[llvm] r325827 - [X86] Make the subus special case in LowerVSETCC self contained

[Craig Topper via llvm-commits]

Author: ctopper
Date: Thu Feb 22 12:24:18 2018
New Revision: 325827

URL: http://llvm.org/viewvc/llvm-project?rev=325827&view=rev
Log:
[X86] Make the subus special case in LowerVSETCC self contained

Previously this code overrode the flags and opcode used by the later code in LowerVSETCC. This makes the code difficult to read and follow.

This patch moves all the SUBUS code into its own function and makes it responsible for creating its own SDNodes on success.

Differential Revision: https://reviews.llvm.org/D43530

Modified:
    llvm/trunk/lib/Target/X86/X86ISelLowering.cpp

Modified: llvm/trunk/lib/Target/X86/X86ISelLowering.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ISelLowering.cpp?rev=325827&r1=325826&r2=325827&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ISelLowering.cpp (original)
+++ llvm/trunk/lib/Target/X86/X86ISelLowering.cpp Thu Feb 22 12:24:18 2018
@@ -17896,6 +17896,51 @@ static SDValue ChangeVSETULTtoVSETULE(co
   return DAG.getBuildVector(VT, dl, ULTOp1);
 }
 
+/// As another special case, use PSUBUS[BW] when it's profitable. E.g. for
+/// Op0 u<= Op1:
+///   t = psubus Op0, Op1
+///   pcmpeq t, <0..0>
+static SDValue LowerVSETCCWithSUBUS(SDValue Op0, SDValue Op1, MVT VT,
+                                    ISD::CondCode Cond, const SDLoc &dl,
+                                    const X86Subtarget &Subtarget,
+                                    SelectionDAG &DAG) {
+  if (!Subtarget.hasSSE2())
+    return SDValue();
+
+  MVT VET = VT.getVectorElementType();
+  if (VET != MVT::i8 && VET != MVT::i16)
+    return SDValue();
+
+  switch (Cond) {
+  default:
+    return SDValue();
+  case ISD::SETULT: {
+    // If the comparison is against a constant we can turn this into a
+    // setule.  With psubus, setule does not require a swap.  This is
+    // beneficial because the constant in the register is no longer
+    // destructed as the destination so it can be hoisted out of a loop.
+    // Only do this pre-AVX since vpcmp* is no longer destructive.
+    if (Subtarget.hasAVX())
+      return SDValue();
+    SDValue ULEOp1 = ChangeVSETULTtoVSETULE(dl, Op1, DAG);
+    if (!ULEOp1)
+      return SDValue();
+    Op1 = ULEOp1;
+    break;
+  }
+  // Psubus is better than flip-sign because it requires no inversion.
+  case ISD::SETUGE:
+    std::swap(Op0, Op1);
+    break;
+  case ISD::SETULE:
+    break;
+  }
+
+  SDValue Result = DAG.getNode(X86ISD::SUBUS, dl, VT, Op0, Op1);
+  return DAG.getNode(X86ISD::PCMPEQ, dl, VT, Result,
+                     getZeroVector(VT, Subtarget, DAG, dl));
+}
+
 static SDValue LowerVSETCC(SDValue Op, const X86Subtarget &Subtarget,
                            SelectionDAG &DAG) {
   SDValue Op0 = Op.getOperand(0);
@@ -18088,6 +18133,10 @@ static SDValue LowerVSETCC(SDValue Op, c
     return Result;
   }
 
+  // Try to use SUBUS and PCMPEQ.
+  if (SDValue V = LowerVSETCCWithSUBUS(Op0, Op1, VT, Cond, dl, Subtarget, DAG))
+    return V;
+
   // We are handling one of the integer comparisons here. Since SSE only has
   // GT and EQ comparisons for integer, swapping operands and multiple
   // operations may be required for some comparisons.
@@ -18098,41 +18147,6 @@ static SDValue LowerVSETCC(SDValue Op, c
   bool Invert = Cond == ISD::SETNE ||
                 (Cond != ISD::SETEQ && ISD::isTrueWhenEqual(Cond));
 
-  MVT VET = VT.getVectorElementType();
-  bool HasSubus = Subtarget.hasSSE2() && (VET == MVT::i8 || VET == MVT::i16);
-  bool Subus = false;
-  if (HasSubus) {
-    // As another special case, use PSUBUS[BW] when it's profitable. E.g. for
-    // Op0 u<= Op1:
-    //   t = psubus Op0, Op1
-    //   pcmpeq t, <0..0>
-    switch (Cond) {
-    default: break;
-    case ISD::SETULT: {
-      // If the comparison is against a constant we can turn this into a
-      // setule.  With psubus, setule does not require a swap.  This is
-      // beneficial because the constant in the register is no longer
-      // destructed as the destination so it can be hoisted out of a loop.
-      // Only do this pre-AVX since vpcmp* is no longer destructive.
-      if (Subtarget.hasAVX())
-        break;
-      if (SDValue ULEOp1 = ChangeVSETULTtoVSETULE(dl, Op1, DAG)) {
-        Op1 = ULEOp1;
-        Subus = true; Invert = false; Swap = false;
-      }
-      break;
-    }
-    // Psubus is better than flip-sign because it requires no inversion.
-    case ISD::SETUGE: Subus = true; Invert = false; Swap = true;  break;
-    case ISD::SETULE: Subus = true; Invert = false; Swap = false; break;
-    }
-
-    if (Subus) {
-      Opc = X86ISD::SUBUS;
-      FlipSigns = false;
-    }
-  }
-
   if (Swap)
     std::swap(Op0, Op1);
 
@@ -18220,10 +18234,6 @@ static SDValue LowerVSETCC(SDValue Op, c
   if (Invert)
     Result = DAG.getNOT(dl, Result, VT);
 
-  if (Subus)
-    Result = DAG.getNode(X86ISD::PCMPEQ, dl, VT, Result,
-                         getZeroVector(VT, Subtarget, DAG, dl));
-
   return Result;
 }