You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
nudt-compiler-cpp/src/ir/passes/CFGSimplify.cpp

238 lines
7.3 KiB

This file contains ambiguous Unicode characters!

This file contains ambiguous Unicode characters that may be confused with others in your current locale. If your use case is intentional and legitimate, you can safely ignore this warning. Use the Escape button to highlight these characters.

// CFG 简化:
// - 删除不可达块、合并空块、简化分支等
// - 改善 IR 结构,便于后续优化与后端生成
#include "ir/IR.h"
#include <algorithm>
#include <unordered_map>
#include <unordered_set>
#include <vector>
namespace ir {
namespace {
// 从入口块开始进行 BFS/DFS标记所有可达的基本块
std::unordered_set<BasicBlock*> FindReachableBlocks(Function* func) {
std::unordered_set<BasicBlock*> reachable;
std::vector<BasicBlock*> worklist;
auto* entry = func->GetEntry();
if (!entry) return reachable;
reachable.insert(entry);
worklist.push_back(entry);
while (!worklist.empty()) {
auto* bb = worklist.back();
worklist.pop_back();
// 遍历该块的所有后继块
for (const auto& inst_ptr : bb->GetInstructions()) {
auto* inst = inst_ptr.get();
// 无条件分支:跳转到目标块
if (auto* br = dynamic_cast<BranchInst*>(inst)) {
auto* target = br->GetTarget();
if (reachable.insert(target).second) {
worklist.push_back(target);
}
}
// 条件分支:跳转到 true 和 false 目标
else if (auto* condbr = dynamic_cast<CondBranchInst*>(inst)) {
auto* true_target = condbr->GetTrueTarget();
auto* false_target = condbr->GetFalseTarget();
if (reachable.insert(true_target).second) {
worklist.push_back(true_target);
}
if (reachable.insert(false_target).second) {
worklist.push_back(false_target);
}
}
// 终止指令,没有后继
if (inst->IsTerminator()) break;
}
}
return reachable;
}
// 获取基本块的 terminator 指令
Instruction* GetTerminator(BasicBlock* bb) {
const auto& insts = bb->GetInstructions();
if (insts.empty()) return nullptr;
return insts.back().get();
}
} // namespace
void RunCFGSimplify(Module& module) {
for (auto& func_ptr : module.GetFunctions()) {
auto* func = func_ptr.get();
if (func->IsExternal()) continue;
auto& blocks = const_cast<std::vector<std::unique_ptr<BasicBlock>>&>(func->GetBlocks());
bool changed = true;
while (changed) {
changed = false;
auto reachable = FindReachableBlocks(func);
std::unordered_set<BasicBlock*> unreachable;
for (auto& bb : blocks) {
if (reachable.find(bb.get()) == reachable.end()) {
unreachable.insert(bb.get());
}
}
for (auto& bb_ptr : blocks) {
auto* bb = bb_ptr.get();
if (unreachable.find(bb) != unreachable.end()) continue;
std::vector<std::pair<PhiInst*, Value*>> phi_replacements;
std::vector<PhiInst*> phi_to_delete;
for (auto& inst_ptr : bb->GetInstructions()) {
auto* phi = dynamic_cast<PhiInst*>(inst_ptr.get());
if (!phi) break;
Value* valid_val = nullptr;
size_t valid_count = 0;
for (size_t i = 0; i < phi->GetNumOperands(); i += 2) {
auto* val = phi->GetOperand(i);
auto* pred = static_cast<BasicBlock*>(phi->GetOperand(i + 1));
if (unreachable.find(pred) == unreachable.end()) {
valid_val = val;
valid_count++;
}
}
if (valid_count == 1 && valid_val) {
phi_replacements.push_back({phi, valid_val});
} else if (valid_count == 0) {
phi_to_delete.push_back(phi);
}
}
for (auto& [phi, val] : phi_replacements) {
phi->ReplaceAllUsesWith(val);
phi_to_delete.push_back(phi);
}
for (auto* phi : phi_to_delete) {
for (size_t i = 0; i < phi->GetNumOperands(); ++i) {
if (auto* op = phi->GetOperand(i)) {
op->RemoveUse(phi, i);
}
}
}
auto& insts = const_cast<std::vector<std::unique_ptr<Instruction>>&>(bb->GetInstructions());
auto new_end = std::remove_if(insts.begin(), insts.end(),
[&phi_to_delete](const std::unique_ptr<Instruction>& inst_ptr) {
return std::find(phi_to_delete.begin(), phi_to_delete.end(), inst_ptr.get()) != phi_to_delete.end();
}
);
insts.erase(new_end, insts.end());
}
for (auto& bb_ptr : blocks) {
if (unreachable.find(bb_ptr.get()) != unreachable.end()) {
for (auto& inst_ptr : bb_ptr->GetInstructions()) {
auto* inst = inst_ptr.get();
for (size_t i = 0; i < inst->GetNumOperands(); ++i) {
if (auto* op = inst->GetOperand(i)) {
op->RemoveUse(inst, i);
}
}
}
}
}
size_t old_size = blocks.size();
blocks.erase(
std::remove_if(blocks.begin(), blocks.end(),
[&reachable](const std::unique_ptr<BasicBlock>& bb_ptr) {
return reachable.find(bb_ptr.get()) == reachable.end();
}
),
blocks.end()
);
if (blocks.size() != old_size) {
changed = true;
}
for (auto& bb_ptr : blocks) {
auto* bb = bb_ptr.get();
auto* term = GetTerminator(bb);
if (!term) continue;
if (auto* condbr = dynamic_cast<CondBranchInst*>(term)) {
auto* cond = condbr->GetCond();
if (auto* const_int = dynamic_cast<ConstantInt*>(cond)) {
auto* true_target = condbr->GetTrueTarget();
auto* false_target = condbr->GetFalseTarget();
auto* target = (const_int->GetValue() != 0) ? true_target : false_target;
bb->RemoveInstruction(condbr);
bb->Append<BranchInst>(Type::GetVoidType(), target);
changed = true;
}
}
}
}
for (auto& bb_ptr : blocks) {
auto* bb = bb_ptr.get();
const auto& preds = bb->GetPredecessors();
if (preds.size() == 1) {
std::vector<std::pair<PhiInst*, Value*>> phi_replacements;
for (auto& inst_ptr : bb->GetInstructions()) {
auto* phi = dynamic_cast<PhiInst*>(inst_ptr.get());
if (!phi) break;
if (phi->GetNumOperands() >= 2) {
Value* incoming_val = phi->GetOperand(0);
phi_replacements.push_back({phi, incoming_val});
}
}
for (auto& [phi, val] : phi_replacements) {
phi->ReplaceAllUsesWith(val);
}
std::vector<PhiInst*> phis_to_clean;
for (auto& inst_ptr : bb->GetInstructions()) {
auto* phi = dynamic_cast<PhiInst*>(inst_ptr.get());
if (!phi) break;
phis_to_clean.push_back(phi);
}
for (auto* phi : phis_to_clean) {
for (size_t i = 0; i < phi->GetNumOperands(); ++i) {
if (auto* op = phi->GetOperand(i)) {
op->RemoveUse(phi, i);
}
}
}
auto& insts = const_cast<std::vector<std::unique_ptr<Instruction>>&>(bb->GetInstructions());
auto new_end = std::remove_if(insts.begin(), insts.end(),
[](const std::unique_ptr<Instruction>& inst_ptr) {
return dynamic_cast<const PhiInst*>(inst_ptr.get()) != nullptr;
}
);
insts.erase(new_end, insts.end());
}
}
}
}
} // namespace ir