CostModel: Add parameter to instruction cost to further classify operand values

On certain architectures we can support efficient vectorized version of
instructions if the operand value is uniform (splat) or a constant scalar.
An example of this is a vector shift on x86.

We can efficiently support

for (i = 0 ; i < ; i += 4)
  w[0:3] = v[0:3] << <2, 2, 2, 2>

but not

for (i = 0; i < ; i += 4)
  w[0:3] = v[0:3] << x[0:3]

This patch adds a parameter to getArithmeticInstrCost to further qualify operand
values as uniform or uniform constant.

Targets can then choose to return a different cost for instructions with such
operand values.

A follow-up commit will test this feature on x86.

radar://13576547

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@178807 91177308-0d34-0410-b5e6-96231b3b80d8

6 files changed, 61 insertions, 13 deletions

diff --git a/include/llvm/Analysis/TargetTransformInfo.h b/include/llvm/Analysis/TargetTransformInfo.h
index ccab19495f..a9d6725d86 100644
--- a/include/llvm/Analysis/TargetTransformInfo.h
+++ b/include/llvm/Analysis/TargetTransformInfo.h
@@ -263,6 +263,13 @@ public:
     SK_ExtractSubvector ///< ExtractSubvector Index indicates start offset.
   };
 
+  /// \brief Additonal information about an operand's possible values.
+  enum OperandValueKind {
+    OK_AnyValue,            // Operand can have any value.
+    OK_UniformValue,        // Operand is uniform (splat of a value).
+    OK_UniformConstantValue // Operand is uniform constant.
+  };
+
   /// \return The number of scalar or vector registers that the target has.
   /// If 'Vectors' is true, it returns the number of vector registers. If it is
   /// set to false, it returns the number of scalar registers.
@@ -277,7 +284,9 @@ public:
   virtual unsigned getMaximumUnrollFactor() const;
 
   /// \return The expected cost of arithmetic ops, such as mul, xor, fsub, etc.
-  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty) const;
+  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                  OperandValueKind Opd1Info = OK_AnyValue,
+                                  OperandValueKind Opd2Info = OK_AnyValue) const;
 
   /// \return The cost of a shuffle instruction of kind Kind and of type Tp.
   /// The index and subtype parameters are used by the subvector insertion and
diff --git a/lib/Analysis/CostModel.cpp b/lib/Analysis/CostModel.cpp
index df70d157b4..98a7780ad9 100644
--- a/lib/Analysis/CostModel.cpp
+++ b/lib/Analysis/CostModel.cpp
@@ -88,6 +88,23 @@ static bool isReverseVectorMask(SmallVector<int, 16> &Mask) {
   return true;
 }
 
+static TargetTransformInfo::OperandValueKind getOperandInfo(Value *V) {
+  TargetTransformInfo::OperandValueKind OpInfo =
+    TargetTransformInfo::OK_AnyValue;
+
+  // Check for a splat of a constant.
+  ConstantDataVector *CDV = 0;
+  if ((CDV = dyn_cast<ConstantDataVector>(V)))
+    if (CDV->getSplatValue() != NULL)
+      OpInfo = TargetTransformInfo::OK_UniformConstantValue;
+  ConstantVector *CV = 0;
+  if ((CV = dyn_cast<ConstantVector>(V)))
+    if (CV->getSplatValue() != NULL)
+      OpInfo = TargetTransformInfo::OK_UniformConstantValue;
+
+  return OpInfo;
+}
+
 unsigned CostModelAnalysis::getInstructionCost(const Instruction *I) const {
   if (!TTI)
     return -1;
@@ -121,7 +138,12 @@ unsigned CostModelAnalysis::getInstructionCost(const Instruction *I) const {
   case Instruction::And:
   case Instruction::Or:
   case Instruction::Xor: {
-    return TTI->getArithmeticInstrCost(I->getOpcode(), I->getType());
+    TargetTransformInfo::OperandValueKind Op1VK =
+      getOperandInfo(I->getOperand(0));
+    TargetTransformInfo::OperandValueKind Op2VK =
+      getOperandInfo(I->getOperand(1));
+    return TTI->getArithmeticInstrCost(I->getOpcode(), I->getType(), Op1VK,
+                                       Op2VK);
   }
   case Instruction::Select: {
     const SelectInst *SI = cast<SelectInst>(I);
diff --git a/lib/Analysis/TargetTransformInfo.cpp b/lib/Analysis/TargetTransformInfo.cpp
index 976cd87321..64f8e96884 100644
--- a/lib/Analysis/TargetTransformInfo.cpp
+++ b/lib/Analysis/TargetTransformInfo.cpp
@@ -150,8 +150,10 @@ unsigned TargetTransformInfo::getMaximumUnrollFactor() const {
 }
 
 unsigned TargetTransformInfo::getArithmeticInstrCost(unsigned Opcode,
-                                                     Type *Ty) const {
-  return PrevTTI->getArithmeticInstrCost(Opcode, Ty);
+                                                Type *Ty,
+                                                OperandValueKind Op1Info,
+                                                OperandValueKind Op2Info) const {
+  return PrevTTI->getArithmeticInstrCost(Opcode, Ty, Op1Info, Op2Info);
 }
 
 unsigned TargetTransformInfo::getShuffleCost(ShuffleKind Kind, Type *Tp,
@@ -495,7 +497,8 @@ struct NoTTI : ImmutablePass, TargetTransformInfo {
     return 1;
   }
 
-  unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty) const {
+  unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty, OperandValueKind,
+                                  OperandValueKind) const {
     return 1;
   }
 
diff --git a/lib/CodeGen/BasicTargetTransformInfo.cpp b/lib/CodeGen/BasicTargetTransformInfo.cpp
index 4cd1b80dc0..012ff8ad83 100644
--- a/lib/CodeGen/BasicTargetTransformInfo.cpp
+++ b/lib/CodeGen/BasicTargetTransformInfo.cpp
@@ -85,7 +85,9 @@ public:
   virtual unsigned getNumberOfRegisters(bool Vector) const;
   virtual unsigned getMaximumUnrollFactor() const;
   virtual unsigned getRegisterBitWidth(bool Vector) const;
-  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty) const;
+  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                          OperandValueKind,
+                                          OperandValueKind) const;
   virtual unsigned getShuffleCost(ShuffleKind Kind, Type *Tp,
                                   int Index, Type *SubTp) const;
   virtual unsigned getCastInstrCost(unsigned Opcode, Type *Dst,
@@ -193,7 +195,9 @@ unsigned BasicTTI::getMaximumUnrollFactor() const {
   return 1;
 }
 
-unsigned BasicTTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty) const {
+unsigned BasicTTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                          OperandValueKind,
+                                          OperandValueKind) const {
   // Check if any of the operands are vector operands.
   int ISD = TLI->InstructionOpcodeToISD(Opcode);
   assert(ISD && "Invalid opcode");
diff --git a/lib/Target/PowerPC/PPCTargetTransformInfo.cpp b/lib/Target/PowerPC/PPCTargetTransformInfo.cpp
index 00037edafc..2504ba70c2 100644
--- a/lib/Target/PowerPC/PPCTargetTransformInfo.cpp
+++ b/lib/Target/PowerPC/PPCTargetTransformInfo.cpp
@@ -86,7 +86,9 @@ public:
   virtual unsigned getNumberOfRegisters(bool Vector) const;
   virtual unsigned getRegisterBitWidth(bool Vector) const;
   virtual unsigned getMaximumUnrollFactor() const;
-  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty) const;
+  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                          OperandValueKind,
+                                          OperandValueKind) const;
   virtual unsigned getShuffleCost(ShuffleKind Kind, Type *Tp,
                                   int Index, Type *SubTp) const;
   virtual unsigned getCastInstrCost(unsigned Opcode, Type *Dst,
@@ -166,11 +168,14 @@ unsigned PPCTTI::getMaximumUnrollFactor() const {
   return 2;
 }
 
-unsigned PPCTTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty) const {
+unsigned PPCTTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                        OperandValueKind Op1Info,
+                                        OperandValueKind Op2Info) const {
   assert(TLI->InstructionOpcodeToISD(Opcode) && "Invalid opcode");
 
   // Fallback to the default implementation.
-  return TargetTransformInfo::getArithmeticInstrCost(Opcode, Ty);
+  return TargetTransformInfo::getArithmeticInstrCost(Opcode, Ty, Op1Info,
+                                                     Op2Info);
 }
 
 unsigned PPCTTI::getShuffleCost(ShuffleKind Kind, Type *Tp, int Index,
diff --git a/lib/Target/X86/X86TargetTransformInfo.cpp b/lib/Target/X86/X86TargetTransformInfo.cpp
index a38b22bd78..e8940bfc23 100644
--- a/lib/Target/X86/X86TargetTransformInfo.cpp
+++ b/lib/Target/X86/X86TargetTransformInfo.cpp
@@ -86,7 +86,9 @@ public:
   virtual unsigned getNumberOfRegisters(bool Vector) const;
   virtual unsigned getRegisterBitWidth(bool Vector) const;
   virtual unsigned getMaximumUnrollFactor() const;
-  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty) const;
+  virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                          OperandValueKind,
+                                          OperandValueKind) const;
   virtual unsigned getShuffleCost(ShuffleKind Kind, Type *Tp,
                                   int Index, Type *SubTp) const;
   virtual unsigned getCastInstrCost(unsigned Opcode, Type *Dst,
@@ -162,7 +164,9 @@ unsigned X86TTI::getMaximumUnrollFactor() const {
   return 2;
 }
 
-unsigned X86TTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty) const {
+unsigned X86TTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty,
+                                        OperandValueKind Op1Info,
+                                        OperandValueKind Op2Info) const {
   // Legalize the type.
   std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(Ty);
 
@@ -278,7 +282,8 @@ unsigned X86TTI::getArithmeticInstrCost(unsigned Opcode, Type *Ty) const {
     return 6;
 
   // Fallback to the default implementation.
-  return TargetTransformInfo::getArithmeticInstrCost(Opcode, Ty);
+  return TargetTransformInfo::getArithmeticInstrCost(Opcode, Ty, Op1Info,
+                                                     Op2Info);
 }
 
 unsigned X86TTI::getShuffleCost(ShuffleKind Kind, Type *Tp, int Index,