator via llvm-commits]

qiucf added inline comments.


================
Comment at: llvm/lib/Target/PowerPC/PPCTargetTransformInfo.cpp:389
+            break;
+          case Intrinsic::experimental_constrained_maxnum:
+            Opcode = ISD::STRICT_FMAXNUM;
----------------
steven.zhang wrote:
> So, how about the constrained fcmp/fcmps ?
It seems normal `fcmp` is also not handled here (when hard float is on).


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D81924/new/

https://reviews.llvm.org/D81924