llvm-6502/lib/Target/SparcV9/ModuloScheduling/ModuloScheduling.cpp

2965 lines
97 KiB
C++

//===-- ModuloScheduling.cpp - ModuloScheduling ----------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by the LLVM research group and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This ModuloScheduling pass is based on the Swing Modulo Scheduling
// algorithm.
//
//===----------------------------------------------------------------------===//
#define DEBUG_TYPE "ModuloSched"
#include "ModuloScheduling.h"
#include "llvm/Constants.h"
#include "llvm/Instructions.h"
#include "llvm/Function.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/Passes.h"
#include "llvm/Support/CFG.h"
#include "llvm/Target/TargetSchedInfo.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/GraphWriter.h"
#include "llvm/ADT/SCCIterator.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Support/Timer.h"
#include <cmath>
#include <algorithm>
#include <fstream>
#include <sstream>
#include <utility>
#include <vector>
#include "../MachineCodeForInstruction.h"
#include "../SparcV9TmpInstr.h"
#include "../SparcV9Internals.h"
#include "../SparcV9RegisterInfo.h"
using namespace llvm;
/// Create ModuloSchedulingPass
///
FunctionPass *llvm::createModuloSchedulingPass(TargetMachine & targ) {
DEBUG(std::cerr << "Created ModuloSchedulingPass\n");
return new ModuloSchedulingPass(targ);
}
//Graph Traits for printing out the dependence graph
template<typename GraphType>
static void WriteGraphToFile(std::ostream &O, const std::string &GraphName,
const GraphType &GT) {
std::string Filename = GraphName + ".dot";
O << "Writing '" << Filename << "'...";
std::ofstream F(Filename.c_str());
if (F.good())
WriteGraph(F, GT);
else
O << " error opening file for writing!";
O << "\n";
};
#if 1
#define TIME_REGION(VARNAME, DESC) \
NamedRegionTimer VARNAME(DESC)
#else
#define TIME_REGION(VARNAME, DESC)
#endif
//Graph Traits for printing out the dependence graph
namespace llvm {
//Loop statistics
Statistic<> ValidLoops("modulosched-validLoops", "Number of candidate loops modulo-scheduled");
Statistic<> JumboBB("modulosched-jumboBB", "Basic Blocks with more then 100 instructions");
Statistic<> LoopsWithCalls("modulosched-loopCalls", "Loops with calls");
Statistic<> LoopsWithCondMov("modulosched-loopCondMov", "Loops with conditional moves");
Statistic<> InvalidLoops("modulosched-invalidLoops", "Loops with unknown trip counts or loop invariant trip counts");
Statistic<> SingleBBLoops("modulosched-singeBBLoops", "Number of single basic block loops");
//Scheduling Statistics
Statistic<> MSLoops("modulosched-schedLoops", "Number of loops successfully modulo-scheduled");
Statistic<> NoSched("modulosched-noSched", "No schedule");
Statistic<> SameStage("modulosched-sameStage", "Max stage is 0");
Statistic<> ResourceConstraint("modulosched-resourceConstraint", "Loops constrained by resources");
Statistic<> RecurrenceConstraint("modulosched-recurrenceConstraint", "Loops constrained by recurrences");
Statistic<> FinalIISum("modulosched-finalIISum", "Sum of all final II");
Statistic<> IISum("modulosched-IISum", "Sum of all theoretical II");
template<>
struct DOTGraphTraits<MSchedGraph*> : public DefaultDOTGraphTraits {
static std::string getGraphName(MSchedGraph *F) {
return "Dependence Graph";
}
static std::string getNodeLabel(MSchedGraphNode *Node, MSchedGraph *Graph) {
if (Node->getInst()) {
std::stringstream ss;
ss << *(Node->getInst());
return ss.str(); //((MachineInstr*)Node->getInst());
}
else
return "No Inst";
}
static std::string getEdgeSourceLabel(MSchedGraphNode *Node,
MSchedGraphNode::succ_iterator I) {
//Label each edge with the type of dependence
std::string edgelabel = "";
switch (I.getEdge().getDepOrderType()) {
case MSchedGraphEdge::TrueDep:
edgelabel = "True";
break;
case MSchedGraphEdge::AntiDep:
edgelabel = "Anti";
break;
case MSchedGraphEdge::OutputDep:
edgelabel = "Output";
break;
default:
edgelabel = "Unknown";
break;
}
//FIXME
int iteDiff = I.getEdge().getIteDiff();
std::string intStr = "(IteDiff: ";
intStr += itostr(iteDiff);
intStr += ")";
edgelabel += intStr;
return edgelabel;
}
};
}
#include <unistd.h>
/// ModuloScheduling::runOnFunction - main transformation entry point
/// The Swing Modulo Schedule algorithm has three basic steps:
/// 1) Computation and Analysis of the dependence graph
/// 2) Ordering of the nodes
/// 3) Scheduling
///
bool ModuloSchedulingPass::runOnFunction(Function &F) {
alarm(100);
bool Changed = false;
int numMS = 0;
DEBUG(std::cerr << "Creating ModuloSchedGraph for each valid BasicBlock in " + F.getName() + "\n");
//Get MachineFunction
MachineFunction &MF = MachineFunction::get(&F);
DependenceAnalyzer &DA = getAnalysis<DependenceAnalyzer>();
//Worklist
std::vector<MachineBasicBlock*> Worklist;
//Iterate over BasicBlocks and put them into our worklist if they are valid
for (MachineFunction::iterator BI = MF.begin(); BI != MF.end(); ++BI)
if(MachineBBisValid(BI)) {
if(BI->size() < 100) {
Worklist.push_back(&*BI);
++ValidLoops;
}
else
++JumboBB;
}
defaultInst = 0;
DEBUG(if(Worklist.size() == 0) std::cerr << "No single basic block loops in function to ModuloSchedule\n");
//Iterate over the worklist and perform scheduling
for(std::vector<MachineBasicBlock*>::iterator BI = Worklist.begin(),
BE = Worklist.end(); BI != BE; ++BI) {
//Print out BB for debugging
DEBUG(std::cerr << "BB Size: " << (*BI)->size() << "\n");
DEBUG(std::cerr << "ModuloScheduling BB: \n"; (*BI)->print(std::cerr));
//Print out LLVM BB
DEBUG(std::cerr << "ModuloScheduling LLVMBB: \n"; (*BI)->getBasicBlock()->print(std::cerr));
//Catch the odd case where we only have TmpInstructions and no real Value*s
if(!CreateDefMap(*BI)) {
//Clear out our maps for the next basic block that is processed
nodeToAttributesMap.clear();
partialOrder.clear();
recurrenceList.clear();
FinalNodeOrder.clear();
schedule.clear();
defMap.clear();
continue;
}
MSchedGraph *MSG = new MSchedGraph(*BI, target, indVarInstrs[*BI], DA, machineTollvm[*BI]);
//Write Graph out to file
DEBUG(WriteGraphToFile(std::cerr, F.getName(), MSG));
DEBUG(MSG->print(std::cerr));
//Calculate Resource II
int ResMII = calculateResMII(*BI);
//Calculate Recurrence II
int RecMII = calculateRecMII(MSG, ResMII);
DEBUG(std::cerr << "Number of reccurrences found: " << recurrenceList.size() << "\n");
//Our starting initiation interval is the maximum of RecMII and ResMII
if(RecMII < ResMII)
++RecurrenceConstraint;
else
++ResourceConstraint;
II = std::max(RecMII, ResMII);
int mII = II;
//Print out II, RecMII, and ResMII
DEBUG(std::cerr << "II starts out as " << II << " ( RecMII=" << RecMII << " and ResMII=" << ResMII << ")\n");
//Dump node properties if in debug mode
DEBUG(for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(),
E = nodeToAttributesMap.end(); I !=E; ++I) {
std::cerr << "Node: " << *(I->first) << " ASAP: " << I->second.ASAP << " ALAP: "
<< I->second.ALAP << " MOB: " << I->second.MOB << " Depth: " << I->second.depth
<< " Height: " << I->second.height << "\n";
});
//Calculate Node Properties
calculateNodeAttributes(MSG, ResMII);
//Dump node properties if in debug mode
DEBUG(for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(),
E = nodeToAttributesMap.end(); I !=E; ++I) {
std::cerr << "Node: " << *(I->first) << " ASAP: " << I->second.ASAP << " ALAP: "
<< I->second.ALAP << " MOB: " << I->second.MOB << " Depth: " << I->second.depth
<< " Height: " << I->second.height << "\n";
});
//Put nodes in order to schedule them
computePartialOrder();
//Dump out partial order
DEBUG(for(std::vector<std::set<MSchedGraphNode*> >::iterator I = partialOrder.begin(),
E = partialOrder.end(); I !=E; ++I) {
std::cerr << "Start set in PO\n";
for(std::set<MSchedGraphNode*>::iterator J = I->begin(), JE = I->end(); J != JE; ++J)
std::cerr << "PO:" << **J << "\n";
});
//Place nodes in final order
orderNodes();
//Dump out order of nodes
DEBUG(for(std::vector<MSchedGraphNode*>::iterator I = FinalNodeOrder.begin(), E = FinalNodeOrder.end(); I != E; ++I) {
std::cerr << "FO:" << **I << "\n";
});
//Finally schedule nodes
bool haveSched = computeSchedule(*BI, MSG);
//Print out final schedule
DEBUG(schedule.print(std::cerr));
//Final scheduling step is to reconstruct the loop only if we actual have
//stage > 0
if(haveSched) {
reconstructLoop(*BI);
++MSLoops;
Changed = true;
FinalIISum += II;
IISum += mII;
if(schedule.getMaxStage() == 0)
++SameStage;
}
else {
++NoSched;
}
//Clear out our maps for the next basic block that is processed
nodeToAttributesMap.clear();
partialOrder.clear();
recurrenceList.clear();
FinalNodeOrder.clear();
schedule.clear();
defMap.clear();
//Clean up. Nuke old MachineBB and llvmBB
//BasicBlock *llvmBB = (BasicBlock*) (*BI)->getBasicBlock();
//Function *parent = (Function*) llvmBB->getParent();
//Should't std::find work??
//parent->getBasicBlockList().erase(std::find(parent->getBasicBlockList().begin(), parent->getBasicBlockList().end(), *llvmBB));
//parent->getBasicBlockList().erase(llvmBB);
//delete(llvmBB);
//delete(*BI);
}
alarm(0);
return Changed;
}
bool ModuloSchedulingPass::CreateDefMap(MachineBasicBlock *BI) {
defaultInst = 0;
for(MachineBasicBlock::iterator I = BI->begin(), E = BI->end(); I != E; ++I) {
for(unsigned opNum = 0; opNum < I->getNumOperands(); ++opNum) {
const MachineOperand &mOp = I->getOperand(opNum);
if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isDef()) {
//assert if this is the second def we have seen
//DEBUG(std::cerr << "Putting " << *(mOp.getVRegValue()) << " into map\n");
//assert(!defMap.count(mOp.getVRegValue()) && "Def already in the map");
if(defMap.count(mOp.getVRegValue()))
return false;
defMap[mOp.getVRegValue()] = &*I;
}
//See if we can use this Value* as our defaultInst
if(!defaultInst && mOp.getType() == MachineOperand::MO_VirtualRegister) {
Value *V = mOp.getVRegValue();
if(!isa<TmpInstruction>(V) && !isa<Argument>(V) && !isa<Constant>(V) && !isa<PHINode>(V))
defaultInst = (Instruction*) V;
}
}
}
if(!defaultInst)
return false;
return true;
}
/// This function checks if a Machine Basic Block is valid for modulo
/// scheduling. This means that it has no control flow (if/else or
/// calls) in the block. Currently ModuloScheduling only works on
/// single basic block loops.
bool ModuloSchedulingPass::MachineBBisValid(const MachineBasicBlock *BI) {
bool isLoop = false;
//Check first if its a valid loop
for(succ_const_iterator I = succ_begin(BI->getBasicBlock()),
E = succ_end(BI->getBasicBlock()); I != E; ++I) {
if (*I == BI->getBasicBlock()) // has single block loop
isLoop = true;
}
if(!isLoop)
return false;
//Check that we have a conditional branch (avoiding MS infinite loops)
if(BranchInst *b = dyn_cast<BranchInst>(((BasicBlock*) BI->getBasicBlock())->getTerminator()))
if(b->isUnconditional())
return false;
//Check size of our basic block.. make sure we have more then just the terminator in it
if(BI->getBasicBlock()->size() == 1)
return false;
//Increase number of single basic block loops for stats
++SingleBBLoops;
//Get Target machine instruction info
const TargetInstrInfo *TMI = target.getInstrInfo();
//Check each instruction and look for calls, keep map to get index later
std::map<const MachineInstr*, unsigned> indexMap;
unsigned count = 0;
for(MachineBasicBlock::const_iterator I = BI->begin(), E = BI->end(); I != E; ++I) {
//Get opcode to check instruction type
MachineOpCode OC = I->getOpcode();
//Look for calls
if(TMI->isCall(OC)) {
++LoopsWithCalls;
return false;
}
//Look for conditional move
if(OC == V9::MOVRZr || OC == V9::MOVRZi || OC == V9::MOVRLEZr || OC == V9::MOVRLEZi
|| OC == V9::MOVRLZr || OC == V9::MOVRLZi || OC == V9::MOVRNZr || OC == V9::MOVRNZi
|| OC == V9::MOVRGZr || OC == V9::MOVRGZi || OC == V9::MOVRGEZr
|| OC == V9::MOVRGEZi || OC == V9::MOVLEr || OC == V9::MOVLEi || OC == V9::MOVLEUr
|| OC == V9::MOVLEUi || OC == V9::MOVFLEr || OC == V9::MOVFLEi
|| OC == V9::MOVNEr || OC == V9::MOVNEi || OC == V9::MOVNEGr || OC == V9::MOVNEGi
|| OC == V9::MOVFNEr || OC == V9::MOVFNEi || OC == V9::MOVGr || OC == V9::MOVGi) {
++LoopsWithCondMov;
return false;
}
indexMap[I] = count;
if(TMI->isNop(OC))
continue;
++count;
}
//Apply a simple pattern match to make sure this loop can be modulo scheduled
//This means only loops with a branch associated to the iteration count
//Get the branch
BranchInst *b = dyn_cast<BranchInst>(((BasicBlock*) BI->getBasicBlock())->getTerminator());
//Get the condition for the branch (we already checked if it was conditional)
Value *cond = b->getCondition();
DEBUG(std::cerr << "Condition: " << *cond << "\n");
//List of instructions associated with induction variable
std::set<Instruction*> indVar;
std::vector<Instruction*> stack;
BasicBlock *BB = (BasicBlock*) BI->getBasicBlock();
//Add branch
indVar.insert(b);
if(Instruction *I = dyn_cast<Instruction>(cond))
if(I->getParent() == BB) {
if (!assocIndVar(I, indVar, stack, BB)) {
++InvalidLoops;
return false;
}
}
else {
++InvalidLoops;
return false;
}
else {
++InvalidLoops;
return false;
}
//The indVar set must be >= 3 instructions for this loop to match (FIX ME!)
if(indVar.size() < 3 )
return false;
//Dump out instructions associate with indvar for debug reasons
DEBUG(for(std::set<Instruction*>::iterator N = indVar.begin(), NE = indVar.end(); N != NE; ++N) {
std::cerr << **N << "\n";
});
//Create map of machine instr to llvm instr
std::map<MachineInstr*, Instruction*> mllvm;
for(BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E; ++I) {
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(I);
for (unsigned j = 0; j < tempMvec.size(); j++) {
mllvm[tempMvec[j]] = I;
}
}
//Convert list of LLVM Instructions to list of Machine instructions
std::map<const MachineInstr*, unsigned> mIndVar;
for(std::set<Instruction*>::iterator N = indVar.begin(), NE = indVar.end(); N != NE; ++N) {
//If we have a load, we can't handle this loop because there is no way to preserve dependences
//between loads and stores
if(isa<LoadInst>(*N))
return false;
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(*N);
for (unsigned j = 0; j < tempMvec.size(); j++) {
MachineOpCode OC = (tempMvec[j])->getOpcode();
if(TMI->isNop(OC))
continue;
if(!indexMap.count(tempMvec[j]))
continue;
mIndVar[(MachineInstr*) tempMvec[j]] = indexMap[(MachineInstr*) tempMvec[j]];
DEBUG(std::cerr << *(tempMvec[j]) << " at index " << indexMap[(MachineInstr*) tempMvec[j]] << "\n");
}
}
//Must have some guts to the loop body (more then 1 instr, dont count nops in size)
if(mIndVar.size() >= (BI->size()-3))
return false;
//Put into a map for future access
indVarInstrs[BI] = mIndVar;
machineTollvm[BI] = mllvm;
return true;
}
bool ModuloSchedulingPass::assocIndVar(Instruction *I, std::set<Instruction*> &indVar,
std::vector<Instruction*> &stack, BasicBlock *BB) {
stack.push_back(I);
//If this is a phi node, check if its the canonical indvar
if(PHINode *PN = dyn_cast<PHINode>(I)) {
if (Instruction *Inc =
dyn_cast<Instruction>(PN->getIncomingValueForBlock(BB)))
if (Inc->getOpcode() == Instruction::Add && Inc->getOperand(0) == PN)
if (ConstantInt *CI = dyn_cast<ConstantInt>(Inc->getOperand(1)))
if (CI->equalsInt(1)) {
//We have found the indvar, so add the stack, and inc instruction to the set
indVar.insert(stack.begin(), stack.end());
indVar.insert(Inc);
stack.pop_back();
return true;
}
return false;
}
else {
//Loop over each of the instructions operands, check if they are an instruction and in this BB
for(unsigned i = 0; i < I->getNumOperands(); ++i) {
if(Instruction *N = dyn_cast<Instruction>(I->getOperand(i))) {
if(N->getParent() == BB)
if(!assocIndVar(N, indVar, stack, BB))
return false;
}
}
}
stack.pop_back();
return true;
}
//ResMII is calculated by determining the usage count for each resource
//and using the maximum.
//FIXME: In future there should be a way to get alternative resources
//for each instruction
int ModuloSchedulingPass::calculateResMII(const MachineBasicBlock *BI) {
TIME_REGION(X, "calculateResMII");
const TargetInstrInfo *mii = target.getInstrInfo();
const TargetSchedInfo *msi = target.getSchedInfo();
int ResMII = 0;
//Map to keep track of usage count of each resource
std::map<unsigned, unsigned> resourceUsageCount;
for(MachineBasicBlock::const_iterator I = BI->begin(), E = BI->end(); I != E; ++I) {
//Get resource usage for this instruction
InstrRUsage rUsage = msi->getInstrRUsage(I->getOpcode());
std::vector<std::vector<resourceId_t> > resources = rUsage.resourcesByCycle;
//Loop over resources in each cycle and increments their usage count
for(unsigned i=0; i < resources.size(); ++i)
for(unsigned j=0; j < resources[i].size(); ++j) {
if(!resourceUsageCount.count(resources[i][j])) {
resourceUsageCount[resources[i][j]] = 1;
}
else {
resourceUsageCount[resources[i][j]] = resourceUsageCount[resources[i][j]] + 1;
}
}
}
//Find maximum usage count
//Get max number of instructions that can be issued at once. (FIXME)
int issueSlots = msi->maxNumIssueTotal;
for(std::map<unsigned,unsigned>::iterator RB = resourceUsageCount.begin(), RE = resourceUsageCount.end(); RB != RE; ++RB) {
//Get the total number of the resources in our cpu
int resourceNum = CPUResource::getCPUResource(RB->first)->maxNumUsers;
//Get total usage count for this resources
unsigned usageCount = RB->second;
//Divide the usage count by either the max number we can issue or the number of
//resources (whichever is its upper bound)
double finalUsageCount;
DEBUG(std::cerr << "Resource Num: " << RB->first << " Usage: " << usageCount << " TotalNum: " << resourceNum << "\n");
if( resourceNum <= issueSlots)
finalUsageCount = ceil(1.0 * usageCount / resourceNum);
else
finalUsageCount = ceil(1.0 * usageCount / issueSlots);
//Only keep track of the max
ResMII = std::max( (int) finalUsageCount, ResMII);
}
return ResMII;
}
/// calculateRecMII - Calculates the value of the highest recurrence
/// By value we mean the total latency
int ModuloSchedulingPass::calculateRecMII(MSchedGraph *graph, int MII) {
/*std::vector<MSchedGraphNode*> vNodes;
//Loop over all nodes in the graph
for(MSchedGraph::iterator I = graph->begin(), E = graph->end(); I != E; ++I) {
findAllReccurrences(I->second, vNodes, MII);
vNodes.clear();
}*/
TIME_REGION(X, "calculateRecMII");
findAllCircuits(graph, MII);
int RecMII = 0;
for(std::set<std::pair<int, std::vector<MSchedGraphNode*> > >::iterator I = recurrenceList.begin(), E=recurrenceList.end(); I !=E; ++I) {
RecMII = std::max(RecMII, I->first);
}
return MII;
}
/// calculateNodeAttributes - The following properties are calculated for
/// each node in the dependence graph: ASAP, ALAP, Depth, Height, and
/// MOB.
void ModuloSchedulingPass::calculateNodeAttributes(MSchedGraph *graph, int MII) {
TIME_REGION(X, "calculateNodeAttributes");
assert(nodeToAttributesMap.empty() && "Node attribute map was not cleared");
//Loop over the nodes and add them to the map
for(MSchedGraph::iterator I = graph->begin(), E = graph->end(); I != E; ++I) {
DEBUG(std::cerr << "Inserting node into attribute map: " << *I->second << "\n");
//Assert if its already in the map
assert(nodeToAttributesMap.count(I->second) == 0 &&
"Node attributes are already in the map");
//Put into the map with default attribute values
nodeToAttributesMap[I->second] = MSNodeAttributes();
}
//Create set to deal with reccurrences
std::set<MSchedGraphNode*> visitedNodes;
//Now Loop over map and calculate the node attributes
for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(), E = nodeToAttributesMap.end(); I != E; ++I) {
calculateASAP(I->first, MII, (MSchedGraphNode*) 0);
visitedNodes.clear();
}
int maxASAP = findMaxASAP();
//Calculate ALAP which depends on ASAP being totally calculated
for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(), E = nodeToAttributesMap.end(); I != E; ++I) {
calculateALAP(I->first, MII, maxASAP, (MSchedGraphNode*) 0);
visitedNodes.clear();
}
//Calculate MOB which depends on ASAP being totally calculated, also do depth and height
for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(), E = nodeToAttributesMap.end(); I != E; ++I) {
(I->second).MOB = std::max(0,(I->second).ALAP - (I->second).ASAP);
DEBUG(std::cerr << "MOB: " << (I->second).MOB << " (" << *(I->first) << ")\n");
calculateDepth(I->first, (MSchedGraphNode*) 0);
calculateHeight(I->first, (MSchedGraphNode*) 0);
}
}
/// ignoreEdge - Checks to see if this edge of a recurrence should be ignored or not
bool ModuloSchedulingPass::ignoreEdge(MSchedGraphNode *srcNode, MSchedGraphNode *destNode) {
if(destNode == 0 || srcNode ==0)
return false;
bool findEdge = edgesToIgnore.count(std::make_pair(srcNode, destNode->getInEdgeNum(srcNode)));
DEBUG(std::cerr << "Ignoring edge? from: " << *srcNode << " to " << *destNode << "\n");
return findEdge;
}
/// calculateASAP - Calculates the
int ModuloSchedulingPass::calculateASAP(MSchedGraphNode *node, int MII, MSchedGraphNode *destNode) {
DEBUG(std::cerr << "Calculating ASAP for " << *node << "\n");
//Get current node attributes
MSNodeAttributes &attributes = nodeToAttributesMap.find(node)->second;
if(attributes.ASAP != -1)
return attributes.ASAP;
int maxPredValue = 0;
//Iterate over all of the predecessors and find max
for(MSchedGraphNode::pred_iterator P = node->pred_begin(), E = node->pred_end(); P != E; ++P) {
//Only process if we are not ignoring the edge
if(!ignoreEdge(*P, node)) {
int predASAP = -1;
predASAP = calculateASAP(*P, MII, node);
assert(predASAP != -1 && "ASAP has not been calculated");
int iteDiff = node->getInEdge(*P).getIteDiff();
int currentPredValue = predASAP + (*P)->getLatency() - (iteDiff * MII);
DEBUG(std::cerr << "pred ASAP: " << predASAP << ", iteDiff: " << iteDiff << ", PredLatency: " << (*P)->getLatency() << ", Current ASAP pred: " << currentPredValue << "\n");
maxPredValue = std::max(maxPredValue, currentPredValue);
}
}
attributes.ASAP = maxPredValue;
DEBUG(std::cerr << "ASAP: " << attributes.ASAP << " (" << *node << ")\n");
return maxPredValue;
}
int ModuloSchedulingPass::calculateALAP(MSchedGraphNode *node, int MII,
int maxASAP, MSchedGraphNode *srcNode) {
DEBUG(std::cerr << "Calculating ALAP for " << *node << "\n");
MSNodeAttributes &attributes = nodeToAttributesMap.find(node)->second;
if(attributes.ALAP != -1)
return attributes.ALAP;
if(node->hasSuccessors()) {
//Trying to deal with the issue where the node has successors, but
//we are ignoring all of the edges to them. So this is my hack for
//now.. there is probably a more elegant way of doing this (FIXME)
bool processedOneEdge = false;
//FIXME, set to something high to start
int minSuccValue = 9999999;
//Iterate over all of the predecessors and fine max
for(MSchedGraphNode::succ_iterator P = node->succ_begin(),
E = node->succ_end(); P != E; ++P) {
//Only process if we are not ignoring the edge
if(!ignoreEdge(node, *P)) {
processedOneEdge = true;
int succALAP = -1;
succALAP = calculateALAP(*P, MII, maxASAP, node);
assert(succALAP != -1 && "Successors ALAP should have been caclulated");
int iteDiff = P.getEdge().getIteDiff();
int currentSuccValue = succALAP - node->getLatency() + iteDiff * MII;
DEBUG(std::cerr << "succ ALAP: " << succALAP << ", iteDiff: " << iteDiff << ", SuccLatency: " << (*P)->getLatency() << ", Current ALAP succ: " << currentSuccValue << "\n");
minSuccValue = std::min(minSuccValue, currentSuccValue);
}
}
if(processedOneEdge)
attributes.ALAP = minSuccValue;
else
attributes.ALAP = maxASAP;
}
else
attributes.ALAP = maxASAP;
DEBUG(std::cerr << "ALAP: " << attributes.ALAP << " (" << *node << ")\n");
if(attributes.ALAP < 0)
attributes.ALAP = 0;
return attributes.ALAP;
}
int ModuloSchedulingPass::findMaxASAP() {
int maxASAP = 0;
for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(),
E = nodeToAttributesMap.end(); I != E; ++I)
maxASAP = std::max(maxASAP, I->second.ASAP);
return maxASAP;
}
int ModuloSchedulingPass::calculateHeight(MSchedGraphNode *node,MSchedGraphNode *srcNode) {
MSNodeAttributes &attributes = nodeToAttributesMap.find(node)->second;
if(attributes.height != -1)
return attributes.height;
int maxHeight = 0;
//Iterate over all of the predecessors and find max
for(MSchedGraphNode::succ_iterator P = node->succ_begin(),
E = node->succ_end(); P != E; ++P) {
if(!ignoreEdge(node, *P)) {
int succHeight = calculateHeight(*P, node);
assert(succHeight != -1 && "Successors Height should have been caclulated");
int currentHeight = succHeight + node->getLatency();
maxHeight = std::max(maxHeight, currentHeight);
}
}
attributes.height = maxHeight;
DEBUG(std::cerr << "Height: " << attributes.height << " (" << *node << ")\n");
return maxHeight;
}
int ModuloSchedulingPass::calculateDepth(MSchedGraphNode *node,
MSchedGraphNode *destNode) {
MSNodeAttributes &attributes = nodeToAttributesMap.find(node)->second;
if(attributes.depth != -1)
return attributes.depth;
int maxDepth = 0;
//Iterate over all of the predecessors and fine max
for(MSchedGraphNode::pred_iterator P = node->pred_begin(), E = node->pred_end(); P != E; ++P) {
if(!ignoreEdge(*P, node)) {
int predDepth = -1;
predDepth = calculateDepth(*P, node);
assert(predDepth != -1 && "Predecessors ASAP should have been caclulated");
int currentDepth = predDepth + (*P)->getLatency();
maxDepth = std::max(maxDepth, currentDepth);
}
}
attributes.depth = maxDepth;
DEBUG(std::cerr << "Depth: " << attributes.depth << " (" << *node << "*)\n");
return maxDepth;
}
void ModuloSchedulingPass::addReccurrence(std::vector<MSchedGraphNode*> &recurrence, int II, MSchedGraphNode *srcBENode, MSchedGraphNode *destBENode) {
//Check to make sure that this recurrence is unique
bool same = false;
//Loop over all recurrences already in our list
for(std::set<std::pair<int, std::vector<MSchedGraphNode*> > >::iterator R = recurrenceList.begin(), RE = recurrenceList.end(); R != RE; ++R) {
bool all_same = true;
//First compare size
if(R->second.size() == recurrence.size()) {
for(std::vector<MSchedGraphNode*>::const_iterator node = R->second.begin(), end = R->second.end(); node != end; ++node) {
if(std::find(recurrence.begin(), recurrence.end(), *node) == recurrence.end()) {
all_same = all_same && false;
break;
}
else
all_same = all_same && true;
}
if(all_same) {
same = true;
break;
}
}
}
if(!same) {
srcBENode = recurrence.back();
destBENode = recurrence.front();
//FIXME
if(destBENode->getInEdge(srcBENode).getIteDiff() == 0) {
//DEBUG(std::cerr << "NOT A BACKEDGE\n");
//find actual backedge HACK HACK
for(unsigned i=0; i< recurrence.size()-1; ++i) {
if(recurrence[i+1]->getInEdge(recurrence[i]).getIteDiff() == 1) {
srcBENode = recurrence[i];
destBENode = recurrence[i+1];
break;
}
}
}
DEBUG(std::cerr << "Back Edge to Remove: " << *srcBENode << " to " << *destBENode << "\n");
edgesToIgnore.insert(std::make_pair(srcBENode, destBENode->getInEdgeNum(srcBENode)));
recurrenceList.insert(std::make_pair(II, recurrence));
}
}
int CircCount;
void ModuloSchedulingPass::unblock(MSchedGraphNode *u, std::set<MSchedGraphNode*> &blocked,
std::map<MSchedGraphNode*, std::set<MSchedGraphNode*> > &B) {
//Unblock u
DEBUG(std::cerr << "Unblocking: " << *u << "\n");
blocked.erase(u);
//std::set<MSchedGraphNode*> toErase;
while (!B[u].empty()) {
MSchedGraphNode *W = *B[u].begin();
B[u].erase(W);
//toErase.insert(*W);
DEBUG(std::cerr << "Removed: " << *W << "from B-List\n");
if(blocked.count(W))
unblock(W, blocked, B);
}
}
bool ModuloSchedulingPass::circuit(MSchedGraphNode *v, std::vector<MSchedGraphNode*> &stack,
std::set<MSchedGraphNode*> &blocked, std::vector<MSchedGraphNode*> &SCC,
MSchedGraphNode *s, std::map<MSchedGraphNode*, std::set<MSchedGraphNode*> > &B,
int II, std::map<MSchedGraphNode*, MSchedGraphNode*> &newNodes) {
bool f = false;
DEBUG(std::cerr << "Finding Circuits Starting with: ( " << v << ")"<< *v << "\n");
//Push node onto the stack
stack.push_back(v);
//block this node
blocked.insert(v);
//Loop over all successors of node v that are in the scc, create Adjaceny list
std::set<MSchedGraphNode*> AkV;
for(MSchedGraphNode::succ_iterator I = v->succ_begin(), E = v->succ_end(); I != E; ++I) {
if((std::find(SCC.begin(), SCC.end(), *I) != SCC.end())) {
AkV.insert(*I);
}
}
for(std::set<MSchedGraphNode*>::iterator I = AkV.begin(), E = AkV.end(); I != E; ++I) {
if(*I == s) {
//We have a circuit, so add it to our list
addRecc(stack, newNodes);
f = true;
}
else if(!blocked.count(*I)) {
if(circuit(*I, stack, blocked, SCC, s, B, II, newNodes))
f = true;
}
else
DEBUG(std::cerr << "Blocked: " << **I << "\n");
}
if(f) {
unblock(v, blocked, B);
}
else {
for(std::set<MSchedGraphNode*>::iterator I = AkV.begin(), E = AkV.end(); I != E; ++I)
B[*I].insert(v);
}
//Pop v
stack.pop_back();
return f;
}
void ModuloSchedulingPass::addRecc(std::vector<MSchedGraphNode*> &stack, std::map<MSchedGraphNode*, MSchedGraphNode*> &newNodes) {
std::vector<MSchedGraphNode*> recc;
//Dump recurrence for now
DEBUG(std::cerr << "Starting Recc\n");
int totalDelay = 0;
int totalDistance = 0;
MSchedGraphNode *lastN = 0;
MSchedGraphNode *start = 0;
MSchedGraphNode *end = 0;
//Loop over recurrence, get delay and distance
for(std::vector<MSchedGraphNode*>::iterator N = stack.begin(), NE = stack.end(); N != NE; ++N) {
DEBUG(std::cerr << **N << "\n");
totalDelay += (*N)->getLatency();
if(lastN) {
int iteDiff = (*N)->getInEdge(lastN).getIteDiff();
totalDistance += iteDiff;
if(iteDiff > 0) {
start = lastN;
end = *N;
}
}
//Get the original node
lastN = *N;
recc.push_back(newNodes[*N]);
}
//Get the loop edge
totalDistance += lastN->getIteDiff(*stack.begin());
DEBUG(std::cerr << "End Recc\n");
CircCount++;
if(start && end) {
//Insert reccurrence into the list
DEBUG(std::cerr << "Ignore Edge from!!: " << *start << " to " << *end << "\n");
edgesToIgnore.insert(std::make_pair(newNodes[start], (newNodes[end])->getInEdgeNum(newNodes[start])));
}
else {
//Insert reccurrence into the list
DEBUG(std::cerr << "Ignore Edge from: " << *lastN << " to " << **stack.begin() << "\n");
edgesToIgnore.insert(std::make_pair(newNodes[lastN], newNodes[(*stack.begin())]->getInEdgeNum(newNodes[lastN])));
}
//Adjust II until we get close to the inequality delay - II*distance <= 0
int RecMII = II; //Starting value
int value = totalDelay-(RecMII * totalDistance);
int lastII = II;
while(value < 0) {
lastII = RecMII;
RecMII--;
value = totalDelay-(RecMII * totalDistance);
}
recurrenceList.insert(std::make_pair(lastII, recc));
}
void ModuloSchedulingPass::addSCC(std::vector<MSchedGraphNode*> &SCC, std::map<MSchedGraphNode*, MSchedGraphNode*> &newNodes) {
int totalDelay = 0;
int totalDistance = 0;
std::vector<MSchedGraphNode*> recc;
MSchedGraphNode *start = 0;
MSchedGraphNode *end = 0;
//Loop over recurrence, get delay and distance
for(std::vector<MSchedGraphNode*>::iterator N = SCC.begin(), NE = SCC.end(); N != NE; ++N) {
DEBUG(std::cerr << **N << "\n");
totalDelay += (*N)->getLatency();
for(unsigned i = 0; i < (*N)->succ_size(); ++i) {
MSchedGraphEdge *edge = (*N)->getSuccessor(i);
if(find(SCC.begin(), SCC.end(), edge->getDest()) != SCC.end()) {
totalDistance += edge->getIteDiff();
if(edge->getIteDiff() > 0)
if(!start && !end) {
start = *N;
end = edge->getDest();
}
}
}
//Get the original node
recc.push_back(newNodes[*N]);
}
DEBUG(std::cerr << "End Recc\n");
CircCount++;
assert( (start && end) && "Must have start and end node to ignore edge for SCC");
if(start && end) {
//Insert reccurrence into the list
DEBUG(std::cerr << "Ignore Edge from!!: " << *start << " to " << *end << "\n");
edgesToIgnore.insert(std::make_pair(newNodes[start], (newNodes[end])->getInEdgeNum(newNodes[start])));
}
int lastII = totalDelay / totalDistance;
recurrenceList.insert(std::make_pair(lastII, recc));
}
void ModuloSchedulingPass::findAllCircuits(MSchedGraph *g, int II) {
CircCount = 0;
//Keep old to new node mapping information
std::map<MSchedGraphNode*, MSchedGraphNode*> newNodes;
//copy the graph
MSchedGraph *MSG = new MSchedGraph(*g, newNodes);
DEBUG(std::cerr << "Finding All Circuits\n");
//Set of blocked nodes
std::set<MSchedGraphNode*> blocked;
//Stack holding current circuit
std::vector<MSchedGraphNode*> stack;
//Map for B Lists
std::map<MSchedGraphNode*, std::set<MSchedGraphNode*> > B;
//current node
MSchedGraphNode *s;
//Iterate over the graph until its down to one node or empty
while(MSG->size() > 1) {
//Write Graph out to file
//WriteGraphToFile(std::cerr, "Graph" + utostr(MSG->size()), MSG);
DEBUG(std::cerr << "Graph Size: " << MSG->size() << "\n");
DEBUG(std::cerr << "Finding strong component Vk with least vertex\n");
//Iterate over all the SCCs in the graph
std::set<MSchedGraphNode*> Visited;
std::vector<MSchedGraphNode*> Vk;
MSchedGraphNode* s = 0;
int numEdges = 0;
//Find scc with the least vertex
for (MSchedGraph::iterator GI = MSG->begin(), E = MSG->end(); GI != E; ++GI)
if (Visited.insert(GI->second).second) {
for (scc_iterator<MSchedGraphNode*> SCCI = scc_begin(GI->second),
E = scc_end(GI->second); SCCI != E; ++SCCI) {
std::vector<MSchedGraphNode*> &nextSCC = *SCCI;
if (Visited.insert(nextSCC[0]).second) {
Visited.insert(nextSCC.begin()+1, nextSCC.end());
if(nextSCC.size() > 1) {
std::cerr << "SCC size: " << nextSCC.size() << "\n";
for(unsigned i = 0; i < nextSCC.size(); ++i) {
//Loop over successor and see if in scc, then count edge
MSchedGraphNode *node = nextSCC[i];
for(MSchedGraphNode::succ_iterator S = node->succ_begin(), SE = node->succ_end(); S != SE; ++S) {
if(find(nextSCC.begin(), nextSCC.end(), *S) != nextSCC.end())
numEdges++;
}
}
std::cerr << "Num Edges: " << numEdges << "\n";
}
//Ignore self loops
if(nextSCC.size() > 1) {
//Get least vertex in Vk
if(!s) {
s = nextSCC[0];
Vk = nextSCC;
}
for(unsigned i = 0; i < nextSCC.size(); ++i) {
if(nextSCC[i] < s) {
s = nextSCC[i];
Vk = nextSCC;
}
}
}
}
}
}
//Process SCC
DEBUG(for(std::vector<MSchedGraphNode*>::iterator N = Vk.begin(), NE = Vk.end();
N != NE; ++N) { std::cerr << *((*N)->getInst()); });
//Iterate over all nodes in this scc
for(std::vector<MSchedGraphNode*>::iterator N = Vk.begin(), NE = Vk.end();
N != NE; ++N) {
blocked.erase(*N);
B[*N].clear();
}
if(Vk.size() > 1) {
if(numEdges < 98)
circuit(s, stack, blocked, Vk, s, B, II, newNodes);
else
addSCC(Vk, newNodes);
//Delete nodes from the graph
//Find all nodes up to s and delete them
std::vector<MSchedGraphNode*> nodesToRemove;
nodesToRemove.push_back(s);
for(MSchedGraph::iterator N = MSG->begin(), NE = MSG->end(); N != NE; ++N) {
if(N->second < s )
nodesToRemove.push_back(N->second);
}
for(std::vector<MSchedGraphNode*>::iterator N = nodesToRemove.begin(), NE = nodesToRemove.end(); N != NE; ++N) {
DEBUG(std::cerr << "Deleting Node: " << **N << "\n");
MSG->deleteNode(*N);
}
}
else
break;
}
DEBUG(std::cerr << "Num Circuits found: " << CircCount << "\n");
}
void ModuloSchedulingPass::findAllReccurrences(MSchedGraphNode *node,
std::vector<MSchedGraphNode*> &visitedNodes,
int II) {
if(std::find(visitedNodes.begin(), visitedNodes.end(), node) != visitedNodes.end()) {
std::vector<MSchedGraphNode*> recurrence;
bool first = true;
int delay = 0;
int distance = 0;
int RecMII = II; //Starting value
MSchedGraphNode *last = node;
MSchedGraphNode *srcBackEdge = 0;
MSchedGraphNode *destBackEdge = 0;
for(std::vector<MSchedGraphNode*>::iterator I = visitedNodes.begin(), E = visitedNodes.end();
I !=E; ++I) {
if(*I == node)
first = false;
if(first)
continue;
delay = delay + (*I)->getLatency();
if(*I != node) {
int diff = (*I)->getInEdge(last).getIteDiff();
distance += diff;
if(diff > 0) {
srcBackEdge = last;
destBackEdge = *I;
}
}
recurrence.push_back(*I);
last = *I;
}
//Get final distance calc
distance += node->getInEdge(last).getIteDiff();
DEBUG(std::cerr << "Reccurrence Distance: " << distance << "\n");
//Adjust II until we get close to the inequality delay - II*distance <= 0
int value = delay-(RecMII * distance);
int lastII = II;
while(value <= 0) {
lastII = RecMII;
RecMII--;
value = delay-(RecMII * distance);
}
DEBUG(std::cerr << "Final II for this recurrence: " << lastII << "\n");
addReccurrence(recurrence, lastII, srcBackEdge, destBackEdge);
assert(distance != 0 && "Recurrence distance should not be zero");
return;
}
unsigned count = 0;
for(MSchedGraphNode::succ_iterator I = node->succ_begin(), E = node->succ_end(); I != E; ++I) {
visitedNodes.push_back(node);
//if(!edgesToIgnore.count(std::make_pair(node, count)))
findAllReccurrences(*I, visitedNodes, II);
visitedNodes.pop_back();
count++;
}
}
void ModuloSchedulingPass::searchPath(MSchedGraphNode *node,
std::vector<MSchedGraphNode*> &path,
std::set<MSchedGraphNode*> &nodesToAdd,
std::set<MSchedGraphNode*> &new_reccurrence) {
//Push node onto the path
path.push_back(node);
//Loop over all successors and see if there is a path from this node to
//a recurrence in the partial order, if so.. add all nodes to be added to recc
for(MSchedGraphNode::succ_iterator S = node->succ_begin(), SE = node->succ_end(); S != SE;
++S) {
//Check if we should ignore this edge first
if(ignoreEdge(node,*S))
continue;
//check if successor is in this recurrence, we will get to it eventually
if(new_reccurrence.count(*S))
continue;
//If this node exists in a recurrence already in the partial
//order, then add all nodes in the path to the set of nodes to add
//Check if its already in our partial order, if not add it to the
//final vector
bool found = false;
for(std::vector<std::set<MSchedGraphNode*> >::iterator PO = partialOrder.begin(),
PE = partialOrder.end(); PO != PE; ++PO) {
if(PO->count(*S)) {
found = true;
break;
}
}
if(!found) {
nodesToAdd.insert(*S);
searchPath(*S, path, nodesToAdd, new_reccurrence);
}
}
//Pop Node off the path
path.pop_back();
}
void ModuloSchedulingPass::pathToRecc(MSchedGraphNode *node,
std::vector<MSchedGraphNode*> &path,
std::set<MSchedGraphNode*> &poSet,
std::set<MSchedGraphNode*> &lastNodes) {
//Push node onto the path
path.push_back(node);
DEBUG(std::cerr << "Current node: " << *node << "\n");
//Loop over all successors and see if there is a path from this node to
//a recurrence in the partial order, if so.. add all nodes to be added to recc
for(MSchedGraphNode::succ_iterator S = node->succ_begin(), SE = node->succ_end(); S != SE;
++S) {
DEBUG(std::cerr << "Succ:" << **S << "\n");
//Check if we should ignore this edge first
if(ignoreEdge(node,*S))
continue;
if(poSet.count(*S)) {
DEBUG(std::cerr << "Found path to recc from no pred\n");
//Loop over path, if it exists in lastNodes, then add to poset, and remove from lastNodes
for(std::vector<MSchedGraphNode*>::iterator I = path.begin(), IE = path.end(); I != IE; ++I) {
if(lastNodes.count(*I)) {
DEBUG(std::cerr << "Inserting node into recc: " << **I << "\n");
poSet.insert(*I);
lastNodes.erase(*I);
}
}
}
else
pathToRecc(*S, path, poSet, lastNodes);
}
//Pop Node off the path
path.pop_back();
}
void ModuloSchedulingPass::computePartialOrder() {
TIME_REGION(X, "calculatePartialOrder");
DEBUG(std::cerr << "Computing Partial Order\n");
//Only push BA branches onto the final node order, we put other
//branches after it FIXME: Should we really be pushing branches on
//it a specific order instead of relying on BA being there?
std::vector<MSchedGraphNode*> branches;
//Steps to add a recurrence to the partial order 1) Find reccurrence
//with the highest RecMII. Add it to the partial order. 2) For each
//recurrence with decreasing RecMII, add it to the partial order
//along with any nodes that connect this recurrence to recurrences
//already in the partial order
for(std::set<std::pair<int, std::vector<MSchedGraphNode*> > >::reverse_iterator
I = recurrenceList.rbegin(), E=recurrenceList.rend(); I !=E; ++I) {
std::set<MSchedGraphNode*> new_recurrence;
//Loop through recurrence and remove any nodes already in the partial order
for(std::vector<MSchedGraphNode*>::const_iterator N = I->second.begin(),
NE = I->second.end(); N != NE; ++N) {
bool found = false;
for(std::vector<std::set<MSchedGraphNode*> >::iterator PO = partialOrder.begin(),
PE = partialOrder.end(); PO != PE; ++PO) {
if(PO->count(*N))
found = true;
}
//Check if its a branch, and remove to handle special
if(!found) {
if((*N)->isBranch() && !(*N)->hasPredecessors()) {
branches.push_back(*N);
}
else
new_recurrence.insert(*N);
}
}
if(new_recurrence.size() > 0) {
std::vector<MSchedGraphNode*> path;
std::set<MSchedGraphNode*> nodesToAdd;
//Dump recc we are dealing with (minus nodes already in PO)
DEBUG(std::cerr << "Recc: ");
DEBUG(for(std::set<MSchedGraphNode*>::iterator R = new_recurrence.begin(), RE = new_recurrence.end(); R != RE; ++R) { std::cerr << **R ; });
//Add nodes that connect this recurrence to recurrences in the partial path
for(std::set<MSchedGraphNode*>::iterator N = new_recurrence.begin(),
NE = new_recurrence.end(); N != NE; ++N)
searchPath(*N, path, nodesToAdd, new_recurrence);
//Add nodes to this recurrence if they are not already in the partial order
for(std::set<MSchedGraphNode*>::iterator N = nodesToAdd.begin(), NE = nodesToAdd.end();
N != NE; ++N) {
bool found = false;
for(std::vector<std::set<MSchedGraphNode*> >::iterator PO = partialOrder.begin(),
PE = partialOrder.end(); PO != PE; ++PO) {
if(PO->count(*N))
found = true;
}
if(!found) {
assert("FOUND CONNECTOR");
new_recurrence.insert(*N);
}
}
partialOrder.push_back(new_recurrence);
//Dump out partial order
DEBUG(for(std::vector<std::set<MSchedGraphNode*> >::iterator I = partialOrder.begin(),
E = partialOrder.end(); I !=E; ++I) {
std::cerr << "Start set in PO\n";
for(std::set<MSchedGraphNode*>::iterator J = I->begin(), JE = I->end(); J != JE; ++J)
std::cerr << "PO:" << **J << "\n";
});
}
}
//Add any nodes that are not already in the partial order
//Add them in a set, one set per connected component
std::set<MSchedGraphNode*> lastNodes;
std::set<MSchedGraphNode*> noPredNodes;
for(std::map<MSchedGraphNode*, MSNodeAttributes>::iterator I = nodeToAttributesMap.begin(),
E = nodeToAttributesMap.end(); I != E; ++I) {
bool found = false;
//Check if its already in our partial order, if not add it to the final vector
for(std::vector<std::set<MSchedGraphNode*> >::iterator PO = partialOrder.begin(),
PE = partialOrder.end(); PO != PE; ++PO) {
if(PO->count(I->first))
found = true;
}
if(!found)
lastNodes.insert(I->first);
}
//For each node w/out preds, see if there is a path to one of the
//recurrences, and if so add them to that current recc
/*for(std::set<MSchedGraphNode*>::iterator N = noPredNodes.begin(), NE = noPredNodes.end();
N != NE; ++N) {
DEBUG(std::cerr << "No Pred Path from: " << **N << "\n");
for(std::vector<std::set<MSchedGraphNode*> >::iterator PO = partialOrder.begin(),
PE = partialOrder.end(); PO != PE; ++PO) {
std::vector<MSchedGraphNode*> path;
pathToRecc(*N, path, *PO, lastNodes);
}
}*/
//Break up remaining nodes that are not in the partial order
///into their connected compoenents
while(lastNodes.size() > 0) {
std::set<MSchedGraphNode*> ccSet;
connectedComponentSet(*(lastNodes.begin()),ccSet, lastNodes);
if(ccSet.size() > 0)
partialOrder.push_back(ccSet);
}
//Clean up branches by putting them in final order
assert(branches.size() == 0 && "We should not have any branches in our graph");
}
void ModuloSchedulingPass::connectedComponentSet(MSchedGraphNode *node, std::set<MSchedGraphNode*> &ccSet, std::set<MSchedGraphNode*> &lastNodes) {
//Add to final set
if( !ccSet.count(node) && lastNodes.count(node)) {
lastNodes.erase(node);
ccSet.insert(node);
}
else
return;
//Loop over successors and recurse if we have not seen this node before
for(MSchedGraphNode::succ_iterator node_succ = node->succ_begin(), end=node->succ_end(); node_succ != end; ++node_succ) {
connectedComponentSet(*node_succ, ccSet, lastNodes);
}
}
void ModuloSchedulingPass::predIntersect(std::set<MSchedGraphNode*> &CurrentSet, std::set<MSchedGraphNode*> &IntersectResult) {
for(unsigned j=0; j < FinalNodeOrder.size(); ++j) {
for(MSchedGraphNode::pred_iterator P = FinalNodeOrder[j]->pred_begin(),
E = FinalNodeOrder[j]->pred_end(); P != E; ++P) {
//Check if we are supposed to ignore this edge or not
if(ignoreEdge(*P,FinalNodeOrder[j]))
continue;
if(CurrentSet.count(*P))
if(std::find(FinalNodeOrder.begin(), FinalNodeOrder.end(), *P) == FinalNodeOrder.end())
IntersectResult.insert(*P);
}
}
}
void ModuloSchedulingPass::succIntersect(std::set<MSchedGraphNode*> &CurrentSet, std::set<MSchedGraphNode*> &IntersectResult) {
for(unsigned j=0; j < FinalNodeOrder.size(); ++j) {
for(MSchedGraphNode::succ_iterator P = FinalNodeOrder[j]->succ_begin(),
E = FinalNodeOrder[j]->succ_end(); P != E; ++P) {
//Check if we are supposed to ignore this edge or not
if(ignoreEdge(FinalNodeOrder[j],*P))
continue;
if(CurrentSet.count(*P))
if(std::find(FinalNodeOrder.begin(), FinalNodeOrder.end(), *P) == FinalNodeOrder.end())
IntersectResult.insert(*P);
}
}
}
void dumpIntersection(std::set<MSchedGraphNode*> &IntersectCurrent) {
std::cerr << "Intersection (";
for(std::set<MSchedGraphNode*>::iterator I = IntersectCurrent.begin(), E = IntersectCurrent.end(); I != E; ++I)
std::cerr << **I << ", ";
std::cerr << ")\n";
}
void ModuloSchedulingPass::orderNodes() {
TIME_REGION(X, "orderNodes");
int BOTTOM_UP = 0;
int TOP_DOWN = 1;
//Set default order
int order = BOTTOM_UP;
//Loop over all the sets and place them in the final node order
for(std::vector<std::set<MSchedGraphNode*> >::iterator CurrentSet = partialOrder.begin(), E= partialOrder.end(); CurrentSet != E; ++CurrentSet) {
DEBUG(std::cerr << "Processing set in S\n");
DEBUG(dumpIntersection(*CurrentSet));
//Result of intersection
std::set<MSchedGraphNode*> IntersectCurrent;
predIntersect(*CurrentSet, IntersectCurrent);
//If the intersection of predecessor and current set is not empty
//sort nodes bottom up
if(IntersectCurrent.size() != 0) {
DEBUG(std::cerr << "Final Node Order Predecessors and Current Set interesection is NOT empty\n");
order = BOTTOM_UP;
}
//If empty, use successors
else {
DEBUG(std::cerr << "Final Node Order Predecessors and Current Set interesection is empty\n");
succIntersect(*CurrentSet, IntersectCurrent);
//sort top-down
if(IntersectCurrent.size() != 0) {
DEBUG(std::cerr << "Final Node Order Successors and Current Set interesection is NOT empty\n");
order = TOP_DOWN;
}
else {
DEBUG(std::cerr << "Final Node Order Successors and Current Set interesection is empty\n");
//Find node with max ASAP in current Set
MSchedGraphNode *node;
int maxASAP = 0;
DEBUG(std::cerr << "Using current set of size " << CurrentSet->size() << "to find max ASAP\n");
for(std::set<MSchedGraphNode*>::iterator J = CurrentSet->begin(), JE = CurrentSet->end(); J != JE; ++J) {
//Get node attributes
MSNodeAttributes nodeAttr= nodeToAttributesMap.find(*J)->second;
//assert(nodeAttr != nodeToAttributesMap.end() && "Node not in attributes map!");
if(maxASAP <= nodeAttr.ASAP) {
maxASAP = nodeAttr.ASAP;
node = *J;
}
}
assert(node != 0 && "In node ordering node should not be null");
IntersectCurrent.insert(node);
order = BOTTOM_UP;
}
}
//Repeat until all nodes are put into the final order from current set
while(IntersectCurrent.size() > 0) {
if(order == TOP_DOWN) {
DEBUG(std::cerr << "Order is TOP DOWN\n");
while(IntersectCurrent.size() > 0) {
DEBUG(std::cerr << "Intersection is not empty, so find heighest height\n");
int MOB = 0;
int height = 0;
MSchedGraphNode *highestHeightNode = *(IntersectCurrent.begin());
//Find node in intersection with highest heigh and lowest MOB
for(std::set<MSchedGraphNode*>::iterator I = IntersectCurrent.begin(),
E = IntersectCurrent.end(); I != E; ++I) {
//Get current nodes properties
MSNodeAttributes nodeAttr= nodeToAttributesMap.find(*I)->second;
if(height < nodeAttr.height) {
highestHeightNode = *I;
height = nodeAttr.height;
MOB = nodeAttr.MOB;
}
else if(height == nodeAttr.height) {
if(MOB > nodeAttr.height) {
highestHeightNode = *I;
height = nodeAttr.height;
MOB = nodeAttr.MOB;
}
}
}
//Append our node with greatest height to the NodeOrder
if(std::find(FinalNodeOrder.begin(), FinalNodeOrder.end(), highestHeightNode) == FinalNodeOrder.end()) {
DEBUG(std::cerr << "Adding node to Final Order: " << *highestHeightNode << "\n");
FinalNodeOrder.push_back(highestHeightNode);
}
//Remove V from IntersectOrder
IntersectCurrent.erase(std::find(IntersectCurrent.begin(),
IntersectCurrent.end(), highestHeightNode));
//Intersect V's successors with CurrentSet
for(MSchedGraphNode::succ_iterator P = highestHeightNode->succ_begin(),
E = highestHeightNode->succ_end(); P != E; ++P) {
//if(lower_bound(CurrentSet->begin(),
// CurrentSet->end(), *P) != CurrentSet->end()) {
if(std::find(CurrentSet->begin(), CurrentSet->end(), *P) != CurrentSet->end()) {
if(ignoreEdge(highestHeightNode, *P))
continue;
//If not already in Intersect, add
if(!IntersectCurrent.count(*P))
IntersectCurrent.insert(*P);
}
}
} //End while loop over Intersect Size
//Change direction
order = BOTTOM_UP;
//Reset Intersect to reflect changes in OrderNodes
IntersectCurrent.clear();
predIntersect(*CurrentSet, IntersectCurrent);
} //End If TOP_DOWN
//Begin if BOTTOM_UP
else {
DEBUG(std::cerr << "Order is BOTTOM UP\n");
while(IntersectCurrent.size() > 0) {
DEBUG(std::cerr << "Intersection of size " << IntersectCurrent.size() << ", finding highest depth\n");
//dump intersection
DEBUG(dumpIntersection(IntersectCurrent));
//Get node with highest depth, if a tie, use one with lowest
//MOB
int MOB = 0;
int depth = 0;
MSchedGraphNode *highestDepthNode = *(IntersectCurrent.begin());
for(std::set<MSchedGraphNode*>::iterator I = IntersectCurrent.begin(),
E = IntersectCurrent.end(); I != E; ++I) {
//Find node attribute in graph
MSNodeAttributes nodeAttr= nodeToAttributesMap.find(*I)->second;
if(depth < nodeAttr.depth) {
highestDepthNode = *I;
depth = nodeAttr.depth;
MOB = nodeAttr.MOB;
}
else if(depth == nodeAttr.depth) {
if(MOB > nodeAttr.MOB) {
highestDepthNode = *I;
depth = nodeAttr.depth;
MOB = nodeAttr.MOB;
}
}
}
//Append highest depth node to the NodeOrder
if(std::find(FinalNodeOrder.begin(), FinalNodeOrder.end(), highestDepthNode) == FinalNodeOrder.end()) {
DEBUG(std::cerr << "Adding node to Final Order: " << *highestDepthNode << "\n");
FinalNodeOrder.push_back(highestDepthNode);
}
//Remove heightestDepthNode from IntersectOrder
IntersectCurrent.erase(highestDepthNode);
//Intersect heightDepthNode's pred with CurrentSet
for(MSchedGraphNode::pred_iterator P = highestDepthNode->pred_begin(),
E = highestDepthNode->pred_end(); P != E; ++P) {
if(CurrentSet->count(*P)) {
if(ignoreEdge(*P, highestDepthNode))
continue;
//If not already in Intersect, add
if(!IntersectCurrent.count(*P))
IntersectCurrent.insert(*P);
}
}
} //End while loop over Intersect Size
//Change order
order = TOP_DOWN;
//Reset IntersectCurrent to reflect changes in OrderNodes
IntersectCurrent.clear();
succIntersect(*CurrentSet, IntersectCurrent);
} //End if BOTTOM_DOWN
DEBUG(std::cerr << "Current Intersection Size: " << IntersectCurrent.size() << "\n");
}
//End Wrapping while loop
DEBUG(std::cerr << "Ending Size of Current Set: " << CurrentSet->size() << "\n");
}//End for over all sets of nodes
//FIXME: As the algorithm stands it will NEVER add an instruction such as ba (with no
//data dependencies) to the final order. We add this manually. It will always be
//in the last set of S since its not part of a recurrence
//Loop over all the sets and place them in the final node order
std::vector<std::set<MSchedGraphNode*> > ::reverse_iterator LastSet = partialOrder.rbegin();
for(std::set<MSchedGraphNode*>::iterator CurrentNode = LastSet->begin(), LastNode = LastSet->end();
CurrentNode != LastNode; ++CurrentNode) {
if((*CurrentNode)->getInst()->getOpcode() == V9::BA)
FinalNodeOrder.push_back(*CurrentNode);
}
//Return final Order
//return FinalNodeOrder;
}
bool ModuloSchedulingPass::computeSchedule(const MachineBasicBlock *BB, MSchedGraph *MSG) {
TIME_REGION(X, "computeSchedule");
bool success = false;
//FIXME: Should be set to max II of the original loop
//Cap II in order to prevent infinite loop
int capII = MSG->totalDelay();
while(!success) {
//Keep track of branches, but do not insert into the schedule
std::vector<MSchedGraphNode*> branches;
//Loop over the final node order and process each node
for(std::vector<MSchedGraphNode*>::iterator I = FinalNodeOrder.begin(),
E = FinalNodeOrder.end(); I != E; ++I) {
//CalculateEarly and Late start
bool initialLSVal = false;
bool initialESVal = false;
int EarlyStart = 0;
int LateStart = 0;
bool hasSucc = false;
bool hasPred = false;
bool sched;
if((*I)->isBranch())
if((*I)->hasPredecessors())
sched = true;
else
sched = false;
else
sched = true;
if(sched) {
//Loop over nodes in the schedule and determine if they are predecessors
//or successors of the node we are trying to schedule
for(MSSchedule::schedule_iterator nodesByCycle = schedule.begin(), nodesByCycleEnd = schedule.end();
nodesByCycle != nodesByCycleEnd; ++nodesByCycle) {
//For this cycle, get the vector of nodes schedule and loop over it
for(std::vector<MSchedGraphNode*>::iterator schedNode = nodesByCycle->second.begin(), SNE = nodesByCycle->second.end(); schedNode != SNE; ++schedNode) {
if((*I)->isPredecessor(*schedNode)) {
int diff = (*I)->getInEdge(*schedNode).getIteDiff();
int ES_Temp = nodesByCycle->first + (*schedNode)->getLatency() - diff * II;
DEBUG(std::cerr << "Diff: " << diff << " Cycle: " << nodesByCycle->first << "\n");
DEBUG(std::cerr << "Temp EarlyStart: " << ES_Temp << " Prev EarlyStart: " << EarlyStart << "\n");
if(initialESVal)
EarlyStart = std::max(EarlyStart, ES_Temp);
else {
EarlyStart = ES_Temp;
initialESVal = true;
}
hasPred = true;
}
if((*I)->isSuccessor(*schedNode)) {
int diff = (*schedNode)->getInEdge(*I).getIteDiff();
int LS_Temp = nodesByCycle->first - (*I)->getLatency() + diff * II;
DEBUG(std::cerr << "Diff: " << diff << " Cycle: " << nodesByCycle->first << "\n");
DEBUG(std::cerr << "Temp LateStart: " << LS_Temp << " Prev LateStart: " << LateStart << "\n");
if(initialLSVal)
LateStart = std::min(LateStart, LS_Temp);
else {
LateStart = LS_Temp;
initialLSVal = true;
}
hasSucc = true;
}
}
}
}
else {
branches.push_back(*I);
continue;
}
//Check if this node is a pred or succ to a branch, and restrict its placement
//even though the branch is not in the schedule
/*int count = branches.size();
for(std::vector<MSchedGraphNode*>::iterator B = branches.begin(), BE = branches.end();
B != BE; ++B) {
if((*I)->isPredecessor(*B)) {
int diff = (*I)->getInEdge(*B).getIteDiff();
int ES_Temp = (II+count-1) + (*B)->getLatency() - diff * II;
DEBUG(std::cerr << "Diff: " << diff << " Cycle: " << (II+count)-1 << "\n");
DEBUG(std::cerr << "Temp EarlyStart: " << ES_Temp << " Prev EarlyStart: " << EarlyStart << "\n");
EarlyStart = std::max(EarlyStart, ES_Temp);
hasPred = true;
}
if((*I)->isSuccessor(*B)) {
int diff = (*B)->getInEdge(*I).getIteDiff();
int LS_Temp = (II+count-1) - (*I)->getLatency() + diff * II;
DEBUG(std::cerr << "Diff: " << diff << " Cycle: " << (II+count-1) << "\n");
DEBUG(std::cerr << "Temp LateStart: " << LS_Temp << " Prev LateStart: " << LateStart << "\n");
LateStart = std::min(LateStart, LS_Temp);
hasSucc = true;
}
count--;
}*/
//Check if the node has no pred or successors and set Early Start to its ASAP
if(!hasSucc && !hasPred)
EarlyStart = nodeToAttributesMap.find(*I)->second.ASAP;
DEBUG(std::cerr << "Has Successors: " << hasSucc << ", Has Pred: " << hasPred << "\n");
DEBUG(std::cerr << "EarlyStart: " << EarlyStart << ", LateStart: " << LateStart << "\n");
//Now, try to schedule this node depending upon its pred and successor in the schedule
//already
if(!hasSucc && hasPred)
success = scheduleNode(*I, EarlyStart, (EarlyStart + II -1));
else if(!hasPred && hasSucc)
success = scheduleNode(*I, LateStart, (LateStart - II +1));
else if(hasPred && hasSucc) {
if(EarlyStart > LateStart) {
success = false;
//LateStart = EarlyStart;
DEBUG(std::cerr << "Early Start can not be later then the late start cycle, schedule fails\n");
}
else
success = scheduleNode(*I, EarlyStart, std::min(LateStart, (EarlyStart + II -1)));
}
else
success = scheduleNode(*I, EarlyStart, EarlyStart + II - 1);
if(!success) {
++II;
schedule.clear();
break;
}
}
if(success) {
DEBUG(std::cerr << "Constructing Schedule Kernel\n");
success = schedule.constructKernel(II, branches, indVarInstrs[BB]);
DEBUG(std::cerr << "Done Constructing Schedule Kernel\n");
if(!success) {
++II;
schedule.clear();
}
DEBUG(std::cerr << "Final II: " << II << "\n");
}
if(II >= capII) {
DEBUG(std::cerr << "Maximum II reached, giving up\n");
return false;
}
assert(II < capII && "The II should not exceed the original loop number of cycles");
}
return true;
}
bool ModuloSchedulingPass::scheduleNode(MSchedGraphNode *node,
int start, int end) {
bool success = false;
DEBUG(std::cerr << *node << " (Start Cycle: " << start << ", End Cycle: " << end << ")\n");
//Make sure start and end are not negative
//if(start < 0) {
//start = 0;
//}
//if(end < 0)
//end = 0;
bool forward = true;
if(start > end)
forward = false;
bool increaseSC = true;
int cycle = start ;
while(increaseSC) {
increaseSC = false;
increaseSC = schedule.insert(node, cycle, II);
if(!increaseSC)
return true;
//Increment cycle to try again
if(forward) {
++cycle;
DEBUG(std::cerr << "Increase cycle: " << cycle << "\n");
if(cycle > end)
return false;
}
else {
--cycle;
DEBUG(std::cerr << "Decrease cycle: " << cycle << "\n");
if(cycle < end)
return false;
}
}
return success;
}
void ModuloSchedulingPass::writePrologues(std::vector<MachineBasicBlock *> &prologues, MachineBasicBlock *origBB, std::vector<BasicBlock*> &llvm_prologues, std::map<const Value*, std::pair<const MachineInstr*, int> > &valuesToSave, std::map<Value*, std::map<int, Value*> > &newValues, std::map<Value*, MachineBasicBlock*> &newValLocation) {
//Keep a map to easily know whats in the kernel
std::map<int, std::set<const MachineInstr*> > inKernel;
int maxStageCount = 0;
//Keep a map of new values we consumed in case they need to be added back
std::map<Value*, std::map<int, Value*> > consumedValues;
MSchedGraphNode *branch = 0;
MSchedGraphNode *BAbranch = 0;
DEBUG(schedule.print(std::cerr));
std::vector<MSchedGraphNode*> branches;
for(MSSchedule::kernel_iterator I = schedule.kernel_begin(), E = schedule.kernel_end(); I != E; ++I) {
maxStageCount = std::max(maxStageCount, I->second);
//Put int the map so we know what instructions in each stage are in the kernel
DEBUG(std::cerr << "Inserting instruction " << *(I->first) << " into map at stage " << I->second << "\n");
inKernel[I->second].insert(I->first);
}
//Get target information to look at machine operands
const TargetInstrInfo *mii = target.getInstrInfo();
//Now write the prologues
for(int i = 0; i < maxStageCount; ++i) {
BasicBlock *llvmBB = new BasicBlock("PROLOGUE", (Function*) (origBB->getBasicBlock()->getParent()));
MachineBasicBlock *machineBB = new MachineBasicBlock(llvmBB);
DEBUG(std::cerr << "i=" << i << "\n");
for(int j = i; j >= 0; --j) {
for(MachineBasicBlock::const_iterator MI = origBB->begin(), ME = origBB->end(); ME != MI; ++MI) {
if(inKernel[j].count(&*MI)) {
MachineInstr *instClone = MI->clone();
machineBB->push_back(instClone);
//If its a branch, insert a nop
if(mii->isBranch(instClone->getOpcode()))
BuildMI(machineBB, V9::NOP, 0);
DEBUG(std::cerr << "Cloning: " << *MI << "\n");
//After cloning, we may need to save the value that this instruction defines
for(unsigned opNum=0; opNum < MI->getNumOperands(); ++opNum) {
Instruction *tmp;
//get machine operand
MachineOperand &mOp = instClone->getOperand(opNum);
if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isDef()) {
//Check if this is a value we should save
if(valuesToSave.count(mOp.getVRegValue())) {
//Save copy in tmpInstruction
tmp = new TmpInstruction(mOp.getVRegValue());
//Add TmpInstruction to safe LLVM Instruction MCFI
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
tempMvec.addTemp((Value*) tmp);
DEBUG(std::cerr << "Value: " << *(mOp.getVRegValue()) << " New Value: " << *tmp << " Stage: " << i << "\n");
newValues[mOp.getVRegValue()][i]= tmp;
newValLocation[tmp] = machineBB;
DEBUG(std::cerr << "Machine Instr Operands: " << *(mOp.getVRegValue()) << ", 0, " << *tmp << "\n");
//Create machine instruction and put int machineBB
MachineInstr *saveValue;
if(mOp.getVRegValue()->getType() == Type::FloatTy)
saveValue = BuildMI(machineBB, V9::FMOVS, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else if(mOp.getVRegValue()->getType() == Type::DoubleTy)
saveValue = BuildMI(machineBB, V9::FMOVD, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else
saveValue = BuildMI(machineBB, V9::ORr, 3).addReg(mOp.getVRegValue()).addImm(0).addRegDef(tmp);
DEBUG(std::cerr << "Created new machine instr: " << *saveValue << "\n");
}
}
//We may also need to update the value that we use if its from an earlier prologue
if(j != 0) {
if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isUse()) {
if(newValues.count(mOp.getVRegValue())) {
if(newValues[mOp.getVRegValue()].count(i-1)) {
Value *oldV = mOp.getVRegValue();
DEBUG(std::cerr << "Replaced this value: " << mOp.getVRegValue() << " With:" << (newValues[mOp.getVRegValue()][i-1]) << "\n");
//Update the operand with the right value
mOp.setValueReg(newValues[mOp.getVRegValue()][i-1]);
//Remove this value since we have consumed it
//NOTE: Should this only be done if j != maxStage?
consumedValues[oldV][i-1] = (newValues[oldV][i-1]);
DEBUG(std::cerr << "Deleted value: " << consumedValues[oldV][i-1] << "\n");
newValues[oldV].erase(i-1);
}
}
else
if(consumedValues.count(mOp.getVRegValue()))
assert(!consumedValues[mOp.getVRegValue()].count(i-1) && "Found a case where we need the value");
}
}
}
}
}
}
MachineFunction *F = (((MachineBasicBlock*)origBB)->getParent());
MachineFunction::BasicBlockListType &BL = F->getBasicBlockList();
MachineFunction::BasicBlockListType::iterator BLI = origBB;
assert(BLI != BL.end() && "Must find original BB in machine function\n");
BL.insert(BLI,machineBB);
prologues.push_back(machineBB);
llvm_prologues.push_back(llvmBB);
}
}
void ModuloSchedulingPass::writeEpilogues(std::vector<MachineBasicBlock *> &epilogues, const MachineBasicBlock *origBB, std::vector<BasicBlock*> &llvm_epilogues, std::map<const Value*, std::pair<const MachineInstr*, int> > &valuesToSave, std::map<Value*, std::map<int, Value*> > &newValues,std::map<Value*, MachineBasicBlock*> &newValLocation, std::map<Value*, std::map<int, Value*> > &kernelPHIs ) {
std::map<int, std::set<const MachineInstr*> > inKernel;
for(MSSchedule::kernel_iterator I = schedule.kernel_begin(), E = schedule.kernel_end(); I != E; ++I) {
//Ignore the branch, we will handle this separately
//if(I->first->isBranch())
//continue;
//Put int the map so we know what instructions in each stage are in the kernel
inKernel[I->second].insert(I->first);
}
std::map<Value*, Value*> valPHIs;
//some debug stuff, will remove later
DEBUG(for(std::map<Value*, std::map<int, Value*> >::iterator V = newValues.begin(), E = newValues.end(); V !=E; ++V) {
std::cerr << "Old Value: " << *(V->first) << "\n";
for(std::map<int, Value*>::iterator I = V->second.begin(), IE = V->second.end(); I != IE; ++I)
std::cerr << "Stage: " << I->first << " Value: " << *(I->second) << "\n";
});
//some debug stuff, will remove later
DEBUG(for(std::map<Value*, std::map<int, Value*> >::iterator V = kernelPHIs.begin(), E = kernelPHIs.end(); V !=E; ++V) {
std::cerr << "Old Value: " << *(V->first) << "\n";
for(std::map<int, Value*>::iterator I = V->second.begin(), IE = V->second.end(); I != IE; ++I)
std::cerr << "Stage: " << I->first << " Value: " << *(I->second) << "\n";
});
//Now write the epilogues
for(int i = schedule.getMaxStage()-1; i >= 0; --i) {
BasicBlock *llvmBB = new BasicBlock("EPILOGUE", (Function*) (origBB->getBasicBlock()->getParent()));
MachineBasicBlock *machineBB = new MachineBasicBlock(llvmBB);
DEBUG(std::cerr << " Epilogue #: " << i << "\n");
std::map<Value*, int> inEpilogue;
for(MachineBasicBlock::const_iterator MI = origBB->begin(), ME = origBB->end(); ME != MI; ++MI) {
for(int j=schedule.getMaxStage(); j > i; --j) {
if(inKernel[j].count(&*MI)) {
DEBUG(std::cerr << "Cloning instruction " << *MI << "\n");
MachineInstr *clone = MI->clone();
//Update operands that need to use the result from the phi
for(unsigned opNum=0; opNum < clone->getNumOperands(); ++opNum) {
//get machine operand
const MachineOperand &mOp = clone->getOperand(opNum);
if((mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isUse())) {
DEBUG(std::cerr << "Writing PHI for " << (mOp.getVRegValue()) << "\n");
//If this is the last instructions for the max iterations ago, don't update operands
if(inEpilogue.count(mOp.getVRegValue()))
if(inEpilogue[mOp.getVRegValue()] == i)
continue;
//Quickly write appropriate phis for this operand
if(newValues.count(mOp.getVRegValue())) {
if(newValues[mOp.getVRegValue()].count(i)) {
Instruction *tmp = new TmpInstruction(newValues[mOp.getVRegValue()][i]);
//Get machine code for this instruction
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
tempMvec.addTemp((Value*) tmp);
//assert of no kernelPHI for this value
assert(kernelPHIs[mOp.getVRegValue()][i] !=0 && "Must have final kernel phi to construct epilogue phi");
MachineInstr *saveValue = BuildMI(machineBB, V9::PHI, 3).addReg(newValues[mOp.getVRegValue()][i]).addReg(kernelPHIs[mOp.getVRegValue()][i]).addRegDef(tmp);
DEBUG(std::cerr << "Resulting PHI: " << *saveValue << "\n");
valPHIs[mOp.getVRegValue()] = tmp;
}
}
if(valPHIs.count(mOp.getVRegValue())) {
//Update the operand in the cloned instruction
clone->getOperand(opNum).setValueReg(valPHIs[mOp.getVRegValue()]);
}
}
else if((mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isDef())) {
inEpilogue[mOp.getVRegValue()] = i;
}
}
machineBB->push_back(clone);
}
}
}
MachineFunction *F = (((MachineBasicBlock*)origBB)->getParent());
MachineFunction::BasicBlockListType &BL = F->getBasicBlockList();
MachineFunction::BasicBlockListType::iterator BLI = (MachineBasicBlock*) origBB;
assert(BLI != BL.end() && "Must find original BB in machine function\n");
BL.insert(BLI,machineBB);
epilogues.push_back(machineBB);
llvm_epilogues.push_back(llvmBB);
DEBUG(std::cerr << "EPILOGUE #" << i << "\n");
DEBUG(machineBB->print(std::cerr));
}
}
void ModuloSchedulingPass::writeKernel(BasicBlock *llvmBB, MachineBasicBlock *machineBB, std::map<const Value*, std::pair<const MachineInstr*, int> > &valuesToSave, std::map<Value*, std::map<int, Value*> > &newValues, std::map<Value*, MachineBasicBlock*> &newValLocation, std::map<Value*, std::map<int, Value*> > &kernelPHIs) {
//Keep track of operands that are read and saved from a previous iteration. The new clone
//instruction will use the result of the phi instead.
std::map<Value*, Value*> finalPHIValue;
std::map<Value*, Value*> kernelValue;
//Branches are a special case
std::vector<MachineInstr*> branches;
//Get target information to look at machine operands
const TargetInstrInfo *mii = target.getInstrInfo();
//Create TmpInstructions for the final phis
for(MSSchedule::kernel_iterator I = schedule.kernel_begin(), E = schedule.kernel_end(); I != E; ++I) {
DEBUG(std::cerr << "Stage: " << I->second << " Inst: " << *(I->first) << "\n";);
//Clone instruction
const MachineInstr *inst = I->first;
MachineInstr *instClone = inst->clone();
//Insert into machine basic block
machineBB->push_back(instClone);
if(mii->isBranch(instClone->getOpcode()))
BuildMI(machineBB, V9::NOP, 0);
DEBUG(std::cerr << "Cloned Inst: " << *instClone << "\n");
//Loop over Machine Operands
for(unsigned i=0; i < inst->getNumOperands(); ++i) {
//get machine operand
const MachineOperand &mOp = inst->getOperand(i);
if(I->second != 0) {
if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isUse()) {
//Check to see where this operand is defined if this instruction is from max stage
if(I->second == schedule.getMaxStage()) {
DEBUG(std::cerr << "VREG: " << *(mOp.getVRegValue()) << "\n");
}
//If its in the value saved, we need to create a temp instruction and use that instead
if(valuesToSave.count(mOp.getVRegValue())) {
//Check if we already have a final PHI value for this
if(!finalPHIValue.count(mOp.getVRegValue())) {
//Only create phi if the operand def is from a stage before this one
if(schedule.defPreviousStage(mOp.getVRegValue(), I->second)) {
TmpInstruction *tmp = new TmpInstruction(mOp.getVRegValue());
//Get machine code for this instruction
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
tempMvec.addTemp((Value*) tmp);
//Update the operand in the cloned instruction
instClone->getOperand(i).setValueReg(tmp);
//save this as our final phi
finalPHIValue[mOp.getVRegValue()] = tmp;
newValLocation[tmp] = machineBB;
}
}
else {
//Use the previous final phi value
instClone->getOperand(i).setValueReg(finalPHIValue[mOp.getVRegValue()]);
}
}
}
}
if(I->second != schedule.getMaxStage()) {
if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isDef()) {
if(valuesToSave.count(mOp.getVRegValue())) {
TmpInstruction *tmp = new TmpInstruction(mOp.getVRegValue());
//Get machine code for this instruction
MachineCodeForInstruction & tempVec = MachineCodeForInstruction::get(defaultInst);
tempVec.addTemp((Value*) tmp);
//Create new machine instr and put in MBB
MachineInstr *saveValue;
if(mOp.getVRegValue()->getType() == Type::FloatTy)
saveValue = BuildMI(machineBB, V9::FMOVS, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else if(mOp.getVRegValue()->getType() == Type::DoubleTy)
saveValue = BuildMI(machineBB, V9::FMOVD, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else
saveValue = BuildMI(machineBB, V9::ORr, 3).addReg(mOp.getVRegValue()).addImm(0).addRegDef(tmp);
//Save for future cleanup
kernelValue[mOp.getVRegValue()] = tmp;
newValLocation[tmp] = machineBB;
kernelPHIs[mOp.getVRegValue()][schedule.getMaxStage()-1] = tmp;
}
}
}
}
}
//Add branches
for(std::vector<MachineInstr*>::iterator I = branches.begin(), E = branches.end(); I != E; ++I) {
machineBB->push_back(*I);
BuildMI(machineBB, V9::NOP, 0);
}
DEBUG(std::cerr << "KERNEL before PHIs\n");
DEBUG(machineBB->print(std::cerr));
//Loop over each value we need to generate phis for
for(std::map<Value*, std::map<int, Value*> >::iterator V = newValues.begin(),
E = newValues.end(); V != E; ++V) {
DEBUG(std::cerr << "Writing phi for" << *(V->first));
DEBUG(std::cerr << "\nMap of Value* for this phi\n");
DEBUG(for(std::map<int, Value*>::iterator I = V->second.begin(),
IE = V->second.end(); I != IE; ++I) {
std::cerr << "Stage: " << I->first;
std::cerr << " Value: " << *(I->second) << "\n";
});
//If we only have one current iteration live, its safe to set lastPhi = to kernel value
if(V->second.size() == 1) {
assert(kernelValue[V->first] != 0 && "Kernel value* must exist to create phi");
MachineInstr *saveValue = BuildMI(*machineBB, machineBB->begin(),V9::PHI, 3).addReg(V->second.begin()->second).addReg(kernelValue[V->first]).addRegDef(finalPHIValue[V->first]);
DEBUG(std::cerr << "Resulting PHI (one live): " << *saveValue << "\n");
kernelPHIs[V->first][V->second.begin()->first] = kernelValue[V->first];
DEBUG(std::cerr << "Put kernel phi in at stage: " << schedule.getMaxStage()-1 << " (map stage = " << V->second.begin()->first << ")\n");
}
else {
//Keep track of last phi created.
Instruction *lastPhi = 0;
unsigned count = 1;
//Loop over the the map backwards to generate phis
for(std::map<int, Value*>::reverse_iterator I = V->second.rbegin(), IE = V->second.rend();
I != IE; ++I) {
if(count < (V->second).size()) {
if(lastPhi == 0) {
lastPhi = new TmpInstruction(I->second);
//Get machine code for this instruction
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
tempMvec.addTemp((Value*) lastPhi);
MachineInstr *saveValue = BuildMI(*machineBB, machineBB->begin(), V9::PHI, 3).addReg(kernelValue[V->first]).addReg(I->second).addRegDef(lastPhi);
DEBUG(std::cerr << "Resulting PHI: " << *saveValue << "\n");
newValLocation[lastPhi] = machineBB;
}
else {
Instruction *tmp = new TmpInstruction(I->second);
//Get machine code for this instruction
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
tempMvec.addTemp((Value*) tmp);
MachineInstr *saveValue = BuildMI(*machineBB, machineBB->begin(), V9::PHI, 3).addReg(lastPhi).addReg(I->second).addRegDef(tmp);
DEBUG(std::cerr << "Resulting PHI: " << *saveValue << "\n");
lastPhi = tmp;
kernelPHIs[V->first][I->first] = lastPhi;
newValLocation[lastPhi] = machineBB;
}
}
//Final phi value
else {
//The resulting value must be the Value* we created earlier
assert(lastPhi != 0 && "Last phi is NULL!\n");
MachineInstr *saveValue = BuildMI(*machineBB, machineBB->begin(), V9::PHI, 3).addReg(lastPhi).addReg(I->second).addRegDef(finalPHIValue[V->first]);
DEBUG(std::cerr << "Resulting PHI: " << *saveValue << "\n");
kernelPHIs[V->first][I->first] = finalPHIValue[V->first];
}
++count;
}
}
}
DEBUG(std::cerr << "KERNEL after PHIs\n");
DEBUG(machineBB->print(std::cerr));
}
void ModuloSchedulingPass::removePHIs(const MachineBasicBlock *origBB, std::vector<MachineBasicBlock *> &prologues, std::vector<MachineBasicBlock *> &epilogues, MachineBasicBlock *kernelBB, std::map<Value*, MachineBasicBlock*> &newValLocation) {
//Worklist to delete things
std::vector<std::pair<MachineBasicBlock*, MachineBasicBlock::iterator> > worklist;
//Worklist of TmpInstructions that need to be added to a MCFI
std::vector<Instruction*> addToMCFI;
//Worklist to add OR instructions to end of kernel so not to invalidate the iterator
//std::vector<std::pair<Instruction*, Value*> > newORs;
const TargetInstrInfo *TMI = target.getInstrInfo();
//Start with the kernel and for each phi insert a copy for the phi def and for each arg
for(MachineBasicBlock::iterator I = kernelBB->begin(), E = kernelBB->end(); I != E; ++I) {
DEBUG(std::cerr << "Looking at Instr: " << *I << "\n");
//Get op code and check if its a phi
if(I->getOpcode() == V9::PHI) {
DEBUG(std::cerr << "Replacing PHI: " << *I << "\n");
Instruction *tmp = 0;
for(unsigned i = 0; i < I->getNumOperands(); ++i) {
//Get Operand
const MachineOperand &mOp = I->getOperand(i);
assert(mOp.getType() == MachineOperand::MO_VirtualRegister && "Should be a Value*\n");
if(!tmp) {
tmp = new TmpInstruction(mOp.getVRegValue());
addToMCFI.push_back(tmp);
}
//Now for all our arguments we read, OR to the new TmpInstruction that we created
if(mOp.isUse()) {
DEBUG(std::cerr << "Use: " << mOp << "\n");
//Place a copy at the end of its BB but before the branches
assert(newValLocation.count(mOp.getVRegValue()) && "We must know where this value is located\n");
//Reverse iterate to find the branches, we can safely assume no instructions have been
//put in the nop positions
for(MachineBasicBlock::iterator inst = --(newValLocation[mOp.getVRegValue()])->end(), endBB = (newValLocation[mOp.getVRegValue()])->begin(); inst != endBB; --inst) {
MachineOpCode opc = inst->getOpcode();
if(TMI->isBranch(opc) || TMI->isNop(opc))
continue;
else {
if(mOp.getVRegValue()->getType() == Type::FloatTy)
BuildMI(*(newValLocation[mOp.getVRegValue()]), ++inst, V9::FMOVS, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else if(mOp.getVRegValue()->getType() == Type::DoubleTy)
BuildMI(*(newValLocation[mOp.getVRegValue()]), ++inst, V9::FMOVD, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else
BuildMI(*(newValLocation[mOp.getVRegValue()]), ++inst, V9::ORr, 3).addReg(mOp.getVRegValue()).addImm(0).addRegDef(tmp);
break;
}
}
}
else {
//Remove the phi and replace it with an OR
DEBUG(std::cerr << "Def: " << mOp << "\n");
//newORs.push_back(std::make_pair(tmp, mOp.getVRegValue()));
if(tmp->getType() == Type::FloatTy)
BuildMI(*kernelBB, I, V9::FMOVS, 3).addReg(tmp).addRegDef(mOp.getVRegValue());
else if(tmp->getType() == Type::DoubleTy)
BuildMI(*kernelBB, I, V9::FMOVD, 3).addReg(tmp).addRegDef(mOp.getVRegValue());
else
BuildMI(*kernelBB, I, V9::ORr, 3).addReg(tmp).addImm(0).addRegDef(mOp.getVRegValue());
worklist.push_back(std::make_pair(kernelBB, I));
}
}
}
}
//Add TmpInstructions to some MCFI
if(addToMCFI.size() > 0) {
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
for(unsigned x = 0; x < addToMCFI.size(); ++x) {
tempMvec.addTemp(addToMCFI[x]);
}
addToMCFI.clear();
}
//Remove phis from epilogue
for(std::vector<MachineBasicBlock*>::iterator MB = epilogues.begin(), ME = epilogues.end(); MB != ME; ++MB) {
for(MachineBasicBlock::iterator I = (*MB)->begin(), E = (*MB)->end(); I != E; ++I) {
DEBUG(std::cerr << "Looking at Instr: " << *I << "\n");
//Get op code and check if its a phi
if(I->getOpcode() == V9::PHI) {
Instruction *tmp = 0;
for(unsigned i = 0; i < I->getNumOperands(); ++i) {
//Get Operand
const MachineOperand &mOp = I->getOperand(i);
assert(mOp.getType() == MachineOperand::MO_VirtualRegister && "Should be a Value*\n");
if(!tmp) {
tmp = new TmpInstruction(mOp.getVRegValue());
addToMCFI.push_back(tmp);
}
//Now for all our arguments we read, OR to the new TmpInstruction that we created
if(mOp.isUse()) {
DEBUG(std::cerr << "Use: " << mOp << "\n");
//Place a copy at the end of its BB but before the branches
assert(newValLocation.count(mOp.getVRegValue()) && "We must know where this value is located\n");
//Reverse iterate to find the branches, we can safely assume no instructions have been
//put in the nop positions
for(MachineBasicBlock::iterator inst = --(newValLocation[mOp.getVRegValue()])->end(), endBB = (newValLocation[mOp.getVRegValue()])->begin(); inst != endBB; --inst) {
MachineOpCode opc = inst->getOpcode();
if(TMI->isBranch(opc) || TMI->isNop(opc))
continue;
else {
if(mOp.getVRegValue()->getType() == Type::FloatTy)
BuildMI(*(newValLocation[mOp.getVRegValue()]), ++inst, V9::FMOVS, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else if(mOp.getVRegValue()->getType() == Type::DoubleTy)
BuildMI(*(newValLocation[mOp.getVRegValue()]), ++inst, V9::FMOVD, 3).addReg(mOp.getVRegValue()).addRegDef(tmp);
else
BuildMI(*(newValLocation[mOp.getVRegValue()]), ++inst, V9::ORr, 3).addReg(mOp.getVRegValue()).addImm(0).addRegDef(tmp);
break;
}
}
}
else {
//Remove the phi and replace it with an OR
DEBUG(std::cerr << "Def: " << mOp << "\n");
if(tmp->getType() == Type::FloatTy)
BuildMI(**MB, I, V9::FMOVS, 3).addReg(tmp).addRegDef(mOp.getVRegValue());
else if(tmp->getType() == Type::DoubleTy)
BuildMI(**MB, I, V9::FMOVD, 3).addReg(tmp).addRegDef(mOp.getVRegValue());
else
BuildMI(**MB, I, V9::ORr, 3).addReg(tmp).addImm(0).addRegDef(mOp.getVRegValue());
worklist.push_back(std::make_pair(*MB,I));
}
}
}
}
}
if(addToMCFI.size() > 0) {
MachineCodeForInstruction & tempMvec = MachineCodeForInstruction::get(defaultInst);
for(unsigned x = 0; x < addToMCFI.size(); ++x) {
tempMvec.addTemp(addToMCFI[x]);
}
addToMCFI.clear();
}
//Delete the phis
for(std::vector<std::pair<MachineBasicBlock*, MachineBasicBlock::iterator> >::iterator I = worklist.begin(), E = worklist.end(); I != E; ++I) {
DEBUG(std::cerr << "Deleting PHI " << *I->second << "\n");
I->first->erase(I->second);
}
assert((addToMCFI.size() == 0) && "We should have added all TmpInstructions to some MachineCodeForInstruction");
}
void ModuloSchedulingPass::reconstructLoop(MachineBasicBlock *BB) {
TIME_REGION(X, "reconstructLoop");
DEBUG(std::cerr << "Reconstructing Loop\n");
//First find the value *'s that we need to "save"
std::map<const Value*, std::pair<const MachineInstr*, int> > valuesToSave;
//Keep track of instructions we have already seen and their stage because
//we don't want to "save" values if they are used in the kernel immediately
std::map<const MachineInstr*, int> lastInstrs;
std::map<const Value*, int> phiUses;
//Loop over kernel and only look at instructions from a stage > 0
//Look at its operands and save values *'s that are read
for(MSSchedule::kernel_iterator I = schedule.kernel_begin(), E = schedule.kernel_end(); I != E; ++I) {
if(I->second !=0) {
//For this instruction, get the Value*'s that it reads and put them into the set.
//Assert if there is an operand of another type that we need to save
const MachineInstr *inst = I->first;
lastInstrs[inst] = I->second;
for(unsigned i=0; i < inst->getNumOperands(); ++i) {
//get machine operand
const MachineOperand &mOp = inst->getOperand(i);
if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isUse()) {
//find the value in the map
if (const Value* srcI = mOp.getVRegValue()) {
if(isa<Constant>(srcI) || isa<Argument>(srcI))
continue;
//Before we declare this Value* one that we should save
//make sure its def is not of the same stage as this instruction
//because it will be consumed before its used
Instruction *defInst = (Instruction*) srcI;
//Should we save this value?
bool save = true;
//Continue if not in the def map, loop invariant code does not need to be saved
if(!defMap.count(srcI))
continue;
MachineInstr *defInstr = defMap[srcI];
if(lastInstrs.count(defInstr)) {
if(lastInstrs[defInstr] == I->second) {
save = false;
}
}
if(save) {
assert(!phiUses.count(srcI) && "Did not expect to see phi use twice");
if(isa<PHINode>(srcI))
phiUses[srcI] = I->second;
valuesToSave[srcI] = std::make_pair(I->first, i);
}
}
}
else if(mOp.getType() == MachineOperand::MO_VirtualRegister && mOp.isDef()) {
if (const Value* destI = mOp.getVRegValue()) {
if(!isa<PHINode>(destI))
continue;
if(phiUses.count(destI)) {
if(phiUses[destI] == I->second) {
//remove from save list
valuesToSave.erase(destI);
}
}
}
}
if(mOp.getType() != MachineOperand::MO_VirtualRegister && mOp.isUse()) {
assert("Our assumption is wrong. We have another type of register that needs to be saved\n");
}
}
}
}
//The new loop will consist of one or more prologues, the kernel, and one or more epilogues.
//Map to keep track of old to new values
std::map<Value*, std::map<int, Value*> > newValues;
//Map to keep track of old to new values in kernel
std::map<Value*, std::map<int, Value*> > kernelPHIs;
//Another map to keep track of what machine basic blocks these new value*s are in since
//they have no llvm instruction equivalent
std::map<Value*, MachineBasicBlock*> newValLocation;
std::vector<MachineBasicBlock*> prologues;
std::vector<BasicBlock*> llvm_prologues;
//Write prologue
if(schedule.getMaxStage() != 0)
writePrologues(prologues, BB, llvm_prologues, valuesToSave, newValues, newValLocation);
//Print out epilogues and prologue
DEBUG(for(std::vector<MachineBasicBlock*>::iterator I = prologues.begin(), E = prologues.end();
I != E; ++I) {
std::cerr << "PROLOGUE\n";
(*I)->print(std::cerr);
});
BasicBlock *llvmKernelBB = new BasicBlock("Kernel", (Function*) (BB->getBasicBlock()->getParent()));
MachineBasicBlock *machineKernelBB = new MachineBasicBlock(llvmKernelBB);
MachineFunction *F = (((MachineBasicBlock*)BB)->getParent());
MachineFunction::BasicBlockListType &BL = F->getBasicBlockList();
MachineFunction::BasicBlockListType::iterator BLI = BB;
assert(BLI != BL.end() && "Must find original BB in machine function\n");
BL.insert(BLI,machineKernelBB);
//(((MachineBasicBlock*)BB)->getParent())->getBasicBlockList().push_back(machineKernelBB);
writeKernel(llvmKernelBB, machineKernelBB, valuesToSave, newValues, newValLocation, kernelPHIs);
std::vector<MachineBasicBlock*> epilogues;
std::vector<BasicBlock*> llvm_epilogues;
//Write epilogues
if(schedule.getMaxStage() != 0)
writeEpilogues(epilogues, BB, llvm_epilogues, valuesToSave, newValues, newValLocation, kernelPHIs);
//Fix our branches
fixBranches(prologues, llvm_prologues, machineKernelBB, llvmKernelBB, epilogues, llvm_epilogues, BB);
//Remove phis
removePHIs(BB, prologues, epilogues, machineKernelBB, newValLocation);
//Print out epilogues and prologue
DEBUG(for(std::vector<MachineBasicBlock*>::iterator I = prologues.begin(), E = prologues.end();
I != E; ++I) {
std::cerr << "PROLOGUE\n";
(*I)->print(std::cerr);
});
DEBUG(std::cerr << "KERNEL\n");
DEBUG(machineKernelBB->print(std::cerr));
DEBUG(for(std::vector<MachineBasicBlock*>::iterator I = epilogues.begin(), E = epilogues.end();
I != E; ++I) {
std::cerr << "EPILOGUE\n";
(*I)->print(std::cerr);
});
DEBUG(std::cerr << "New Machine Function" << "\n");
DEBUG(std::cerr << BB->getParent() << "\n");
}
void ModuloSchedulingPass::fixBranches(std::vector<MachineBasicBlock *> &prologues, std::vector<BasicBlock*> &llvm_prologues, MachineBasicBlock *machineKernelBB, BasicBlock *llvmKernelBB, std::vector<MachineBasicBlock *> &epilogues, std::vector<BasicBlock*> &llvm_epilogues, MachineBasicBlock *BB) {
const TargetInstrInfo *TMI = target.getInstrInfo();
if(schedule.getMaxStage() != 0) {
//Fix prologue branches
for(unsigned I = 0; I < prologues.size(); ++I) {
//Find terminator since getFirstTerminator does not work!
for(MachineBasicBlock::reverse_iterator mInst = prologues[I]->rbegin(), mInstEnd = prologues[I]->rend(); mInst != mInstEnd; ++mInst) {
MachineOpCode OC = mInst->getOpcode();
//If its a branch update its branchto
if(TMI->isBranch(OC)) {
for(unsigned opNum = 0; opNum < mInst->getNumOperands(); ++opNum) {
MachineOperand &mOp = mInst->getOperand(opNum);
if (mOp.getType() == MachineOperand::MO_PCRelativeDisp) {
//Check if we are branching to the kernel, if not branch to epilogue
if(mOp.getVRegValue() == BB->getBasicBlock()) {
if(I == prologues.size()-1)
mOp.setValueReg(llvmKernelBB);
else
mOp.setValueReg(llvm_prologues[I+1]);
}
else {
mOp.setValueReg(llvm_epilogues[(llvm_epilogues.size()-1-I)]);
}
}
}
DEBUG(std::cerr << "New Prologue Branch: " << *mInst << "\n");
}
}
//Update llvm basic block with our new branch instr
DEBUG(std::cerr << BB->getBasicBlock()->getTerminator() << "\n");
const BranchInst *branchVal = dyn_cast<BranchInst>(BB->getBasicBlock()->getTerminator());
if(I == prologues.size()-1) {
TerminatorInst *newBranch = new BranchInst(llvmKernelBB,
llvm_epilogues[(llvm_epilogues.size()-1-I)],
branchVal->getCondition(),
llvm_prologues[I]);
}
else
TerminatorInst *newBranch = new BranchInst(llvm_prologues[I+1],
llvm_epilogues[(llvm_epilogues.size()-1-I)],
branchVal->getCondition(),
llvm_prologues[I]);
}
}
Value *origBranchExit = 0;
//Fix up kernel machine branches
for(MachineBasicBlock::reverse_iterator mInst = machineKernelBB->rbegin(), mInstEnd = machineKernelBB->rend(); mInst != mInstEnd; ++mInst) {
MachineOpCode OC = mInst->getOpcode();
if(TMI->isBranch(OC)) {
for(unsigned opNum = 0; opNum < mInst->getNumOperands(); ++opNum) {
MachineOperand &mOp = mInst->getOperand(opNum);
if(mOp.getType() == MachineOperand::MO_PCRelativeDisp) {
if(mOp.getVRegValue() == BB->getBasicBlock())
mOp.setValueReg(llvmKernelBB);
else
if(llvm_epilogues.size() > 0) {
assert(origBranchExit == 0 && "There should only be one branch out of the loop");
origBranchExit = mOp.getVRegValue();
mOp.setValueReg(llvm_epilogues[0]);
}
else
origBranchExit = mOp.getVRegValue();
}
}
}
}
//Update kernelLLVM branches
const BranchInst *branchVal = dyn_cast<BranchInst>(BB->getBasicBlock()->getTerminator());
assert(origBranchExit != 0 && "We must have the original bb the kernel exits to!");
if(epilogues.size() > 0) {
TerminatorInst *newBranch = new BranchInst(llvmKernelBB,
llvm_epilogues[0],
branchVal->getCondition(),
llvmKernelBB);
}
else {
BasicBlock *origBBExit = dyn_cast<BasicBlock>(origBranchExit);
assert(origBBExit !=0 && "Original exit basic block must be set");
TerminatorInst *newBranch = new BranchInst(llvmKernelBB,
origBBExit,
branchVal->getCondition(),
llvmKernelBB);
}
if(schedule.getMaxStage() != 0) {
//Lastly add unconditional branches for the epilogues
for(unsigned I = 0; I < epilogues.size(); ++I) {
//Now since we don't have fall throughs, add a unconditional branch to the next prologue
if(I != epilogues.size()-1) {
BuildMI(epilogues[I], V9::BA, 1).addPCDisp(llvm_epilogues[I+1]);
//Add unconditional branch to end of epilogue
TerminatorInst *newBranch = new BranchInst(llvm_epilogues[I+1],
llvm_epilogues[I]);
}
else {
BuildMI(epilogues[I], V9::BA, 1).addPCDisp(origBranchExit);
//Update last epilogue exit branch
BranchInst *branchVal = (BranchInst*) dyn_cast<BranchInst>(BB->getBasicBlock()->getTerminator());
//Find where we are supposed to branch to
BasicBlock *nextBlock = 0;
for(unsigned j=0; j <branchVal->getNumSuccessors(); ++j) {
if(branchVal->getSuccessor(j) != BB->getBasicBlock())
nextBlock = branchVal->getSuccessor(j);
}
assert((nextBlock != 0) && "Next block should not be null!");
TerminatorInst *newBranch = new BranchInst(nextBlock, llvm_epilogues[I]);
}
//Add one more nop!
BuildMI(epilogues[I], V9::NOP, 0);
}
}
//FIX UP Machine BB entry!!
//We are looking at the predecesor of our loop basic block and we want to change its ba instruction
//Find all llvm basic blocks that branch to the loop entry and change to our first prologue.
const BasicBlock *llvmBB = BB->getBasicBlock();
std::vector<const BasicBlock*>Preds (pred_begin(llvmBB), pred_end(llvmBB));
//for(pred_const_iterator P = pred_begin(llvmBB), PE = pred_end(llvmBB); P != PE; ++PE) {
for(std::vector<const BasicBlock*>::iterator P = Preds.begin(), PE = Preds.end(); P != PE; ++P) {
if(*P == llvmBB)
continue;
else {
DEBUG(std::cerr << "Found our entry BB\n");
//Get the Terminator instruction for this basic block and print it out
DEBUG(std::cerr << *((*P)->getTerminator()) << "\n");
//Update the terminator
TerminatorInst *term = ((BasicBlock*)*P)->getTerminator();
for(unsigned i=0; i < term->getNumSuccessors(); ++i) {
if(term->getSuccessor(i) == llvmBB) {
DEBUG(std::cerr << "Replacing successor bb\n");
if(llvm_prologues.size() > 0) {
term->setSuccessor(i, llvm_prologues[0]);
//Also update its corresponding machine instruction
MachineCodeForInstruction & tempMvec =
MachineCodeForInstruction::get(term);
for (unsigned j = 0; j < tempMvec.size(); j++) {
MachineInstr *temp = tempMvec[j];
MachineOpCode opc = temp->getOpcode();
if(TMI->isBranch(opc)) {
DEBUG(std::cerr << *temp << "\n");
//Update branch
for(unsigned opNum = 0; opNum < temp->getNumOperands(); ++opNum) {
MachineOperand &mOp = temp->getOperand(opNum);
if (mOp.getType() == MachineOperand::MO_PCRelativeDisp) {
if(mOp.getVRegValue() == llvmBB)
mOp.setValueReg(llvm_prologues[0]);
}
}
}
}
}
else {
term->setSuccessor(i, llvmKernelBB);
//Also update its corresponding machine instruction
MachineCodeForInstruction & tempMvec =
MachineCodeForInstruction::get(term);
for (unsigned j = 0; j < tempMvec.size(); j++) {
MachineInstr *temp = tempMvec[j];
MachineOpCode opc = temp->getOpcode();
if(TMI->isBranch(opc)) {
DEBUG(std::cerr << *temp << "\n");
//Update branch
for(unsigned opNum = 0; opNum < temp->getNumOperands(); ++opNum) {
MachineOperand &mOp = temp->getOperand(opNum);
if (mOp.getType() == MachineOperand::MO_PCRelativeDisp) {
if(mOp.getVRegValue() == llvmBB)
mOp.setValueReg(llvmKernelBB);
}
}
}
}
}
}
}
break;
}
}
//BB->getParent()->getBasicBlockList().erase(BB);
}