add codegen summary

.vscode/launch.json

@@ -3,15 +3,15 @@
         {
             "type": "lldb",
             "request": "launch",
-			"name": "Compile & Debug 1.ll",
+            "name": "Compile & Gen 1.ll",
             // 要调试的程序
             "program": "${workspaceFolder}/build/cminusfc",
             // 命令行参数
             "args": [
-				"-o",
-				"./build/1.ll",
                 "-emit-llvm",
-				"./build/1.cminus"
+                "./build/1.cminus",
+                "-o",
+                "./build/1.ll"
             ],
             // 程序运行的目录
             "cwd": "${workspaceFolder}",
@@ -24,21 +24,42 @@
         {
             "type": "lldb",
             "request": "launch",
-			"name": "Debug 1.ll",
+            "name": "Gen 1.ll",
             // 要调试的程序
             "program": "${workspaceFolder}/build/cminusfc",
             // 命令行参数
             "args": [
-				"-o",
-				"./build/1.ll",
                 "-emit-llvm",
-				"./build/1.cminus"
+                "./build/1.cminus",
+                "-o",
+                "./build/1.ll"
             ],
             // 程序运行的目录
             "cwd": "${workspaceFolder}",
             "initCommands": [
                 "command script import ${workspaceFolder}/lldb_formatters.py"
             ]
+        },
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "Compile & Gen 1.s",
+            // 要调试的程序
+            "program": "${workspaceFolder}/build/cminusfc",
+            // 命令行参数
+            "args": [
+                "-S",
+                "./build/1.cminus",
+                "-o",
+                "./build/1.s"
+            ],
+            // 程序运行的目录
+            "cwd": "${workspaceFolder}",
+            // 程序运行前运行的命令(例如 build)
+            "preLaunchTask": "make cminusfc",
+            "initCommands": [
+                "command script import ${workspaceFolder}/lldb_formatters.py"
+            ]
         }
     ]
 }
\ No newline at end of file

include/codegen/ASMInstruction.hpp

@@ -4,7 +4,7 @@
 #include <string>
 
 struct ASMInstruction {
-    enum InstType { Instruction, Atrribute, Label, Comment } type;
+    enum InstType { Instruction, Attribute, Label, Comment } type;
     std::string content;
 
     explicit ASMInstruction(std::string s, InstType ty = Instruction)
@@ -13,7 +13,7 @@ struct ASMInstruction {
     std::string format() const {
         switch (type) {
         case ASMInstruction::Instruction:
-        case ASMInstruction::Atrribute:
+        case ASMInstruction::Attribute:
             return "\t" + content + "\n";
         case ASMInstruction::Label:
             return content + ":\n";
@@ -22,4 +22,7 @@ struct ASMInstruction {
         }
         assert(false && "unreachable");
     }
+
+    // 用于 lldb 调试生成 summary
+    std::string safe_print() const;
 };

include/codegen/CodeGen.hpp

@@ -8,13 +8,16 @@
 class CodeGen {
   public:
     explicit CodeGen(Module *module) : m(module) {}
+    ~CodeGen(){
+        for(auto i : output) delete i;
+    }
 
     std::string print() const;
 
     void run();
 
     template <class... Args> void append_inst(Args... arg) {
-        output.emplace_back(arg...);
+        output.emplace_back(new ASMInstruction(arg...));
     }
 
     void
@@ -26,7 +29,7 @@ class CodeGen {
         }
         content.pop_back();
         content.pop_back();
-        output.emplace_back(content, ty);
+        output.emplace_back(new ASMInstruction(content, ty));
     }
 
   private:
@@ -98,5 +101,5 @@ class CodeGen {
     } context;
 
     Module *m;
-    std::list<ASMInstruction> output;
+    std::list<ASMInstruction*> output;
 };

include/codegen/Register.hpp

@@ -32,6 +32,9 @@ struct Reg {
 
     std::string print() const;
 
+    // 用于 lldb 调试生成 summary
+    std::string safe_print() const;
+
     static Reg zero() { return Reg(0); }
     static Reg ra() { return Reg(1); }
     static Reg sp() { return Reg(3); }
@@ -62,17 +65,20 @@ struct FReg {
     std::string print() const;
 
     static FReg fa(unsigned i) {
-        assert(0 <= i and i <= 7);
+        assert(i <= 7);
         return FReg(i);
     }
     static FReg ft(unsigned i) {
-        assert(0 <= i and i <= 15);
+        assert(i <= 15);
         return FReg(i + 8);
     }
     static FReg fs(unsigned i) {
-        assert(0 <= i and i <= 7);
+        assert(i <= 7);
         return FReg(i + 24);
     }
+
+    // 用于 lldb 调试生成 summary
+    std::string safe_print() const;
 };
 
 struct CFReg {
@@ -82,4 +88,7 @@ struct CFReg {
     bool operator==(const CFReg &other) { return id == other.id; }
 
     std::string print() const { return "$fcc" + std::to_string(id); }
+
+    // 用于 lldb 调试生成 summary
+    std::string safe_print() const;
 };

include/lightir/Function.hpp

@@ -40,7 +40,7 @@ class Function : public Value {
     BasicBlock *get_entry_block() const { return basic_blocks_.front(); }
 
     std::list<BasicBlock*> &get_basic_blocks() { return basic_blocks_; }
-    std::list<Argument> &get_args() { return arguments_; }
+    std::list<Argument*> &get_args() { return arguments_; }
 
     bool is_declaration() const { return basic_blocks_.empty(); }
 
@@ -58,7 +58,7 @@ class Function : public Value {
 
   private:
     std::list<BasicBlock*> basic_blocks_;
-    std::list<Argument> arguments_;
+    std::list<Argument*> arguments_;
     Module *parent_;
     unsigned seq_cnt_; // print use
 };

include/lightir/Names.hpp

@@ -6,8 +6,8 @@
 class Names
 {
 public:
-	explicit Names(bool use_underline, std::string prefix, std::string append)
-		: use_underline_(use_underline), default_prefix_used_count_(0), default_prefix_(std::move(prefix)),
+    explicit Names(std::string prefix, std::string append)
+        : default_prefix_used_count_(0), default_prefix_(std::move(prefix)),
           appended_prefix_(std::move(append))
     {
     }
@@ -18,7 +18,6 @@ public:
     std::string get_name();
 
 private:
-	bool use_underline_;
     int default_prefix_used_count_;
     std::string default_prefix_;
     std::string appended_prefix_;

lldb_formatters.py

@@ -9,7 +9,13 @@ def parseString(val : lldb.SBValue):
     return ""
 
 def SafePrintSummary(valobj: lldb.SBValue, internal_dict):
-    name: lldb.SBValue = valobj.EvaluateExpression("this->safe_print()")
+    if valobj.GetType().IsPointerType():
+        deref: lldb.SBValue = valobj.Dereference()
+        if deref.IsValid():
+            valobj = deref
+        else:
+            return ""
+    name: lldb.SBValue = valobj.EvaluateExpression("safe_print()")
     return parseString(name)
 
 def __lldb_init_module(debugger: lldb.SBDebugger, internal_dict):
@@ -18,7 +24,8 @@ def __lldb_init_module(debugger: lldb.SBDebugger, internal_dict):
              , "Function", "Argument"
              , "BasicBlock"
              , "GlobalVariable"
-             , "Instruction", "IBinaryInst", "FBinaryInst", "ICmpInst", "FCmpInst", "CallInst", "BranchInst", "ReturnInst", "GetElementPtrInst", "StoreInst", "LoadInst", "AllocaInst", "ZextInst", "FpToSiInst", "SiToFpInst", "PhiInst" ]
+             , "Instruction", "IBinaryInst", "FBinaryInst", "ICmpInst", "FCmpInst", "CallInst", "BranchInst", "ReturnInst", "GetElementPtrInst", "StoreInst", "LoadInst", "AllocaInst", "ZextInst", "FpToSiInst", "SiToFpInst", "PhiInst"
+             , "ASMInstruction", "Reg", "FReg", "CFReg" ]
     for i in types:
         debugger.HandleCommand(
             f"type summary add -F lldb_formatters.SafePrintSummary {i} -w my"

src/cminusfc/cminusf_builder.cpp

@@ -142,8 +142,8 @@ Value* CminusfBuilder::visit(ASTFunDeclaration &node) {
     scope.enter();
     context.pre_enter_scope = true;
     std::vector<Value *> args;
-    for (auto &arg : func->get_args()) {
-        args.push_back(&arg);
+    for (auto arg : func->get_args()) {
+        args.push_back(arg);
     }
     for (unsigned int i = 0; i < node.params.size(); ++i) {
         if (node.params[i]->isarray) {

src/codegen/CodeGen.cpp

 #include "CodeGen.hpp"
 
+#include "ASMInstruction.hpp"
 #include "CodeGenUtil.hpp"
+#include "Register.hpp"
+
+std::string ASMInstruction::safe_print() const
+{
+    switch (type) {
+    case Instruction:
+    case Attribute:
+        return content;
+    case Label:
+        return content + ":";
+    case Comment:
+        return "# " + content;
+    }
+    return "<error>";
+}
 
 void CodeGen::allocate() {
     // 备份 $ra $fp
@@ -8,9 +24,9 @@ void CodeGen::allocate() {
 
     // 为每个参数分配栈空间
     for (auto &arg : context.func->get_args()) {
-        auto size = arg.get_type()->get_size();
+        auto size = arg->get_type()->get_size();
         offset = offset + size;
-        context.offset_map[&arg] = -static_cast<int>(offset);
+        context.offset_map[arg] = -static_cast<int>(offset);
     }
 
     // 为指令结果分配栈空间
@@ -205,11 +221,11 @@ void CodeGen::gen_prologue() {
 
     int garg_cnt = 0;
     int farg_cnt = 0;
-    for (auto &arg : context.func->get_args()) {
-        if (arg.get_type()->is_float_type()) {
-            store_from_freg(&arg, FReg::fa(farg_cnt++));
+    for (auto arg : context.func->get_args()) {
+        if (arg->get_type()->is_float_type()) {
+            store_from_freg(arg, FReg::fa(farg_cnt++));
         } else { // int or pointer
-            store_from_greg(&arg, Reg::a(garg_cnt++));
+            store_from_greg(arg, Reg::a(garg_cnt++));
         }
     }
 }
@@ -241,16 +257,16 @@ void CodeGen::gen_binary() {
     load_to_greg(context.inst->get_operand(1), Reg::t(1));
     switch (context.inst->get_instr_type()) {
     case Instruction::add:
-        output.emplace_back("add.w $t2, $t0, $t1");
+        append_inst("add.w $t2, $t0, $t1");
         break;
     case Instruction::sub:
-        output.emplace_back("sub.w $t2, $t0, $t1");
+        append_inst("sub.w $t2, $t0, $t1");
         break;
     case Instruction::mul:
-        output.emplace_back("mul.w $t2, $t0, $t1");
+        append_inst("mul.w $t2, $t0, $t1");
         break;
     case Instruction::sdiv:
-        output.emplace_back("div.w $t2, $t0, $t1");
+        append_inst("div.w $t2, $t0, $t1");
         break;
     default:
         assert(false);
@@ -370,26 +386,26 @@ void CodeGen::run() {
          * - 支持更旧版本的 GNU 汇编器, 因为 `.bss` 伪指令是应该相对较新的指令,
          *   GNU 汇编器在 2023 年 2 月的 2.37 版本才将其引入
          */
-        append_inst(".text", ASMInstruction::Atrribute);
+        append_inst(".text", ASMInstruction::Attribute);
         append_inst(".section", {".bss", "\"aw\"", "@nobits"},
-                    ASMInstruction::Atrribute);
+                    ASMInstruction::Attribute);
         for (auto global : m->get_global_variable()) {
             auto size =
                 global->get_type()->get_pointer_element_type()->get_size();
             append_inst(".globl", {global->get_name()},
-                        ASMInstruction::Atrribute);
+                        ASMInstruction::Attribute);
             append_inst(".type", {global->get_name(), "@object"},
-                        ASMInstruction::Atrribute);
+                        ASMInstruction::Attribute);
             append_inst(".size", {global->get_name(), std::to_string(size)},
-                        ASMInstruction::Atrribute);
+                        ASMInstruction::Attribute);
             append_inst(global->get_name(), ASMInstruction::Label);
             append_inst(".space", {std::to_string(size)},
-                        ASMInstruction::Atrribute);
+                        ASMInstruction::Attribute);
         }
     }
 
     // 函数代码段
-    output.emplace_back(".text", ASMInstruction::Atrribute);
+    output.emplace_back(new ASMInstruction(".text", ASMInstruction::Attribute));
     for (auto func : m->get_functions()) {
         if (not func->is_declaration()) {
             // 更新 context
@@ -397,9 +413,9 @@ void CodeGen::run() {
             context.func = func;
 
             // 函数信息
-            append_inst(".globl", {func->get_name()}, ASMInstruction::Atrribute);
+            append_inst(".globl", {func->get_name()}, ASMInstruction::Attribute);
             append_inst(".type", {func->get_name(), "@function"},
-                        ASMInstruction::Atrribute);
+                        ASMInstruction::Attribute);
             append_inst(func->get_name(), ASMInstruction::Label);
 
             // 分配函数栈帧
@@ -495,8 +511,8 @@ void CodeGen::run() {
 
 std::string CodeGen::print() const {
     std::string result;
-    for (const auto &inst : output) {
-        result += inst.format();
+    for (const auto inst : output) {
+        result += inst->format();
     }
     return result;
 }

src/codegen/Register.cpp

@@ -26,9 +26,35 @@ std::string Reg::print() const {
     assert(false);
 }
 
+std::string Reg::safe_print() const
+{
+    if (id == 0) {
+        return "$zero";
+    }
+    if (id == 1) {
+        return "$ra";
+    }
+    if (id == 2) {
+        return "$tp";
+    }
+    if (id == 3) {
+        return "$sp";
+    }
+    if (4 <= id and id <= 11) {
+        return "$a" + std::to_string(id - 4);
+    }
+    if (12 <= id and id <= 20) {
+        return "$t" + std::to_string(id - 12);
+    }
+    if (id == 22) {
+        return "$fp";
+    }
+    return "<error id " + std::to_string(id) + ">";
+}
+
 std::string FReg::print() const {
 
-    if (0 <= id and id <= 7) {
+    if (id <= 7) {
         return "$fa" + std::to_string(id);
     }
     if (8 <= id and id <= 23) {
@@ -39,3 +65,22 @@ std::string FReg::print() const {
     }
     assert(false);
 }
+
+std::string FReg::safe_print() const
+{
+    if (id <= 7) {
+        return "$fa" + std::to_string(id);
+    }
+    if (8 <= id and id <= 23) {
+        return "$ft" + std::to_string(id - 8);
+    }
+    if (24 <= id and id <= 31) {
+        return "$fs" + std::to_string(id - 24);
+    }
+    return "<error id " + std::to_string(id) + ">";
+}
+
+std::string CFReg::safe_print() const
+{
+    return "$fcc" + std::to_string(id);
+}

src/lightir/BasicBlock.cpp

@@ -125,4 +125,4 @@ BasicBlock* BasicBlock::get_entry_block_of_same_function() const
     return parent_->get_entry_block();
 }
 
-Names GLOBAL_BASICBLOCK_NAMES_{false, "label", ""};
\ No newline at end of file
+Names GLOBAL_BASICBLOCK_NAMES_{"label", "_"};
\ No newline at end of file

src/lightir/Function.cpp

@@ -3,6 +3,7 @@
 #include "Module.hpp"
 #include "Instruction.hpp"
 #include <cassert>
+#include <string>
 #include <unordered_set>
 #include <queue>
 
@@ -10,24 +11,25 @@ namespace
 {
     std::string chopName(std::string name)
     {
-        if (name.size() > 4) return { name.begin(), name.begin() + 4 };
+        if (name.size() > 3) return { name.begin(), name.begin() + 3 };
         return name;
     }
 }
 
 Function::Function(FunctionType* ty, const std::string& name, Module* parent)
-    : Value(ty, name), names4blocks_(false, "label", chopName(name)), names4insts_(true, "op", ""), parent_(parent), seq_cnt_(0) {
+    : Value(ty, name), names4blocks_("label", chopName(name) + "_"), names4insts_("op", ""), parent_(parent), seq_cnt_(0) {
     // num_args_ = ty->getNumParams();
     parent->add_function(this);
     // build args
     for (unsigned i = 0; i < get_num_of_args(); i++) {
-        arguments_.emplace_back(ty->get_param_type(i), "", this, i);
+        arguments_.emplace_back(new Argument(ty->get_param_type(i), "arg" + std::to_string(i), this, i));
     }
 }
 
 Function::~Function()
 {
     for (auto bb : basic_blocks_) delete bb;
+    for(auto arg: arguments_) delete arg;
 }
 
 Function* Function::create(FunctionType* ty, const std::string& name,
@@ -66,10 +68,10 @@ void Function::add_basic_block(BasicBlock* bb) { basic_blocks_.push_back(bb); }
 void Function::set_instr_name() {
     std::map<Value*, int> seq;
     for (auto& arg : this->get_args()) {
-        if (seq.find(&arg) == seq.end()) {
+        if (seq.find(arg) == seq.end()) {
             auto seq_num = seq.size() + seq_cnt_;
-            if (arg.set_name("arg" + std::to_string(seq_num))) {
-                seq.insert({ &arg, seq_num });
+            if (arg->set_name("arg" + std::to_string(seq_num))) {
+                seq.insert({ arg, seq_num });
             }
         }
     }
@@ -121,7 +123,7 @@ std::string Function::print() {
         for (auto& arg : get_args()) {
             if (&arg != &*get_args().begin())
                 func_ir += ", ";
-            func_ir += arg.print();
+            func_ir += arg->print();
         }
     }
     func_ir += ")";
@@ -160,7 +162,7 @@ std::string Argument::safe_print() const
     auto ty = parent->get_function_type();
     if (ty == nullptr || ty->get_num_of_args() <= arg_no_) return "@" + parent->get_name() + " arg" + std::to_string(arg_no_) + " <unknow type>";
     auto ty2 = ty->get_param_type(arg_no_);
-    return "@" + parent->get_name() + " arg" + std::to_string(arg_no_) + " " + (ty2 == nullptr ? "<null>" : ty2->safe_print());
+    return (ty2 == nullptr ? "<null>" : ty2->safe_print()) + " @" + parent->get_name() + " arg" + std::to_string(arg_no_);
 }
 
 

src/lightir/IRprinter.cpp

@@ -538,10 +538,12 @@ std::string BranchInst::print() {
 
 std::string ReturnInst::print() {
     std::string instr_ir;
-    instr_ir += safe_print_instr_op_name(get_instr_type());
+    instr_ir += get_instr_op_name();
     instr_ir += " ";
     if (!is_void_ret()) {
-        instr_ir += safe_print_op_as_op(this, 0, true);
+        instr_ir += this->get_operand(0)->get_type()->print();
+        instr_ir += " ";
+        instr_ir += print_as_op(this->get_operand(0), false);
     } else {
         instr_ir += "void";
     }

src/lightir/Instruction.cpp

@@ -368,4 +368,4 @@ PhiInst *PhiInst::create_phi(Type *ty, BasicBlock *bb,
     return new PhiInst(ty, vals, val_bbs, bb, name);
 }
 
-Names GLOBAL_INSTRUCTION_NAMES_{ true, "op", "" };
\ No newline at end of file
+Names GLOBAL_INSTRUCTION_NAMES_{"op", "_" };
\ No newline at end of file

src/lightir/Names.cpp

@@ -3,19 +3,25 @@
 std::string Names::get_name(std::string name)
 {
     int ed = static_cast<int>(name.size());
-	while (ed > 0)
+    int bg = 0;
+    while (bg < ed)
+    {
+        char ch = name[bg];
+        if (ch == '_' || (ch >= '0' && ch <= '9')) bg++;
+        else break;
+    }
+    while (ed > bg)
     {
         char ch = name[ed - 1];
-		if ((use_underline_ && ch == '_') || (ch >= '0' && ch <= '9')) ed--;
+        if (ch == '_' || (ch >= '0' && ch <= '9')) ed--;
         else break;
     }
-	if (ed == 0) return get_name();
-	std::string name1 = {name.begin(), name.begin() + ed};
+    if (bg == ed) return get_name();
+    std::string name1 = {name.begin() + bg, name.begin() + ed};
     if (name1 == default_prefix_) return get_name();
     auto get = appended_prefix_ + name1;
     auto idx = allocated_[name1]++;
     if (idx == 0) return get;
-	if (use_underline_) get += "_";
     return get + std::to_string(idx);
 }