[AArch64][TTI][NFC] Overload method 'getVectorInstrCost' to provide vector instruction itself, as a context information for cost estimation.

1) Overloaded (instruction-based) method is a wrapper around the current (opcode-based) method.
2) This patch also changes a few callsites (VectorCombine.cpp,
   SLPVectorizer.cpp, CodeGenPrepare.cpp) to call the overloaded method.
3) This is a split of D128302.

Differential Revision: https://reviews.llvm.org/D131114

16 files changed, 67 insertions, 23 deletions

diff --git a/llvm/include/llvm/Analysis/TargetTransformInfo.h b/llvm/include/llvm/Analysis/TargetTransformInfo.h
index f613cc3fae94..13c9a2eaeefa 100644
--- a/llvm/include/llvm/Analysis/TargetTransformInfo.h
+++ b/llvm/include/llvm/Analysis/TargetTransformInfo.h
@@ -1166,9 +1166,21 @@ public:
 
   /// \return The expected cost of vector Insert and Extract.
   /// Use -1 to indicate that there is no information on the index value.
+  /// This is used when the instruction is not available; a typical use
+  /// case is to provision the cost of vectorization/scalarization in
+  /// vectorizer passes.
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index = -1) const;
 
+  /// \return The expected cost of vector Insert and Extract.
+  /// This is used when instruction is available, and implementation
+  /// asserts 'I' is not nullptr.
+  ///
+  /// A typical suitable use case is cost estimation when vector instruction
+  /// exists (e.g., from basic blocks during transformation).
+  InstructionCost getVectorInstrCost(const Instruction *I, Type *Val,
+                                     unsigned Index = -1) const;
+
   /// \return The cost of replication shuffle of \p VF elements typed \p EltTy
   /// \p ReplicationFactor times.
   ///
@@ -1747,6 +1759,8 @@ public:
                                              const Instruction *I) = 0;
   virtual InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                              unsigned Index) = 0;
+  virtual InstructionCost getVectorInstrCost(const Instruction *I, Type *Val,
+                                             unsigned Index) = 0;
 
   virtual InstructionCost
   getReplicationShuffleCost(Type *EltTy, int ReplicationFactor, int VF,
@@ -2305,6 +2319,10 @@ public:
                                      unsigned Index) override {
     return Impl.getVectorInstrCost(Opcode, Val, Index);
   }
+  InstructionCost getVectorInstrCost(const Instruction *I, Type *Val,
+                                     unsigned Index) override {
+    return Impl.getVectorInstrCost(I, Val, Index);
+  }
   InstructionCost
   getReplicationShuffleCost(Type *EltTy, int ReplicationFactor, int VF,
                             const APInt &DemandedDstElts,
diff --git a/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h b/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
index f3376a3982a6..514a970c5999 100644
--- a/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
+++ b/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
@@ -575,6 +575,11 @@ public:
     return 1;
   }
 
+  InstructionCost getVectorInstrCost(const Instruction *I, Type *Val,
+                                     unsigned Index) const {
+    return 1;
+  }
+
   unsigned getReplicationShuffleCost(Type *EltTy, int ReplicationFactor, int VF,
                                      const APInt &DemandedDstElts,
                                      TTI::TargetCostKind CostKind) {
@@ -1148,7 +1153,7 @@ public:
       if (auto *CI = dyn_cast<ConstantInt>(IE->getOperand(2)))
         if (CI->getValue().getActiveBits() <= 32)
           Idx = CI->getZExtValue();
-      return TargetTTI->getVectorInstrCost(Opcode, Ty, Idx);
+      return TargetTTI->getVectorInstrCost(IE, Ty, Idx);
     }
     case Instruction::ShuffleVector: {
       auto *Shuffle = dyn_cast<ShuffleVectorInst>(U);
@@ -1238,7 +1243,7 @@ public:
         if (CI->getValue().getActiveBits() <= 32)
           Idx = CI->getZExtValue();
       Type *DstTy = U->getOperand(0)->getType();
-      return TargetTTI->getVectorInstrCost(Opcode, DstTy, Idx);
+      return TargetTTI->getVectorInstrCost(EEI, DstTy, Idx);
     }
     }
     // By default, just classify everything as 'basic'.
diff --git a/llvm/include/llvm/CodeGen/BasicTTIImpl.h b/llvm/include/llvm/CodeGen/BasicTTIImpl.h
index 33c93e4c56ac..2b092d48121e 100644
--- a/llvm/include/llvm/CodeGen/BasicTTIImpl.h
+++ b/llvm/include/llvm/CodeGen/BasicTTIImpl.h
@@ -1159,6 +1159,11 @@ public:
     return LT.first;
   }
 
+  InstructionCost getVectorInstrCost(const Instruction *I, Type *Val,
+                                     unsigned Index) {
+    return thisT()->getVectorInstrCost(I->getOpcode(), Val, Index);
+  }
+
   InstructionCost getReplicationShuffleCost(Type *EltTy, int ReplicationFactor,
                                             int VF,
                                             const APInt &DemandedDstElts,
diff --git a/llvm/lib/Analysis/TargetTransformInfo.cpp b/llvm/lib/Analysis/TargetTransformInfo.cpp
index 4d3d549803db..c86279eae9a5 100644
--- a/llvm/lib/Analysis/TargetTransformInfo.cpp
+++ b/llvm/lib/Analysis/TargetTransformInfo.cpp
@@ -869,11 +869,26 @@ InstructionCost TargetTransformInfo::getCmpSelInstrCost(
 InstructionCost TargetTransformInfo::getVectorInstrCost(unsigned Opcode,
                                                         Type *Val,
                                                         unsigned Index) const {
+  // FIXME: Assert that Opcode is either InsertElement or ExtractElement.
+  // This is mentioned in the interface description and respected by all
+  // callers, but never asserted upon.
   InstructionCost Cost = TTIImpl->getVectorInstrCost(Opcode, Val, Index);
   assert(Cost >= 0 && "TTI should not produce negative costs!");
   return Cost;
 }
 
+InstructionCost TargetTransformInfo::getVectorInstrCost(const Instruction *I,
+                                                        Type *Val,
+                                                        unsigned Index) const {
+  assert((I != nullptr) && "Expect not-null instruction pointer");
+  // FIXME: Assert that Opcode is either InsertElement or ExtractElement.
+  // This is mentioned in the interface description and respected by all
+  // callers, but never asserted upon.
+  InstructionCost Cost = TTIImpl->getVectorInstrCost(I, Val, Index);
+  assert(Cost >= 0 && "TTI should not produce negative costs!");
+  return Cost;
+}
+
 InstructionCost TargetTransformInfo::getReplicationShuffleCost(
     Type *EltTy, int ReplicationFactor, int VF, const APInt &DemandedDstElts,
     TTI::TargetCostKind CostKind) {
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index b100fbe2b33c..57423ffbae13 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -7261,7 +7261,7 @@ class VectorPromoteHelper {
     // scalar to vector.
     // The vector chain has to account for the combining cost.
     InstructionCost ScalarCost =
-        TTI.getVectorInstrCost(Transition->getOpcode(), PromotedType, Index);
+        TTI.getVectorInstrCost(Transition, PromotedType, Index);
     InstructionCost VectorCost = StoreExtractCombineCost;
     enum TargetTransformInfo::TargetCostKind CostKind =
       TargetTransformInfo::TCK_RecipThroughput;
diff --git a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h
index fcd0df6f1d06..c4e7135d8b57 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h
+++ b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h
@@ -173,6 +173,7 @@ public:
   InstructionCost getCFInstrCost(unsigned Opcode, TTI::TargetCostKind CostKind,
                                  const Instruction *I = nullptr);
 
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
 
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.h b/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.h
index f2260c31e678..eeb304311342 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.h
@@ -159,6 +159,7 @@ public:
   bool isInlineAsmSourceOfDivergence(const CallInst *CI,
                                      ArrayRef<unsigned> Indices = {}) const;
 
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *ValTy,
                                      unsigned Index);
   bool isSourceOfDivergence(const Value *V) const;
diff --git a/llvm/lib/Target/AMDGPU/R600TargetTransformInfo.h b/llvm/lib/Target/AMDGPU/R600TargetTransformInfo.h
index 544292bc4fd9..f1a198fd14e4 100644
--- a/llvm/lib/Target/AMDGPU/R600TargetTransformInfo.h
+++ b/llvm/lib/Target/AMDGPU/R600TargetTransformInfo.h
@@ -60,6 +60,7 @@ public:
   unsigned getMaxInterleaveFactor(unsigned VF);
   InstructionCost getCFInstrCost(unsigned Opcode, TTI::TargetCostKind CostKind,
                                  const Instruction *I = nullptr);
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *ValTy,
                                      unsigned Index);
 };
diff --git a/llvm/lib/Target/ARM/ARMTargetTransformInfo.h b/llvm/lib/Target/ARM/ARMTargetTransformInfo.h
index 5e39e85bac87..a3aed48f6beb 100644
--- a/llvm/lib/Target/ARM/ARMTargetTransformInfo.h
+++ b/llvm/lib/Target/ARM/ARMTargetTransformInfo.h
@@ -237,6 +237,7 @@ public:
                                      TTI::TargetCostKind CostKind,
                                      const Instruction *I = nullptr);
 
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
 
diff --git a/llvm/lib/Target/Hexagon/HexagonTargetTransformInfo.h b/llvm/lib/Target/Hexagon/HexagonTargetTransformInfo.h
index 7bbaf7ae9cb2..9d263642173a 100644
--- a/llvm/lib/Target/Hexagon/HexagonTargetTransformInfo.h
+++ b/llvm/lib/Target/Hexagon/HexagonTargetTransformInfo.h
@@ -151,6 +151,7 @@ public:
                                    TTI::CastContextHint CCH,
                                    TTI::TargetCostKind CostKind,
                                    const Instruction *I = nullptr);
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
 
diff --git a/llvm/lib/Target/PowerPC/PPCTargetTransformInfo.h b/llvm/lib/Target/PowerPC/PPCTargetTransformInfo.h
index 790eb0b42afa..9a5b26bd1b90 100644
--- a/llvm/lib/Target/PowerPC/PPCTargetTransformInfo.h
+++ b/llvm/lib/Target/PowerPC/PPCTargetTransformInfo.h
@@ -123,6 +123,7 @@ public:
                                      CmpInst::Predicate VecPred,
                                      TTI::TargetCostKind CostKind,
                                      const Instruction *I = nullptr);
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
   InstructionCost getMemoryOpCost(unsigned Opcode, Type *Src,
diff --git a/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.h b/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.h
index 33317e799eab..0804d0330de4 100644
--- a/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.h
+++ b/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.h
@@ -107,6 +107,7 @@ public:
                                      CmpInst::Predicate VecPred,
                                      TTI::TargetCostKind CostKind,
                                      const Instruction *I = nullptr);
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
   bool isFoldableLoad(const LoadInst *Ld, const Instruction *&FoldedValue);
diff --git a/llvm/lib/Target/WebAssembly/WebAssemblyTargetTransformInfo.h b/llvm/lib/Target/WebAssembly/WebAssemblyTargetTransformInfo.h
index fde58a9587b6..10179eff2620 100644
--- a/llvm/lib/Target/WebAssembly/WebAssemblyTargetTransformInfo.h
+++ b/llvm/lib/Target/WebAssembly/WebAssemblyTargetTransformInfo.h
@@ -67,6 +67,7 @@ public:
       TTI::OperandValueProperties Opd2PropInfo = TTI::OP_None,
       ArrayRef<const Value *> Args = ArrayRef<const Value *>(),
       const Instruction *CxtI = nullptr);
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
 
diff --git a/llvm/lib/Target/X86/X86TargetTransformInfo.h b/llvm/lib/Target/X86/X86TargetTransformInfo.h
index bd3c3fb1bb2f..9f83c0461b56 100644
--- a/llvm/lib/Target/X86/X86TargetTransformInfo.h
+++ b/llvm/lib/Target/X86/X86TargetTransformInfo.h
@@ -146,6 +146,7 @@ public:
                                      CmpInst::Predicate VecPred,
                                      TTI::TargetCostKind CostKind,
                                      const Instruction *I = nullptr);
+  using BaseT::getVectorInstrCost;
   InstructionCost getVectorInstrCost(unsigned Opcode, Type *Val,
                                      unsigned Index);
   InstructionCost getScalarizationOverhead(VectorType *Ty,
diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index d69d1e3d19f3..737fcb386764 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -5882,8 +5882,7 @@ InstructionCost BoUpSLP::getEntryCost(const TreeEntry *E,
           continue;
         }
       }
-      Cost -= TTIRef.getVectorInstrCost(Instruction::ExtractElement,
-                                        EE->getVectorOperandType(), Idx);
+      Cost -= TTIRef.getVectorInstrCost(EE, EE->getVectorOperandType(), Idx);
     }
     // Add a cost for subvector extracts/inserts if required.
     for (const auto &Data : ExtractVectorsTys) {
@@ -6116,9 +6115,8 @@ InstructionCost BoUpSLP::getEntryCost(const TreeEntry *E,
         for (unsigned I : E->ReuseShuffleIndices) {
           if (ShuffleOrOp == Instruction::ExtractElement) {
             auto *EE = cast<ExtractElementInst>(VL[I]);
-            CommonCost -= TTI->getVectorInstrCost(Instruction::ExtractElement,
-                                                  EE->getVectorOperandType(),
-                                                  *getExtractIndex(EE));
+            CommonCost -= TTI->getVectorInstrCost(
+                EE, EE->getVectorOperandType(), *getExtractIndex(EE));
           } else {
             CommonCost -= TTI->getVectorInstrCost(Instruction::ExtractElement,
                                                   VecTy, Idx);
@@ -6129,9 +6127,8 @@ InstructionCost BoUpSLP::getEntryCost(const TreeEntry *E,
         for (Value *V : VL) {
           if (ShuffleOrOp == Instruction::ExtractElement) {
             auto *EE = cast<ExtractElementInst>(V);
-            CommonCost += TTI->getVectorInstrCost(Instruction::ExtractElement,
-                                                  EE->getVectorOperandType(),
-                                                  *getExtractIndex(EE));
+            CommonCost += TTI->getVectorInstrCost(
+                EE, EE->getVectorOperandType(), *getExtractIndex(EE));
           } else {
             --Idx;
             CommonCost += TTI->getVectorInstrCost(Instruction::ExtractElement,
diff --git a/llvm/lib/Transforms/Vectorize/VectorCombine.cpp b/llvm/lib/Transforms/Vectorize/VectorCombine.cpp
index a38936644bd3..37901d8bce4d 100644
--- a/llvm/lib/Transforms/Vectorize/VectorCombine.cpp
+++ b/llvm/lib/Transforms/Vectorize/VectorCombine.cpp
@@ -270,10 +270,8 @@ ExtractElementInst *VectorCombine::getShuffleExtract(
 
   Type *VecTy = Ext0->getVectorOperand()->getType();
   assert(VecTy == Ext1->getVectorOperand()->getType() && "Need matching types");
-  InstructionCost Cost0 =
-      TTI.getVectorInstrCost(Ext0->getOpcode(), VecTy, Index0);
-  InstructionCost Cost1 =
-      TTI.getVectorInstrCost(Ext1->getOpcode(), VecTy, Index1);
+  InstructionCost Cost0 = TTI.getVectorInstrCost(Ext0, VecTy, Index0);
+  InstructionCost Cost1 = TTI.getVectorInstrCost(Ext1, VecTy, Index1);
 
   // If both costs are invalid no shuffle is needed
   if (!Cost0.isValid() && !Cost1.isValid())
@@ -337,10 +335,8 @@ bool VectorCombine::isExtractExtractCheap(ExtractElementInst *Ext0,
   unsigned Ext0Index = Ext0IndexC->getZExtValue();
   unsigned Ext1Index = Ext1IndexC->getZExtValue();
 
-  InstructionCost Extract0Cost =
-      TTI.getVectorInstrCost(Instruction::ExtractElement, VecTy, Ext0Index);
-  InstructionCost Extract1Cost =
-      TTI.getVectorInstrCost(Instruction::ExtractElement, VecTy, Ext1Index);
+  InstructionCost Extract0Cost = TTI.getVectorInstrCost(Ext0, VecTy, Ext0Index);
+  InstructionCost Extract1Cost = TTI.getVectorInstrCost(Ext1, VecTy, Ext1Index);
 
   // A more expensive extract will always be replaced by a splat shuffle.
   // For example, if Ext0 is more expensive:
@@ -754,9 +750,8 @@ bool VectorCombine::foldExtractedCmps(Instruction &I) {
   if (!VecTy)
     return false;
 
-  InstructionCost OldCost =
-      TTI.getVectorInstrCost(Ext0->getOpcode(), VecTy, Index0);
-  OldCost += TTI.getVectorInstrCost(Ext1->getOpcode(), VecTy, Index1);
+  InstructionCost OldCost = TTI.getVectorInstrCost(Ext0, VecTy, Index0);
+  OldCost += TTI.getVectorInstrCost(Ext1, VecTy, Index1);
   OldCost +=
       TTI.getCmpSelInstrCost(CmpOpcode, I0->getType(),
                              CmpInst::makeCmpResultType(I0->getType()), Pred) *
@@ -776,7 +771,7 @@ bool VectorCombine::foldExtractedCmps(Instruction &I) {
   NewCost += TTI.getShuffleCost(TargetTransformInfo::SK_PermuteSingleSrc, CmpTy,
                                 ShufMask);
   NewCost += TTI.getArithmeticInstrCost(I.getOpcode(), CmpTy);
-  NewCost += TTI.getVectorInstrCost(Ext0->getOpcode(), CmpTy, CheapIndex);
+  NewCost += TTI.getVectorInstrCost(Ext0, CmpTy, CheapIndex);
 
   // Aggressively form vector ops if the cost is equal because the transform
   // may enable further optimization.