diff options
Diffstat (limited to 'llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp')
-rw-r--r-- | llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp | 313 |
1 files changed, 211 insertions, 102 deletions
diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp index 3524d84b1f1..62a7adb9d98 100644 --- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp +++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp @@ -268,104 +268,6 @@ static bool CanReuseExtract(ArrayRef<Value *> VL) { return true; } -static void reorderInputsAccordingToOpcode(ArrayRef<Value *> VL, - SmallVectorImpl<Value *> &Left, - SmallVectorImpl<Value *> &Right) { - - SmallVector<Value *, 16> OrigLeft, OrigRight; - - bool AllSameOpcodeLeft = true; - bool AllSameOpcodeRight = true; - for (unsigned i = 0, e = VL.size(); i != e; ++i) { - Instruction *I = cast<Instruction>(VL[i]); - Value *V0 = I->getOperand(0); - Value *V1 = I->getOperand(1); - - OrigLeft.push_back(V0); - OrigRight.push_back(V1); - - Instruction *I0 = dyn_cast<Instruction>(V0); - Instruction *I1 = dyn_cast<Instruction>(V1); - - // Check whether all operands on one side have the same opcode. In this case - // we want to preserve the original order and not make things worse by - // reordering. - AllSameOpcodeLeft = I0; - AllSameOpcodeRight = I1; - - if (i && AllSameOpcodeLeft) { - if(Instruction *P0 = dyn_cast<Instruction>(OrigLeft[i-1])) { - if(P0->getOpcode() != I0->getOpcode()) - AllSameOpcodeLeft = false; - } else - AllSameOpcodeLeft = false; - } - if (i && AllSameOpcodeRight) { - if(Instruction *P1 = dyn_cast<Instruction>(OrigRight[i-1])) { - if(P1->getOpcode() != I1->getOpcode()) - AllSameOpcodeRight = false; - } else - AllSameOpcodeRight = false; - } - - // Sort two opcodes. In the code below we try to preserve the ability to use - // broadcast of values instead of individual inserts. - // vl1 = load - // vl2 = phi - // vr1 = load - // vr2 = vr2 - // = vl1 x vr1 - // = vl2 x vr2 - // If we just sorted according to opcode we would leave the first line in - // tact but we would swap vl2 with vr2 because opcode(phi) > opcode(load). - // = vl1 x vr1 - // = vr2 x vl2 - // Because vr2 and vr1 are from the same load we loose the opportunity of a - // broadcast for the packed right side in the backend: we have [vr1, vl2] - // instead of [vr1, vr2=vr1]. - if (I0 && I1) { - if(!i && I0->getOpcode() > I1->getOpcode()) { - Left.push_back(I1); - Right.push_back(I0); - } else if (i && I0->getOpcode() > I1->getOpcode() && Right[i-1] != I1) { - // Try not to destroy a broad cast for no apparent benefit. - Left.push_back(I1); - Right.push_back(I0); - } else if (i && I0->getOpcode() == I1->getOpcode() && Right[i-1] == I0) { - // Try preserve broadcasts. - Left.push_back(I1); - Right.push_back(I0); - } else if (i && I0->getOpcode() == I1->getOpcode() && Left[i-1] == I1) { - // Try preserve broadcasts. - Left.push_back(I1); - Right.push_back(I0); - } else { - Left.push_back(I0); - Right.push_back(I1); - } - continue; - } - // One opcode, put the instruction on the right. - if (I0) { - Left.push_back(V1); - Right.push_back(I0); - continue; - } - Left.push_back(V0); - Right.push_back(V1); - } - - bool LeftBroadcast = isSplat(Left); - bool RightBroadcast = isSplat(Right); - - // Don't reorder if the operands where good to begin with. - if (!(LeftBroadcast || RightBroadcast) && - (AllSameOpcodeRight || AllSameOpcodeLeft)) { - Left = OrigLeft; - Right = OrigRight; - } -} - /// \returns True if in-tree use also needs extract. This refers to /// possible scalar operand in vectorized instruction. static bool InTreeUserNeedToExtract(Value *Scalar, Instruction *UserInst, @@ -508,6 +410,16 @@ private: /// be beneficial even the tree height is tiny. bool isFullyVectorizableTinyTree(); + /// \reorder commutative operands in alt shuffle if they result in + /// vectorized code. + void reorderAltShuffleOperands(ArrayRef<Value *> VL, + SmallVectorImpl<Value *> &Left, + SmallVectorImpl<Value *> &Right); + /// \reorder commutative operands to get better probability of + /// generating vectorized code. + void reorderInputsAccordingToOpcode(ArrayRef<Value *> VL, + SmallVectorImpl<Value *> &Left, + SmallVectorImpl<Value *> &Right); struct TreeEntry { TreeEntry() : Scalars(), VectorizedValue(nullptr), NeedToGather(0) {} @@ -1441,6 +1353,16 @@ void BoUpSLP::buildTree_rec(ArrayRef<Value *> VL, unsigned Depth) { } newTreeEntry(VL, true); DEBUG(dbgs() << "SLP: added a ShuffleVector op.\n"); + + // Reorder operands if reordering would enable vectorization. + if (isa<BinaryOperator>(VL0)) { + ValueList Left, Right; + reorderAltShuffleOperands(VL, Left, Right); + buildTree_rec(Left, Depth + 1); + buildTree_rec(Right, Depth + 1); + return; + } + for (unsigned i = 0, e = VL0->getNumOperands(); i < e; ++i) { ValueList Operands; // Prepare the operand vector. @@ -1878,6 +1800,195 @@ bool BoUpSLP::isConsecutiveAccess(Value *A, Value *B) { return X == PtrSCEVB; } +// Reorder commutative operations in alternate shuffle if the resulting vectors +// are consecutive loads. This would allow us to vectorize the tree. +// If we have something like- +// load a[0] - load b[0] +// load b[1] + load a[1] +// load a[2] - load b[2] +// load a[3] + load b[3] +// Reordering the second load b[1] load a[1] would allow us to vectorize this +// code. +void BoUpSLP::reorderAltShuffleOperands(ArrayRef<Value *> VL, + SmallVectorImpl<Value *> &Left, + SmallVectorImpl<Value *> &Right) { + + // Push left and right operands of binary operation into Left and Right + for (unsigned i = 0, e = VL.size(); i < e; ++i) { + Left.push_back(cast<Instruction>(VL[i])->getOperand(0)); + Right.push_back(cast<Instruction>(VL[i])->getOperand(1)); + } + + // Reorder if we have a commutative operation and consecutive access + // are on either side of the alternate instructions. + for (unsigned j = 0; j < VL.size() - 1; ++j) { + if (LoadInst *L = dyn_cast<LoadInst>(Left[j])) { + if (LoadInst *L1 = dyn_cast<LoadInst>(Right[j + 1])) { + Instruction *VL1 = cast<Instruction>(VL[j]); + Instruction *VL2 = cast<Instruction>(VL[j + 1]); + if (isConsecutiveAccess(L, L1) && VL1->isCommutative()) { + std::swap(Left[j], Right[j]); + continue; + } else if (isConsecutiveAccess(L, L1) && VL2->isCommutative()) { + std::swap(Left[j + 1], Right[j + 1]); + continue; + } + // else unchanged + } + } + if (LoadInst *L = dyn_cast<LoadInst>(Right[j])) { + if (LoadInst *L1 = dyn_cast<LoadInst>(Left[j + 1])) { + Instruction *VL1 = cast<Instruction>(VL[j]); + Instruction *VL2 = cast<Instruction>(VL[j + 1]); + if (isConsecutiveAccess(L, L1) && VL1->isCommutative()) { + std::swap(Left[j], Right[j]); + continue; + } else if (isConsecutiveAccess(L, L1) && VL2->isCommutative()) { + std::swap(Left[j + 1], Right[j + 1]); + continue; + } + // else unchanged + } + } + } +} + +void BoUpSLP::reorderInputsAccordingToOpcode(ArrayRef<Value *> VL, + SmallVectorImpl<Value *> &Left, + SmallVectorImpl<Value *> &Right) { + + SmallVector<Value *, 16> OrigLeft, OrigRight; + + bool AllSameOpcodeLeft = true; + bool AllSameOpcodeRight = true; + for (unsigned i = 0, e = VL.size(); i != e; ++i) { + Instruction *I = cast<Instruction>(VL[i]); + Value *VLeft = I->getOperand(0); + Value *VRight = I->getOperand(1); + + OrigLeft.push_back(VLeft); + OrigRight.push_back(VRight); + + Instruction *ILeft = dyn_cast<Instruction>(VLeft); + Instruction *IRight = dyn_cast<Instruction>(VRight); + + // Check whether all operands on one side have the same opcode. In this case + // we want to preserve the original order and not make things worse by + // reordering. + if (i && AllSameOpcodeLeft && ILeft) { + if (Instruction *PLeft = dyn_cast<Instruction>(OrigLeft[i - 1])) { + if (PLeft->getOpcode() != ILeft->getOpcode()) + AllSameOpcodeLeft = false; + } else + AllSameOpcodeLeft = false; + } + if (i && AllSameOpcodeRight && IRight) { + if (Instruction *PRight = dyn_cast<Instruction>(OrigRight[i - 1])) { + if (PRight->getOpcode() != IRight->getOpcode()) + AllSameOpcodeRight = false; + } else + AllSameOpcodeRight = false; + } + + // Sort two opcodes. In the code below we try to preserve the ability to use + // broadcast of values instead of individual inserts. + // vl1 = load + // vl2 = phi + // vr1 = load + // vr2 = vr2 + // = vl1 x vr1 + // = vl2 x vr2 + // If we just sorted according to opcode we would leave the first line in + // tact but we would swap vl2 with vr2 because opcode(phi) > opcode(load). + // = vl1 x vr1 + // = vr2 x vl2 + // Because vr2 and vr1 are from the same load we loose the opportunity of a + // broadcast for the packed right side in the backend: we have [vr1, vl2] + // instead of [vr1, vr2=vr1]. + if (ILeft && IRight) { + if (!i && ILeft->getOpcode() > IRight->getOpcode()) { + Left.push_back(IRight); + Right.push_back(ILeft); + } else if (i && ILeft->getOpcode() > IRight->getOpcode() && + Right[i - 1] != IRight) { + // Try not to destroy a broad cast for no apparent benefit. + Left.push_back(IRight); + Right.push_back(ILeft); + } else if (i && ILeft->getOpcode() == IRight->getOpcode() && + Right[i - 1] == ILeft) { + // Try preserve broadcasts. + Left.push_back(IRight); + Right.push_back(ILeft); + } else if (i && ILeft->getOpcode() == IRight->getOpcode() && + Left[i - 1] == IRight) { + // Try preserve broadcasts. + Left.push_back(IRight); + Right.push_back(ILeft); + } else { + Left.push_back(ILeft); + Right.push_back(IRight); + } + continue; + } + // One opcode, put the instruction on the right. + if (ILeft) { + Left.push_back(VRight); + Right.push_back(ILeft); + continue; + } + Left.push_back(VLeft); + Right.push_back(VRight); + } + + bool LeftBroadcast = isSplat(Left); + bool RightBroadcast = isSplat(Right); + + // If operands end up being broadcast return this operand order. + if (LeftBroadcast || RightBroadcast) + return; + + // Don't reorder if the operands where good to begin. + if (AllSameOpcodeRight || AllSameOpcodeLeft) { + Left = OrigLeft; + Right = OrigRight; + } + + // Finally check if we can get longer vectorizable chain by reordering + // without breaking the good operand order detected above. + // E.g. If we have something like- + // load a[0] load b[0] + // load b[1] load a[1] + // load a[2] load b[2] + // load a[3] load b[3] + // Reordering the second load b[1] load a[1] would allow us to vectorize + // this code and we still retain AllSameOpcode property. + // FIXME: This load reordering might break AllSameOpcode in some rare cases + // such as- + // add a[0],c[0] load b[0] + // add a[1],c[2] load b[1] + // b[2] load b[2] + // add a[3],c[3] load b[3] + for (unsigned j = 0; j < VL.size() - 1; ++j) { + if (LoadInst *L = dyn_cast<LoadInst>(Left[j])) { + if (LoadInst *L1 = dyn_cast<LoadInst>(Right[j + 1])) { + if (isConsecutiveAccess(L, L1)) { + std::swap(Left[j + 1], Right[j + 1]); + continue; + } + } + } + if (LoadInst *L = dyn_cast<LoadInst>(Right[j])) { + if (LoadInst *L1 = dyn_cast<LoadInst>(Left[j + 1])) { + if (isConsecutiveAccess(L, L1)) { + std::swap(Left[j + 1], Right[j + 1]); + continue; + } + } + } + // else unchanged + } +} + void BoUpSLP::setInsertPointAfterBundle(ArrayRef<Value *> VL) { Instruction *VL0 = cast<Instruction>(VL[0]); BasicBlock::iterator NextInst = VL0; @@ -2274,10 +2385,8 @@ Value *BoUpSLP::vectorizeTree(TreeEntry *E) { } case Instruction::ShuffleVector: { ValueList LHSVL, RHSVL; - for (int i = 0, e = E->Scalars.size(); i < e; ++i) { - LHSVL.push_back(cast<Instruction>(E->Scalars[i])->getOperand(0)); - RHSVL.push_back(cast<Instruction>(E->Scalars[i])->getOperand(1)); - } + assert(isa<BinaryOperator>(VL0) && "Invalid Shuffle Vector Operand"); + reorderAltShuffleOperands(E->Scalars, LHSVL, RHSVL); setInsertPointAfterBundle(E->Scalars); Value *LHS = vectorizeTree(LHSVL); |