[PATCH] D142756: AMDGPU: Try to select fneg modifier from xor

[Pierre van Houtryve via Phabricator via llvm-commits]

Pierre-vh added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp:2602
+  } else {
+    Src = stripBitcast(Src);
+    if (Src.getOpcode() == ISD::XOR) {
----------------
Can we just do the stripBitcast call earlier? (at `Src = In`)
That way it simplifies the condition a bit (you can have a if/else if)



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp:2603
+    Src = stripBitcast(Src);
+    if (Src.getOpcode() == ISD::XOR) {
+      // Match an integer fneg
----------------
Can you add a comment explaining why we need this special-case (like in the commit message) ?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D142756/new/

https://reviews.llvm.org/D142756