[PATCH] D55163: AMDGPU: Add optimization patterns to combine fp32->fp16 conversions

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.
Herald added a subscriber: kerbowa.


================
Comment at: lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp:2147
+  uint32_t Val;
+  if (getConstantValue(In, Val)) {
+    if (Hi && Val & 0xffff)
----------------
Constant folding during selection is pretty weird. Can we just do this in InstSimplify?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D55163/new/

https://reviews.llvm.org/D55163