[PATCH] D113276: [DAG] FoldConstantVectorArithmetic - remove SDNodeFlags argument

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon created this revision.
RKSimon added reviewers: lebedev.ri, spatel, efriedma, craig.topper.
Herald added subscribers: ecnelises, hiraditya.
RKSimon requested review of this revision.
Herald added a project: LLVM.

Another minor step towards merging FoldConstantVectorArithmetic into FoldConstantArithmetic.

We don't use SDNodeFlags in any constant folding inside DAG, so passing the Flags is a waste of time - an alternative would be to wire up FoldConstantArithmetic to take SDNodeFlags just-in-case we someday start using it, but we don't have any way to test it and I'd prefer to avoid dead code.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D113276

Files:
  llvm/include/llvm/CodeGen/SelectionDAG.h
  llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
  llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp


Index: llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
===================================================================
--- llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
+++ llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
@@ -5401,8 +5401,7 @@
 // TODO: Merge with FoldConstantArithmetic
 SDValue SelectionDAG::FoldConstantVectorArithmetic(unsigned Opcode,
                                                    const SDLoc &DL, EVT VT,
-                                                   ArrayRef<SDValue> Ops,
-                                                   const SDNodeFlags Flags) {
+                                                   ArrayRef<SDValue> Ops) {
   // If the opcode is a target-specific ISD node, there's nothing we can
   // do here and the operand rules may not line up with the below, so
   // bail early.
@@ -5486,7 +5485,7 @@
     }
 
     // Constant fold the scalar operands.
-    SDValue ScalarResult = getNode(Opcode, DL, SVT, ScalarOps, Flags);
+    SDValue ScalarResult = getNode(Opcode, DL, SVT, ScalarOps);
 
     // Legalize the (integer) scalar constant if necessary.
     if (LegalSVT != SVT)
Index: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
===================================================================
--- llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -22270,8 +22270,8 @@
   SDNodeFlags Flags = N->getFlags();
 
   // See if we can constant fold the vector operation.
-  if (SDValue Fold = DAG.FoldConstantVectorArithmetic(
-          Opcode, SDLoc(LHS), LHS.getValueType(), Ops, N->getFlags()))
+  if (SDValue Fold = DAG.FoldConstantVectorArithmetic(Opcode, SDLoc(LHS),
+                                                      LHS.getValueType(), Ops))
     return Fold;
 
   // Move unary shuffles with identical masks after a vector binop:
Index: llvm/include/llvm/CodeGen/SelectionDAG.h
===================================================================
--- llvm/include/llvm/CodeGen/SelectionDAG.h
+++ llvm/include/llvm/CodeGen/SelectionDAG.h
@@ -1733,8 +1733,7 @@
                                  ArrayRef<SDValue> Ops);
 
   SDValue FoldConstantVectorArithmetic(unsigned Opcode, const SDLoc &DL, EVT VT,
-                                       ArrayRef<SDValue> Ops,
-                                       const SDNodeFlags Flags = SDNodeFlags());
+                                       ArrayRef<SDValue> Ops);
 
   /// Fold floating-point operations with 2 operands when both operands are
   /// constants and/or undefined.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D113276.385053.patch
Type: text/x-patch
Size: 2537 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20211105/326e7635/attachment.bin>