[llvm] [InstCombine] Extend `foldICmpBinOp` to `add`-like `or`. (PR #71396)

[Nikita Popov via llvm-commits]

================
@@ -4624,27 +4624,35 @@ Instruction *InstCombinerImpl::foldICmpBinOp(ICmpInst &I,
   }
 
   bool NoOp0WrapProblem = false, NoOp1WrapProblem = false;
-  if (BO0 && isa<OverflowingBinaryOperator>(BO0))
-    NoOp0WrapProblem =
-        ICmpInst::isEquality(Pred) ||
-        (CmpInst::isUnsigned(Pred) && BO0->hasNoUnsignedWrap()) ||
-        (CmpInst::isSigned(Pred) && BO0->hasNoSignedWrap());
-  if (BO1 && isa<OverflowingBinaryOperator>(BO1))
-    NoOp1WrapProblem =
-        ICmpInst::isEquality(Pred) ||
-        (CmpInst::isUnsigned(Pred) && BO1->hasNoUnsignedWrap()) ||
-        (CmpInst::isSigned(Pred) && BO1->hasNoSignedWrap());
-
+  bool Op0HasNUW = false, Op1HasNUW = false;
+  bool Op0HasNSW = false, Op1HasNSW = false;
   // Analyze the case when either Op0 or Op1 is an add instruction.
   // Op0 = A + B (or A and B are null); Op1 = C + D (or C and D are null).
----------------
nikic wrote:

This comment should be below the lambda (above the `Value *A = nullptr, *B = nullptr, *C = nullptr, *D = nullptr;` line).

https://github.com/llvm/llvm-project/pull/71396