[llvm] [X86] Fold generic ADD/SUB with constants to X86ISD::SUB/ADD (PR #164316)
Phoebe Wang via llvm-commits
llvm-commits at lists.llvm.org
Wed Oct 22 19:20:25 PDT 2025
================
@@ -57631,8 +57631,25 @@ static SDValue combineX86AddSub(SDNode *N, SelectionDAG &DAG,
DCI.CombineTo(GenericAddSub, Op);
}
};
- MatchGeneric(LHS, RHS, false);
- MatchGeneric(RHS, LHS, X86ISD::SUB == N->getOpcode());
+ MatchGeneric(GenericOpc, LHS, RHS, false);
+ MatchGeneric(GenericOpc, RHS, LHS, X86ISD::SUB == N->getOpcode());
+
+ if (ConstantSDNode *Const = dyn_cast<ConstantSDNode>(RHS)) {
+ SDValue NegC = DAG.getConstant(-Const->getAPIntValue(), DL, VT);
+ if (X86ISD::SUB == N->getOpcode()) {
+ // With LHS - C, fold LHS + (-C)
+ MatchGeneric(ISD::ADD, LHS, NegC, false);
+ } else {
+ // With -(LHS + C), fold (-C) - LHS
----------------
phoebewang wrote:
I think the comment is missleading. We are folding `LHS + C` to `-((-C) - LHS)`. We are creating a new Negate instead of eliminating.
https://github.com/llvm/llvm-project/pull/164316
More information about the llvm-commits
mailing list