AArch64: improve handling & modelling of FP_TO_XINT nodes.

There's probably no acatual change in behaviour here, just updating
the LowerFP_TO_INT function to be more similar to the reverse
implementation and updating costs to current CodeGen.

llvm-svn: 210985

2 files changed, 25 insertions, 10 deletions

diff --git a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
index fd9e69d3231..c2cf5b7f17f 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -1387,24 +1387,22 @@ static SDValue LowerVectorFP_TO_INT(SDValue Op, SelectionDAG &DAG) {
   EVT InVT = Op.getOperand(0).getValueType();
   EVT VT = Op.getValueType();
 
-  // FP_TO_XINT conversion from the same type are legal.
-  if (VT.getSizeInBits() == InVT.getSizeInBits())
-    return Op;
-
-  if (InVT == MVT::v2f64 || InVT == MVT::v4f32) {
+  if (VT.getSizeInBits() < InVT.getSizeInBits()) {
     SDLoc dl(Op);
     SDValue Cv =
         DAG.getNode(Op.getOpcode(), dl, InVT.changeVectorElementTypeToInteger(),
                     Op.getOperand(0));
     return DAG.getNode(ISD::TRUNCATE, dl, VT, Cv);
-  } else if (InVT == MVT::v2f32) {
+  }
+
+  if (VT.getSizeInBits() > InVT.getSizeInBits()) {
     SDLoc dl(Op);
     SDValue Ext = DAG.getNode(ISD::FP_EXTEND, dl, MVT::v2f64, Op.getOperand(0));
     return DAG.getNode(Op.getOpcode(), dl, VT, Ext);
   }
 
   // Type changing conversions are illegal.
-  return SDValue();
+  return Op;
 }
 
 SDValue AArch64TargetLowering::LowerFP_TO_INT(SDValue Op,
diff --git a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
index 4fae0a53b32..1dac14b96af 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
+++ b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
@@ -315,8 +315,10 @@ unsigned AArch64TTI::getCastInstrCost(unsigned Opcode, Type *Dst,
     // Complex: to v2f32
     { ISD::SINT_TO_FP, MVT::v2f32, MVT::v2i8,  3 },
     { ISD::SINT_TO_FP, MVT::v2f32, MVT::v2i16, 3 },
+    { ISD::SINT_TO_FP, MVT::v2f32, MVT::v2i64, 2 },
     { ISD::UINT_TO_FP, MVT::v2f32, MVT::v2i8,  3 },
     { ISD::UINT_TO_FP, MVT::v2f32, MVT::v2i16, 3 },
+    { ISD::UINT_TO_FP, MVT::v2f32, MVT::v2i64, 2 },
 
     // Complex: to v4f32
     { ISD::SINT_TO_FP, MVT::v4f32, MVT::v4i8,  4 },
@@ -341,12 +343,27 @@ unsigned AArch64TTI::getCastInstrCost(unsigned Opcode, Type *Dst,
     { ISD::FP_TO_UINT, MVT::v4i32, MVT::v4f32, 1 },
     { ISD::FP_TO_UINT, MVT::v2i64, MVT::v2f64, 1 },
 
-    { ISD::FP_TO_SINT, MVT::v2i32, MVT::v2f64, 2 },
+    // Complex, from v2f32: legal type is v2i32 (no cost) or v2i64 (1 ext).
     { ISD::FP_TO_SINT, MVT::v2i64, MVT::v2f32, 2 },
-    { ISD::FP_TO_SINT, MVT::v4i16, MVT::v4f32, 2 },
-    { ISD::FP_TO_UINT, MVT::v2i32, MVT::v2f64, 2 },
+    { ISD::FP_TO_SINT, MVT::v2i16, MVT::v2f32, 1 },
+    { ISD::FP_TO_SINT, MVT::v2i8,  MVT::v2f32, 1 },
     { ISD::FP_TO_UINT, MVT::v2i64, MVT::v2f32, 2 },
+    { ISD::FP_TO_UINT, MVT::v2i16, MVT::v2f32, 1 },
+    { ISD::FP_TO_UINT, MVT::v2i8,  MVT::v2f32, 1 },
+
+    // Complex, from v4f32: legal type is v4i16, 1 narrowing => ~2
+    { ISD::FP_TO_SINT, MVT::v4i16, MVT::v4f32, 2 },
+    { ISD::FP_TO_SINT, MVT::v4i8,  MVT::v4f32, 2 },
     { ISD::FP_TO_UINT, MVT::v4i16, MVT::v4f32, 2 },
+    { ISD::FP_TO_UINT, MVT::v4i8,  MVT::v4f32, 2 },
+
+    // Complex, from v2f64: legal type is v2i32, 1 narrowing => ~2.
+    { ISD::FP_TO_SINT, MVT::v2i32, MVT::v2f64, 2 },
+    { ISD::FP_TO_SINT, MVT::v2i16, MVT::v2f64, 2 },
+    { ISD::FP_TO_SINT, MVT::v2i8,  MVT::v2f64, 2 },
+    { ISD::FP_TO_UINT, MVT::v2i32, MVT::v2f64, 2 },
+    { ISD::FP_TO_UINT, MVT::v2i16, MVT::v2f64, 2 },
+    { ISD::FP_TO_UINT, MVT::v2i8,  MVT::v2f64, 2 },
   };
 
   int Idx = ConvertCostTableLookup<MVT>(