[PATCH] D63558: AMDGPU: Undo sub x, c canonicalization for v2i16

[Matt Arsenault via Phabricator via llvm-commits]

arsenm marked an inline comment as done.
arsenm added inline comments.


================
Comment at: lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp:93
+    uint32_t K = Negate ?
+      (-LHSVal & 0xffff) | (-RHSVal << 16) :
+      (LHSVal & 0xffff) | (RHSVal << 16);
----------------
rampitec wrote:
> Don't you have warnings doing negation on an uint? Making it int may be cleaner.
There's no warning because it's the same operation. Making it int would be worse, because then it would involve multiple casts to get the unsigned shift


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D63558/new/

https://reviews.llvm.org/D63558