#ifndef LLVM_CODEGEN_REGALLOCPBQP_H
#define LLVM_CODEGEN_REGALLOCPBQP_H
-#include "llvm/ADT/DenseMap.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
-#include "llvm/CodeGen/PBQP/Graph.h"
-#include "llvm/CodeGen/PBQP/Solution.h"
-#include <map>
-#include <set>
+#include "llvm/CodeGen/PBQPRAConstraint.h"
+#include "llvm/CodeGen/PBQP/CostAllocator.h"
+#include "llvm/CodeGen/PBQP/ReductionRules.h"
+#include "llvm/Support/ErrorHandling.h"
namespace llvm {
-
- class LiveIntervals;
- class MachineBlockFrequencyInfo;
- class MachineFunction;
- class TargetRegisterInfo;
- template<class T> class OwningPtr;
-
- /// This class wraps up a PBQP instance representing a register allocation
- /// problem, plus the structures necessary to map back from the PBQP solution
- /// to a register allocation solution. (i.e. The PBQP-node <--> vreg map,
- /// and the PBQP option <--> storage location map).
-
- class PBQPRAProblem {
+namespace PBQP {
+namespace RegAlloc {
+
+/// @brief Spill option index.
+inline unsigned getSpillOptionIdx() { return 0; }
+
+/// \brief Metadata to speed allocatability test.
+///
+/// Keeps track of the number of infinities in each row and column.
+class MatrixMetadata {
+private:
+ MatrixMetadata(const MatrixMetadata&);
+ void operator=(const MatrixMetadata&);
+public:
+ MatrixMetadata(const Matrix& M)
+ : WorstRow(0), WorstCol(0),
+ UnsafeRows(new bool[M.getRows() - 1]()),
+ UnsafeCols(new bool[M.getCols() - 1]()) {
+
+ unsigned* ColCounts = new unsigned[M.getCols() - 1]();
+
+ for (unsigned i = 1; i < M.getRows(); ++i) {
+ unsigned RowCount = 0;
+ for (unsigned j = 1; j < M.getCols(); ++j) {
+ if (M[i][j] == std::numeric_limits<PBQPNum>::infinity()) {
+ ++RowCount;
+ ++ColCounts[j - 1];
+ UnsafeRows[i - 1] = true;
+ UnsafeCols[j - 1] = true;
+ }
+ }
+ WorstRow = std::max(WorstRow, RowCount);
+ }
+ unsigned WorstColCountForCurRow =
+ *std::max_element(ColCounts, ColCounts + M.getCols() - 1);
+ WorstCol = std::max(WorstCol, WorstColCountForCurRow);
+ delete[] ColCounts;
+ }
+
+ ~MatrixMetadata() {
+ delete[] UnsafeRows;
+ delete[] UnsafeCols;
+ }
+
+ unsigned getWorstRow() const { return WorstRow; }
+ unsigned getWorstCol() const { return WorstCol; }
+ const bool* getUnsafeRows() const { return UnsafeRows; }
+ const bool* getUnsafeCols() const { return UnsafeCols; }
+
+private:
+ unsigned WorstRow, WorstCol;
+ bool* UnsafeRows;
+ bool* UnsafeCols;
+};
+
+class NodeMetadata {
+public:
+ typedef std::vector<unsigned> OptionToRegMap;
+
+ typedef enum { Unprocessed,
+ OptimallyReducible,
+ ConservativelyAllocatable,
+ NotProvablyAllocatable } ReductionState;
+
+ NodeMetadata() : RS(Unprocessed), DeniedOpts(0), OptUnsafeEdges(nullptr){}
+ ~NodeMetadata() { delete[] OptUnsafeEdges; }
+
+ void setVReg(unsigned VReg) { this->VReg = VReg; }
+ unsigned getVReg() const { return VReg; }
+
+ void setOptionRegs(OptionToRegMap OptionRegs) {
+ this->OptionRegs = std::move(OptionRegs);
+ }
+ const OptionToRegMap& getOptionRegs() const { return OptionRegs; }
+
+ void setup(const Vector& Costs) {
+ NumOpts = Costs.getLength() - 1;
+ OptUnsafeEdges = new unsigned[NumOpts]();
+ }
+
+ ReductionState getReductionState() const { return RS; }
+ void setReductionState(ReductionState RS) { this->RS = RS; }
+
+ void handleAddEdge(const MatrixMetadata& MD, bool Transpose) {
+ DeniedOpts += Transpose ? MD.getWorstCol() : MD.getWorstRow();
+ const bool* UnsafeOpts =
+ Transpose ? MD.getUnsafeCols() : MD.getUnsafeRows();
+ for (unsigned i = 0; i < NumOpts; ++i)
+ OptUnsafeEdges[i] += UnsafeOpts[i];
+ }
+
+ void handleRemoveEdge(const MatrixMetadata& MD, bool Transpose) {
+ DeniedOpts -= Transpose ? MD.getWorstCol() : MD.getWorstRow();
+ const bool* UnsafeOpts =
+ Transpose ? MD.getUnsafeCols() : MD.getUnsafeRows();
+ for (unsigned i = 0; i < NumOpts; ++i)
+ OptUnsafeEdges[i] -= UnsafeOpts[i];
+ }
+
+ bool isConservativelyAllocatable() const {
+ return (DeniedOpts < NumOpts) ||
+ (std::find(OptUnsafeEdges, OptUnsafeEdges + NumOpts, 0) !=
+ OptUnsafeEdges + NumOpts);
+ }
+
+private:
+ ReductionState RS;
+ unsigned NumOpts;
+ unsigned DeniedOpts;
+ unsigned* OptUnsafeEdges;
+ unsigned VReg;
+ OptionToRegMap OptionRegs;
+};
+
+class RegAllocSolverImpl {
+private:
+ typedef MDMatrix<MatrixMetadata> RAMatrix;
+public:
+ typedef PBQP::Vector RawVector;
+ typedef PBQP::Matrix RawMatrix;
+ typedef PBQP::Vector Vector;
+ typedef RAMatrix Matrix;
+ typedef PBQP::PoolCostAllocator<Vector, Matrix> CostAllocator;
+
+ typedef GraphBase::NodeId NodeId;
+ typedef GraphBase::EdgeId EdgeId;
+
+ typedef RegAlloc::NodeMetadata NodeMetadata;
+
+ struct EdgeMetadata { };
+
+ class GraphMetadata {
public:
+ GraphMetadata(MachineFunction &MF,
+ LiveIntervals &LIS,
+ MachineBlockFrequencyInfo &MBFI)
+ : MF(MF), LIS(LIS), MBFI(MBFI) {}
- typedef SmallVector<unsigned, 16> AllowedSet;
+ MachineFunction &MF;
+ LiveIntervals &LIS;
+ MachineBlockFrequencyInfo &MBFI;
- PBQP::Graph& getGraph() { return graph; }
-
- const PBQP::Graph& getGraph() const { return graph; }
-
- /// Record the mapping between the given virtual register and PBQP node,
- /// and the set of allowed pregs for the vreg.
- ///
- /// If you are extending
- /// PBQPBuilder you are unlikely to need this: Nodes and options for all
- /// vregs will already have been set up for you by the base class.
- template <typename AllowedRegsItr>
- void recordVReg(unsigned vreg, PBQP::Graph::NodeId nodeId,
- AllowedRegsItr arBegin, AllowedRegsItr arEnd) {
- assert(node2VReg.find(nodeId) == node2VReg.end() && "Re-mapping node.");
- assert(vreg2Node.find(vreg) == vreg2Node.end() && "Re-mapping vreg.");
- assert(allowedSets[vreg].empty() && "vreg already has pregs.");
-
- node2VReg[nodeId] = vreg;
- vreg2Node[vreg] = nodeId;
- std::copy(arBegin, arEnd, std::back_inserter(allowedSets[vreg]));
+ void setNodeIdForVReg(unsigned VReg, GraphBase::NodeId NId) {
+ VRegToNodeId[VReg] = NId;
}
- /// Get the virtual register corresponding to the given PBQP node.
- unsigned getVRegForNode(PBQP::Graph::NodeId nodeId) const;
-
- /// Get the PBQP node corresponding to the given virtual register.
- PBQP::Graph::NodeId getNodeForVReg(unsigned vreg) const;
-
- /// Returns true if the given PBQP option represents a physical register,
- /// false otherwise.
- bool isPRegOption(unsigned vreg, unsigned option) const {
- // At present we only have spills or pregs, so anything that's not a
- // spill is a preg. (This might be extended one day to support remat).
- return !isSpillOption(vreg, option);
+ GraphBase::NodeId getNodeIdForVReg(unsigned VReg) const {
+ auto VRegItr = VRegToNodeId.find(VReg);
+ if (VRegItr == VRegToNodeId.end())
+ return GraphBase::invalidNodeId();
+ return VRegItr->second;
}
- /// Returns true if the given PBQP option represents spilling, false
- /// otherwise.
- bool isSpillOption(unsigned vreg, unsigned option) const {
- // We hardcode option zero as the spill option.
- return option == 0;
+ void eraseNodeIdForVReg(unsigned VReg) {
+ VRegToNodeId.erase(VReg);
}
- /// Returns the allowed set for the given virtual register.
- const AllowedSet& getAllowedSet(unsigned vreg) const;
-
- /// Get PReg for option.
- unsigned getPRegForOption(unsigned vreg, unsigned option) const;
-
private:
-
- typedef std::map<PBQP::Graph::NodeId, unsigned> Node2VReg;
- typedef DenseMap<unsigned, PBQP::Graph::NodeId> VReg2Node;
- typedef DenseMap<unsigned, AllowedSet> AllowedSetMap;
-
- PBQP::Graph graph;
- Node2VReg node2VReg;
- VReg2Node vreg2Node;
-
- AllowedSetMap allowedSets;
-
+ DenseMap<unsigned, NodeId> VRegToNodeId;
};
- /// Builds PBQP instances to represent register allocation problems. Includes
- /// spill, interference and coalescing costs by default. You can extend this
- /// class to support additional constraints for your architecture.
- class PBQPBuilder {
- private:
- PBQPBuilder(const PBQPBuilder&) LLVM_DELETED_FUNCTION;
- void operator=(const PBQPBuilder&) LLVM_DELETED_FUNCTION;
- public:
-
- typedef std::set<unsigned> RegSet;
-
- /// Default constructor.
- PBQPBuilder() {}
+ typedef PBQP::Graph<RegAllocSolverImpl> Graph;
+
+ RegAllocSolverImpl(Graph &G) : G(G) {}
+
+ Solution solve() {
+ G.setSolver(*this);
+ Solution S;
+ setup();
+ S = backpropagate(G, reduce());
+ G.unsetSolver();
+ return S;
+ }
+
+ void handleAddNode(NodeId NId) {
+ G.getNodeMetadata(NId).setup(G.getNodeCosts(NId));
+ }
+ void handleRemoveNode(NodeId NId) {}
+ void handleSetNodeCosts(NodeId NId, const Vector& newCosts) {}
+
+ void handleAddEdge(EdgeId EId) {
+ handleReconnectEdge(EId, G.getEdgeNode1Id(EId));
+ handleReconnectEdge(EId, G.getEdgeNode2Id(EId));
+ }
+
+ void handleRemoveEdge(EdgeId EId) {
+ handleDisconnectEdge(EId, G.getEdgeNode1Id(EId));
+ handleDisconnectEdge(EId, G.getEdgeNode2Id(EId));
+ }
+
+ void handleDisconnectEdge(EdgeId EId, NodeId NId) {
+ NodeMetadata& NMd = G.getNodeMetadata(NId);
+ const MatrixMetadata& MMd = G.getEdgeCosts(EId).getMetadata();
+ NMd.handleRemoveEdge(MMd, NId == G.getEdgeNode2Id(EId));
+ if (G.getNodeDegree(NId) == 3) {
+ // This node is becoming optimally reducible.
+ moveToOptimallyReducibleNodes(NId);
+ } else if (NMd.getReductionState() ==
+ NodeMetadata::NotProvablyAllocatable &&
+ NMd.isConservativelyAllocatable()) {
+ // This node just became conservatively allocatable.
+ moveToConservativelyAllocatableNodes(NId);
+ }
+ }
+
+ void handleReconnectEdge(EdgeId EId, NodeId NId) {
+ NodeMetadata& NMd = G.getNodeMetadata(NId);
+ const MatrixMetadata& MMd = G.getEdgeCosts(EId).getMetadata();
+ NMd.handleAddEdge(MMd, NId == G.getEdgeNode2Id(EId));
+ }
+
+ void handleSetEdgeCosts(EdgeId EId, const Matrix& NewCosts) {
+ handleRemoveEdge(EId);
+
+ NodeId N1Id = G.getEdgeNode1Id(EId);
+ NodeId N2Id = G.getEdgeNode2Id(EId);
+ NodeMetadata& N1Md = G.getNodeMetadata(N1Id);
+ NodeMetadata& N2Md = G.getNodeMetadata(N2Id);
+ const MatrixMetadata& MMd = NewCosts.getMetadata();
+ N1Md.handleAddEdge(MMd, N1Id != G.getEdgeNode1Id(EId));
+ N2Md.handleAddEdge(MMd, N2Id != G.getEdgeNode1Id(EId));
+ }
+
+private:
+
+ void removeFromCurrentSet(NodeId NId) {
+ switch (G.getNodeMetadata(NId).getReductionState()) {
+ case NodeMetadata::Unprocessed: break;
+ case NodeMetadata::OptimallyReducible:
+ assert(OptimallyReducibleNodes.find(NId) !=
+ OptimallyReducibleNodes.end() &&
+ "Node not in optimally reducible set.");
+ OptimallyReducibleNodes.erase(NId);
+ break;
+ case NodeMetadata::ConservativelyAllocatable:
+ assert(ConservativelyAllocatableNodes.find(NId) !=
+ ConservativelyAllocatableNodes.end() &&
+ "Node not in conservatively allocatable set.");
+ ConservativelyAllocatableNodes.erase(NId);
+ break;
+ case NodeMetadata::NotProvablyAllocatable:
+ assert(NotProvablyAllocatableNodes.find(NId) !=
+ NotProvablyAllocatableNodes.end() &&
+ "Node not in not-provably-allocatable set.");
+ NotProvablyAllocatableNodes.erase(NId);
+ break;
+ }
+ }
+
+ void moveToOptimallyReducibleNodes(NodeId NId) {
+ removeFromCurrentSet(NId);
+ OptimallyReducibleNodes.insert(NId);
+ G.getNodeMetadata(NId).setReductionState(
+ NodeMetadata::OptimallyReducible);
+ }
+
+ void moveToConservativelyAllocatableNodes(NodeId NId) {
+ removeFromCurrentSet(NId);
+ ConservativelyAllocatableNodes.insert(NId);
+ G.getNodeMetadata(NId).setReductionState(
+ NodeMetadata::ConservativelyAllocatable);
+ }
+
+ void moveToNotProvablyAllocatableNodes(NodeId NId) {
+ removeFromCurrentSet(NId);
+ NotProvablyAllocatableNodes.insert(NId);
+ G.getNodeMetadata(NId).setReductionState(
+ NodeMetadata::NotProvablyAllocatable);
+ }
+
+ void setup() {
+ // Set up worklists.
+ for (auto NId : G.nodeIds()) {
+ if (G.getNodeDegree(NId) < 3)
+ moveToOptimallyReducibleNodes(NId);
+ else if (G.getNodeMetadata(NId).isConservativelyAllocatable())
+ moveToConservativelyAllocatableNodes(NId);
+ else
+ moveToNotProvablyAllocatableNodes(NId);
+ }
+ }
+
+ // Compute a reduction order for the graph by iteratively applying PBQP
+ // reduction rules. Locally optimal rules are applied whenever possible (R0,
+ // R1, R2). If no locally-optimal rules apply then any conservatively
+ // allocatable node is reduced. Finally, if no conservatively allocatable
+ // node exists then the node with the lowest spill-cost:degree ratio is
+ // selected.
+ std::vector<GraphBase::NodeId> reduce() {
+ assert(!G.empty() && "Cannot reduce empty graph.");
+
+ typedef GraphBase::NodeId NodeId;
+ std::vector<NodeId> NodeStack;
+
+ // Consume worklists.
+ while (true) {
+ if (!OptimallyReducibleNodes.empty()) {
+ NodeSet::iterator NItr = OptimallyReducibleNodes.begin();
+ NodeId NId = *NItr;
+ OptimallyReducibleNodes.erase(NItr);
+ NodeStack.push_back(NId);
+ switch (G.getNodeDegree(NId)) {
+ case 0:
+ break;
+ case 1:
+ applyR1(G, NId);
+ break;
+ case 2:
+ applyR2(G, NId);
+ break;
+ default: llvm_unreachable("Not an optimally reducible node.");
+ }
+ } else if (!ConservativelyAllocatableNodes.empty()) {
+ // Conservatively allocatable nodes will never spill. For now just
+ // take the first node in the set and push it on the stack. When we
+ // start optimizing more heavily for register preferencing, it may
+ // would be better to push nodes with lower 'expected' or worst-case
+ // register costs first (since early nodes are the most
+ // constrained).
+ NodeSet::iterator NItr = ConservativelyAllocatableNodes.begin();
+ NodeId NId = *NItr;
+ ConservativelyAllocatableNodes.erase(NItr);
+ NodeStack.push_back(NId);
+ G.disconnectAllNeighborsFromNode(NId);
+
+ } else if (!NotProvablyAllocatableNodes.empty()) {
+ NodeSet::iterator NItr =
+ std::min_element(NotProvablyAllocatableNodes.begin(),
+ NotProvablyAllocatableNodes.end(),
+ SpillCostComparator(G));
+ NodeId NId = *NItr;
+ NotProvablyAllocatableNodes.erase(NItr);
+ NodeStack.push_back(NId);
+ G.disconnectAllNeighborsFromNode(NId);
+ } else
+ break;
+ }
- /// Clean up a PBQPBuilder.
- virtual ~PBQPBuilder() {}
+ return NodeStack;
+ }
- /// Build a PBQP instance to represent the register allocation problem for
- /// the given MachineFunction.
- virtual PBQPRAProblem *build(MachineFunction *mf, const LiveIntervals *lis,
- const MachineBlockFrequencyInfo *mbfi,
- const RegSet &vregs);
+ class SpillCostComparator {
+ public:
+ SpillCostComparator(const Graph& G) : G(G) {}
+ bool operator()(NodeId N1Id, NodeId N2Id) {
+ PBQPNum N1SC = G.getNodeCosts(N1Id)[0] / G.getNodeDegree(N1Id);
+ PBQPNum N2SC = G.getNodeCosts(N2Id)[0] / G.getNodeDegree(N2Id);
+ return N1SC < N2SC;
+ }
private:
-
- void addSpillCosts(PBQP::Vector &costVec, PBQP::PBQPNum spillCost);
-
- void addInterferenceCosts(PBQP::Matrix &costMat,
- const PBQPRAProblem::AllowedSet &vr1Allowed,
- const PBQPRAProblem::AllowedSet &vr2Allowed,
- const TargetRegisterInfo *tri);
+ const Graph& G;
};
- /// Extended builder which adds coalescing constraints to a problem.
- class PBQPBuilderWithCoalescing : public PBQPBuilder {
- public:
-
- /// Build a PBQP instance to represent the register allocation problem for
- /// the given MachineFunction.
- virtual PBQPRAProblem *build(MachineFunction *mf, const LiveIntervals *lis,
- const MachineBlockFrequencyInfo *mbfi,
- const RegSet &vregs);
-
- private:
+ Graph& G;
+ typedef std::set<NodeId> NodeSet;
+ NodeSet OptimallyReducibleNodes;
+ NodeSet ConservativelyAllocatableNodes;
+ NodeSet NotProvablyAllocatableNodes;
+};
+
+class PBQPRAGraph : public PBQP::Graph<RegAllocSolverImpl> {
+private:
+ typedef PBQP::Graph<RegAllocSolverImpl> BaseT;
+public:
+ PBQPRAGraph(GraphMetadata Metadata) : BaseT(Metadata) {}
+};
+
+inline Solution solve(PBQPRAGraph& G) {
+ if (G.empty())
+ return Solution();
+ RegAllocSolverImpl RegAllocSolver(G);
+ return RegAllocSolver.solve();
+}
- void addPhysRegCoalesce(PBQP::Vector &costVec, unsigned pregOption,
- PBQP::PBQPNum benefit);
+} // namespace RegAlloc
+} // namespace PBQP
- void addVirtRegCoalesce(PBQP::Matrix &costMat,
- const PBQPRAProblem::AllowedSet &vr1Allowed,
- const PBQPRAProblem::AllowedSet &vr2Allowed,
- PBQP::PBQPNum benefit);
- };
+/// @brief Create a PBQP register allocator instance.
+FunctionPass *
+createPBQPRegisterAllocator(char *customPassID = nullptr);
- FunctionPass* createPBQPRegisterAllocator(OwningPtr<PBQPBuilder> &builder,
- char *customPassID=0);
-}
+} // namespace llvm
#endif /* LLVM_CODEGEN_REGALLOCPBQP_H */
projects / oota-llvm.git / blobdiff