//===-- SpillPlacement.cpp - Optimal Spill Code Placement -----------------===// // // The LLVM Compiler Infrastructure // // This file is distributed under the University of Illinois Open Source // License. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // // This file implements the spill code placement analysis. // // Each edge bundle corresponds to a node in a Hopfield network. Constraints on // basic blocks are weighted by the block frequency and added to become the node // bias. // // Transparent basic blocks have the variable live through, but don't care if it // is spilled or in a register. These blocks become connections in the Hopfield // network, again weighted by block frequency. // // The Hopfield network minimizes (possibly locally) its energy function: // // E = -sum_n V_n * ( B_n + sum_{n, m linked by b} V_m * F_b ) // // The energy function represents the expected spill code execution frequency, // or the cost of spilling. This is a Lyapunov function which never increases // when a node is updated. It is guaranteed to converge to a local minimum. // //===----------------------------------------------------------------------===// #define DEBUG_TYPE "spillplacement" #include "SpillPlacement.h" #include "llvm/CodeGen/EdgeBundles.h" #include "llvm/CodeGen/LiveIntervalAnalysis.h" #include "llvm/CodeGen/MachineBasicBlock.h" #include "llvm/CodeGen/MachineFunction.h" #include "llvm/CodeGen/MachineLoopInfo.h" #include "llvm/CodeGen/Passes.h" #include "llvm/Support/Debug.h" #include "llvm/Support/Format.h" using namespace llvm; char SpillPlacement::ID = 0; INITIALIZE_PASS_BEGIN(SpillPlacement, "spill-code-placement", "Spill Code Placement Analysis", true, true) INITIALIZE_PASS_DEPENDENCY(EdgeBundles) INITIALIZE_PASS_DEPENDENCY(MachineLoopInfo) INITIALIZE_PASS_END(SpillPlacement, "spill-code-placement", "Spill Code Placement Analysis", true, true) char &llvm::SpillPlacementID = SpillPlacement::ID; void SpillPlacement::getAnalysisUsage(AnalysisUsage &AU) const { AU.setPreservesAll(); AU.addRequiredTransitive<EdgeBundles>(); AU.addRequiredTransitive<MachineLoopInfo>(); MachineFunctionPass::getAnalysisUsage(AU); } /// Node - Each edge bundle corresponds to a Hopfield node. /// /// The node contains precomputed frequency data that only depends on the CFG, /// but Bias and Links are computed each time placeSpills is called. /// /// The node Value is positive when the variable should be in a register. The /// value can change when linked nodes change, but convergence is very fast /// because all weights are positive. /// struct SpillPlacement::Node { /// Scale - Inverse block frequency feeding into[0] or out of[1] the bundle. /// Ideally, these two numbers should be identical, but inaccuracies in the /// block frequency estimates means that we need to normalize ingoing and /// outgoing frequencies separately so they are commensurate. float Scale[2]; /// Bias - Normalized contributions from non-transparent blocks. /// A bundle connected to a MustSpill block has a huge negative bias, /// otherwise it is a number in the range [-2;2]. float Bias; /// Value - Output value of this node computed from the Bias and links. /// This is always in the range [-1;1]. A positive number means the variable /// should go in a register through this bundle. float Value; typedef SmallVector<std::pair<float, unsigned>, 4> LinkVector; /// Links - (Weight, BundleNo) for all transparent blocks connecting to other /// bundles. The weights are all positive and add up to at most 2, weights /// from ingoing and outgoing nodes separately add up to a most 1. The weight /// sum can be less than 2 when the variable is not live into / out of some /// connected basic blocks. LinkVector Links; /// preferReg - Return true when this node prefers to be in a register. bool preferReg() const { // Undecided nodes (Value==0) go on the stack. return Value > 0; } /// mustSpill - Return True if this node is so biased that it must spill. bool mustSpill() const { // Actually, we must spill if Bias < sum(weights). // It may be worth it to compute the weight sum here? return Bias < -2.0f; } /// Node - Create a blank Node. Node() { Scale[0] = Scale[1] = 0; } /// clear - Reset per-query data, but preserve frequencies that only depend on // the CFG. void clear() { Bias = Value = 0; Links.clear(); } /// addLink - Add a link to bundle b with weight w. /// out=0 for an ingoing link, and 1 for an outgoing link. void addLink(unsigned b, float w, bool out) { // Normalize w relative to all connected blocks from that direction. w *= Scale[out]; // There can be multiple links to the same bundle, add them up. for (LinkVector::iterator I = Links.begin(), E = Links.end(); I != E; ++I) if (I->second == b) { I->first += w; return; } // This must be the first link to b. Links.push_back(std::make_pair(w, b)); } /// addBias - Bias this node from an ingoing[0] or outgoing[1] link. /// Return the change to the total number of positive biases. void addBias(float w, bool out) { // Normalize w relative to all connected blocks from that direction. w *= Scale[out]; Bias += w; } /// update - Recompute Value from Bias and Links. Return true when node /// preference changes. bool update(const Node nodes[]) { // Compute the weighted sum of inputs. float Sum = Bias; for (LinkVector::iterator I = Links.begin(), E = Links.end(); I != E; ++I) Sum += I->first * nodes[I->second].Value; // The weighted sum is going to be in the range [-2;2]. Ideally, we should // simply set Value = sign(Sum), but we will add a dead zone around 0 for // two reasons: // 1. It avoids arbitrary bias when all links are 0 as is possible during // initial iterations. // 2. It helps tame rounding errors when the links nominally sum to 0. const float Thres = 1e-4f; bool Before = preferReg(); if (Sum < -Thres) Value = -1; else if (Sum > Thres) Value = 1; else Value = 0; return Before != preferReg(); } }; bool SpillPlacement::runOnMachineFunction(MachineFunction &mf) { MF = &mf; bundles = &getAnalysis<EdgeBundles>(); loops = &getAnalysis<MachineLoopInfo>(); assert(!nodes && "Leaking node array"); nodes = new Node[bundles->getNumBundles()]; // Compute total ingoing and outgoing block frequencies for all bundles. BlockFrequency.resize(mf.getNumBlockIDs()); for (MachineFunction::iterator I = mf.begin(), E = mf.end(); I != E; ++I) { float Freq = LiveIntervals::getSpillWeight(true, false, loops->getLoopDepth(I)); unsigned Num = I->getNumber(); BlockFrequency[Num] = Freq; nodes[bundles->getBundle(Num, 1)].Scale[0] += Freq; nodes[bundles->getBundle(Num, 0)].Scale[1] += Freq; } // Scales are reciprocal frequencies. for (unsigned i = 0, e = bundles->getNumBundles(); i != e; ++i) for (unsigned d = 0; d != 2; ++d) if (nodes[i].Scale[d] > 0) nodes[i].Scale[d] = 1 / nodes[i].Scale[d]; // We never change the function. return false; } void SpillPlacement::releaseMemory() { delete[] nodes; nodes = 0; } /// activate - mark node n as active if it wasn't already. void SpillPlacement::activate(unsigned n) { if (ActiveNodes->test(n)) return; ActiveNodes->set(n); nodes[n].clear(); // Very large bundles usually come from big switches, indirect branches, // landing pads, or loops with many 'continue' statements. It is difficult to // allocate registers when so many different blocks are involved. // // Give a small negative bias to large bundles such that 1/32 of the // connected blocks need to be interested before we consider expanding the // region through the bundle. This helps compile time by limiting the number // of blocks visited and the number of links in the Hopfield network. if (bundles->getBlocks(n).size() > 100) nodes[n].Bias = -0.0625f; } /// addConstraints - Compute node biases and weights from a set of constraints. /// Set a bit in NodeMask for each active node. void SpillPlacement::addConstraints(ArrayRef<BlockConstraint> LiveBlocks) { for (ArrayRef<BlockConstraint>::iterator I = LiveBlocks.begin(), E = LiveBlocks.end(); I != E; ++I) { float Freq = getBlockFrequency(I->Number); const float Bias[] = { 0, // DontCare, 1, // PrefReg, -1, // PrefSpill 0, // PrefBoth -HUGE_VALF // MustSpill }; // Live-in to block? if (I->Entry != DontCare) { unsigned ib = bundles->getBundle(I->Number, 0); activate(ib); nodes[ib].addBias(Freq * Bias[I->Entry], 1); } // Live-out from block? if (I->Exit != DontCare) { unsigned ob = bundles->getBundle(I->Number, 1); activate(ob); nodes[ob].addBias(Freq * Bias[I->Exit], 0); } } } /// addPrefSpill - Same as addConstraints(PrefSpill) void SpillPlacement::addPrefSpill(ArrayRef<unsigned> Blocks, bool Strong) { for (ArrayRef<unsigned>::iterator I = Blocks.begin(), E = Blocks.end(); I != E; ++I) { float Freq = getBlockFrequency(*I); if (Strong) Freq += Freq; unsigned ib = bundles->getBundle(*I, 0); unsigned ob = bundles->getBundle(*I, 1); activate(ib); activate(ob); nodes[ib].addBias(-Freq, 1); nodes[ob].addBias(-Freq, 0); } } void SpillPlacement::addLinks(ArrayRef<unsigned> Links) { for (ArrayRef<unsigned>::iterator I = Links.begin(), E = Links.end(); I != E; ++I) { unsigned Number = *I; unsigned ib = bundles->getBundle(Number, 0); unsigned ob = bundles->getBundle(Number, 1); // Ignore self-loops. if (ib == ob) continue; activate(ib); activate(ob); if (nodes[ib].Links.empty() && !nodes[ib].mustSpill()) Linked.push_back(ib); if (nodes[ob].Links.empty() && !nodes[ob].mustSpill()) Linked.push_back(ob); float Freq = getBlockFrequency(Number); nodes[ib].addLink(ob, Freq, 1); nodes[ob].addLink(ib, Freq, 0); } } bool SpillPlacement::scanActiveBundles() { Linked.clear(); RecentPositive.clear(); for (int n = ActiveNodes->find_first(); n>=0; n = ActiveNodes->find_next(n)) { nodes[n].update(nodes); // A node that must spill, or a node without any links is not going to // change its value ever again, so exclude it from iterations. if (nodes[n].mustSpill()) continue; if (!nodes[n].Links.empty()) Linked.push_back(n); if (nodes[n].preferReg()) RecentPositive.push_back(n); } return !RecentPositive.empty(); } /// iterate - Repeatedly update the Hopfield nodes until stability or the /// maximum number of iterations is reached. /// @param Linked - Numbers of linked nodes that need updating. void SpillPlacement::iterate() { // First update the recently positive nodes. They have likely received new // negative bias that will turn them off. while (!RecentPositive.empty()) nodes[RecentPositive.pop_back_val()].update(nodes); if (Linked.empty()) return; // Run up to 10 iterations. The edge bundle numbering is closely related to // basic block numbering, so there is a strong tendency towards chains of // linked nodes with sequential numbers. By scanning the linked nodes // backwards and forwards, we make it very likely that a single node can // affect the entire network in a single iteration. That means very fast // convergence, usually in a single iteration. for (unsigned iteration = 0; iteration != 10; ++iteration) { // Scan backwards, skipping the last node which was just updated. bool Changed = false; for (SmallVectorImpl<unsigned>::const_reverse_iterator I = llvm::next(Linked.rbegin()), E = Linked.rend(); I != E; ++I) { unsigned n = *I; if (nodes[n].update(nodes)) { Changed = true; if (nodes[n].preferReg()) RecentPositive.push_back(n); } } if (!Changed || !RecentPositive.empty()) return; // Scan forwards, skipping the first node which was just updated. Changed = false; for (SmallVectorImpl<unsigned>::const_iterator I = llvm::next(Linked.begin()), E = Linked.end(); I != E; ++I) { unsigned n = *I; if (nodes[n].update(nodes)) { Changed = true; if (nodes[n].preferReg()) RecentPositive.push_back(n); } } if (!Changed || !RecentPositive.empty()) return; } } void SpillPlacement::prepare(BitVector &RegBundles) { Linked.clear(); RecentPositive.clear(); // Reuse RegBundles as our ActiveNodes vector. ActiveNodes = &RegBundles; ActiveNodes->clear(); ActiveNodes->resize(bundles->getNumBundles()); } bool SpillPlacement::finish() { assert(ActiveNodes && "Call prepare() first"); // Write preferences back to ActiveNodes. bool Perfect = true; for (int n = ActiveNodes->find_first(); n>=0; n = ActiveNodes->find_next(n)) if (!nodes[n].preferReg()) { ActiveNodes->reset(n); Perfect = false; } ActiveNodes = 0; return Perfect; }