aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
author_ov4 <hausemasterstream@gmail.com>2026-01-30 13:28:45 +0000
committer_ov4 <hausemasterstream@gmail.com>2026-01-30 13:28:45 +0000
commit93175ba5edb0eacc182f573c4fd4556f089da305 (patch)
tree0e50d5e0b50d7298dcb50342583fe4b718ccd1c0
parentbd799e42b40613983c6e3cb9aedf58866ac68605 (diff)
Added BYREF and BYVAL
-rw-r--r--CMakeLists.txt2
-rw-r--r--include/cps/ArrayHandler.h6
-rw-r--r--include/cps/CodeGen.h2
-rw-r--r--include/cps/FunctionAST.h7
-rw-r--r--include/cps/FunctionGen.h7
-rw-r--r--include/cps/Lexer.h4
-rw-r--r--include/cps/Parser.h2
-rw-r--r--lib/CodeGen/ArrayHandler.cc8
-rw-r--r--lib/CodeGen/CodeGen.cc66
-rw-r--r--lib/CodeGen/FunctionGen.cc36
-rw-r--r--lib/Lexer/Lexer.cc3
-rw-r--r--lib/Parser/ParserFunction.cc (renamed from lib/Parser/Parser_Function.cc)16
12 files changed, 118 insertions, 41 deletions
diff --git a/CMakeLists.txt b/CMakeLists.txt
index f1eefd9..f43204b 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -15,7 +15,7 @@ include_directories(include)
add_library(CPSLexer lib/Lexer/Lexer.cc)
add_library(CPSParser
lib/Parser/Parser.cc
- lib/Parser/Parser_Function.cc
+ lib/Parser/ParserFunction.cc
)
add_library(CPSCodeGen
diff --git a/include/cps/ArrayHandler.h b/include/cps/ArrayHandler.h
index 734f464..1ee91db 100644
--- a/include/cps/ArrayHandler.h
+++ b/include/cps/ArrayHandler.h
@@ -24,7 +24,7 @@ class ArrayHandler {
llvm::LLVMContext *TheContext;
llvm::IRBuilder<> *Builder;
llvm::Module *TheModule;
- std::map<std::string, llvm::AllocaInst*> *NamedValues;
+ std::map<std::string, llvm::Value*> *NamedValues;
std::map<std::string, ArrayMetadata> ArrayTable;
RuntimeCheck &RuntimeChecker;
@@ -37,13 +37,13 @@ class ArrayHandler {
void emitPrintLoop(const std::string &Name, int CurrentDim, std::vector<llvm::Value*> CurrentIndices, llvm::FunctionCallee PrintfFunc, llvm::Value* FmtStr);
public:
- ArrayHandler(llvm::LLVMContext &C, llvm::IRBuilder<> &B, llvm::Module &M, std::map<std::string, llvm::AllocaInst*> &NV);
+ ArrayHandler(llvm::LLVMContext &C, llvm::IRBuilder<> &B, llvm::Module &M, std::map<std::string, llvm::Value*> &NV);
void setupExternalFunctions();
void emitArrayDeclare(ArrayDeclareStmtAST *Stmt, CodeGen &CG);
void emitArrayAssign(ArrayAssignStmtAST *Stmt, CodeGen &CG);
- ArrayHandler(llvm::LLVMContext &C, llvm::IRBuilder<> &B, llvm::Module &M, std::map<std::string, llvm::AllocaInst*> &NV, RuntimeCheck &RC);
+ ArrayHandler(llvm::LLVMContext &C, llvm::IRBuilder<> &B, llvm::Module &M, std::map<std::string, llvm::Value*> &NV, RuntimeCheck &RC);
llvm::Value* emitArrayAccess(ArrayAccessExprAST *Expr, CodeGen &CG);
diff --git a/include/cps/CodeGen.h b/include/cps/CodeGen.h
index 832ba7a..85f605f 100644
--- a/include/cps/CodeGen.h
+++ b/include/cps/CodeGen.h
@@ -16,7 +16,7 @@ class CodeGen {
std::unique_ptr<llvm::LLVMContext> TheContext;
std::unique_ptr<llvm::Module> TheModule;
std::unique_ptr<llvm::IRBuilder<>> Builder;
- std::map<std::string, llvm::AllocaInst*> NamedValues;
+ std::map<std::string, llvm::Value*> NamedValues;
std::unique_ptr<ArrayHandler> Arrays;
std::unique_ptr<RuntimeCheck> RuntimeChecker;
diff --git a/include/cps/FunctionAST.h b/include/cps/FunctionAST.h
index 155bd8c..5b4da6c 100644
--- a/include/cps/FunctionAST.h
+++ b/include/cps/FunctionAST.h
@@ -3,24 +3,25 @@
#include <vector>
#include <string>
#include <memory>
+#include <tuple>
namespace cps {
class PrototypeAST {
std::string Name;
- std::vector<std::pair<std::string, std::string>> Args;
+ std::vector<std::tuple<std::string, std::string, bool>> Args;
std::string ReturnType;
bool IsExternal;
public:
PrototypeAST(const std::string &Name,
- std::vector<std::pair<std::string, std::string>> Args,
+ std::vector<std::tuple<std::string, std::string, bool>> Args,
const std::string &ReturnType,
bool IsExternal = false)
: Name(Name), Args(std::move(Args)), ReturnType(ReturnType), IsExternal(IsExternal) {}
const std::string &getName() const { return Name; }
- const std::vector<std::pair<std::string, std::string>> &getArgs() const { return Args; }
+ const std::vector<std::tuple<std::string, std::string, bool>> &getArgs() const { return Args; }
const std::string &getReturnType() const { return ReturnType; }
bool isExternal() const { return IsExternal; }
};
diff --git a/include/cps/FunctionGen.h b/include/cps/FunctionGen.h
index c1a6ebd..84a2858 100644
--- a/include/cps/FunctionGen.h
+++ b/include/cps/FunctionGen.h
@@ -6,6 +6,7 @@
#include <vector>
#include <functional>
#include <string>
+#include <tuple>
namespace cps {
@@ -14,13 +15,13 @@ class FunctionGen {
llvm::Module &Module;
llvm::IRBuilder<> &Builder;
- std::map<std::string, llvm::AllocaInst*> &NamedValues;
+ std::map<std::string, llvm::Value*> &NamedValues;
- void createArgumentAllocas(llvm::Function *F, const std::vector<std::pair<std::string, std::string>> &Args);
+ void createArgumentAllocas(llvm::Function *F, const std::vector<std::tuple<std::string, std::string, bool>> &Args);
public:
FunctionGen(llvm::LLVMContext &C, llvm::Module &M, llvm::IRBuilder<> &B,
- std::map<std::string, llvm::AllocaInst*> &NV)
+ std::map<std::string, llvm::Value*> &NV)
: Context(C), Module(M), Builder(B), NamedValues(NV) {}
llvm::Type *getLLVMType(const std::string &TypeName);
diff --git a/include/cps/Lexer.h b/include/cps/Lexer.h
index cd34435..4226cc3 100644
--- a/include/cps/Lexer.h
+++ b/include/cps/Lexer.h
@@ -49,7 +49,9 @@ enum Token {
tok_returns = -403,
tok_procedure = -404,
tok_endprocedure = -405,
- tok_call = -406
+ tok_call = -406,
+ tok_byref = -407,
+ tok_byval = -408
};
class Lexer {
diff --git a/include/cps/Parser.h b/include/cps/Parser.h
index f328141..2a9c98a 100644
--- a/include/cps/Parser.h
+++ b/include/cps/Parser.h
@@ -36,7 +36,7 @@ class Parser {
std::unique_ptr<StmtAST> ParseProcedure();
std::unique_ptr<StmtAST> ParseCallStmt();
std::unique_ptr<StmtAST> ParseReturnStmt();
- std::vector<std::pair<std::string, std::string>> ParsePrototypeArgs();
+ std::vector<std::tuple<std::string, std::string, bool>> ParsePrototypeArgs();
public:
Parser(Lexer &L);
diff --git a/lib/CodeGen/ArrayHandler.cc b/lib/CodeGen/ArrayHandler.cc
index 18b9253..5842f77 100644
--- a/lib/CodeGen/ArrayHandler.cc
+++ b/lib/CodeGen/ArrayHandler.cc
@@ -8,7 +8,7 @@
using namespace llvm;
using namespace cps;
-ArrayHandler::ArrayHandler(LLVMContext &C, IRBuilder<> &B, Module &M, std::map<std::string, AllocaInst*> &NV, RuntimeCheck &RC)
+ArrayHandler::ArrayHandler(LLVMContext &C, IRBuilder<> &B, Module &M, std::map<std::string, Value*> &NV, RuntimeCheck &RC)
: TheContext(&C), Builder(&B), TheModule(&M), NamedValues(&NV), RuntimeChecker(RC) {
setupExternalFunctions();
}
@@ -125,7 +125,7 @@ Value* ArrayHandler::emitArrayAccess(ArrayAccessExprAST *Expr, CodeGen &CG) {
Value *Offset = computeFlatIndex(Name, Indices);
- AllocaInst *Alloca = (*NamedValues)[Name];
+ Value *Alloca = (*NamedValues)[Name];
Value *RawPtr = Builder->CreateLoad(PointerType::getUnqual(*TheContext), Alloca);
Value *IntPtr = Builder->CreateBitCast(RawPtr, PointerType::getUnqual(*TheContext));
Value *ElemPtr = Builder->CreateGEP(Type::getInt64Ty(*TheContext), IntPtr, Offset);
@@ -154,7 +154,7 @@ void ArrayHandler::emitArrayAssign(ArrayAssignStmtAST *Stmt, CodeGen &CG) {
Value *Val = CG.emitExpr(Stmt->getExpr());
Value *Offset = computeFlatIndex(Name, Indices);
- AllocaInst *Alloca = (*NamedValues)[Name];
+ Value *Alloca = (*NamedValues)[Name];
Value *RawPtr = Builder->CreateLoad(PointerType::getUnqual(*TheContext), Alloca);
Value *IntPtr = Builder->CreateBitCast(RawPtr, PointerType::getUnqual(*TheContext));
Value *ElemPtr = Builder->CreateGEP(Type::getInt64Ty(*TheContext), IntPtr, Offset);
@@ -197,7 +197,7 @@ void ArrayHandler::emitPrintLoop(const std::string &Name, int CurrentDim, std::v
if (CurrentDim == Meta.Rank) {
Value *Offset = computeFlatIndex(Name, CurrentIndices);
- AllocaInst *Alloca = (*NamedValues)[Name];
+ Value *Alloca = (*NamedValues)[Name];
Value *RawPtr = Builder->CreateLoad(PointerType::getUnqual(*TheContext), Alloca);
Value *IntPtr = Builder->CreateBitCast(RawPtr, PointerType::getUnqual(*TheContext));
Value *ElemPtr = Builder->CreateGEP(Type::getInt64Ty(*TheContext), IntPtr, Offset);
diff --git a/lib/CodeGen/CodeGen.cc b/lib/CodeGen/CodeGen.cc
index 6358699..bfbd8e0 100644
--- a/lib/CodeGen/CodeGen.cc
+++ b/lib/CodeGen/CodeGen.cc
@@ -70,7 +70,7 @@ Value *CodeGen::emitExpr(ExprAST *Expr) {
}
if (auto *Var = dynamic_cast<VariableExprAST*>(Expr)) {
- AllocaInst *A = NamedValues[Var->getName()];
+ Value *A = NamedValues[Var->getName()];
if (!A) {
fprintf(stderr, "Error: Unknown variable name %s\n", Var->getName().c_str());
return nullptr;
@@ -109,9 +109,34 @@ Value *CodeGen::emitExpr(ExprAST *Expr) {
}
if (auto *Call = dynamic_cast<CallExprAST*>(Expr)) {
+ Function *CalleeF = TheModule->getFunction(Call->getCallee());
std::vector<Value*> Args;
- for (const auto &Arg : Call->getArgs()) {
- Args.push_back(emitExpr(Arg.get()));
+
+ for (unsigned i = 0; i < Call->getArgs().size(); ++i) {
+ ExprAST *ArgExpr = Call->getArgs()[i].get();
+ bool isByRef = false;
+
+ if (CalleeF && i < CalleeF->arg_size()) {
+ if (CalleeF->getArg(i)->getType()->isPointerTy()) {
+ isByRef = true;
+ }
+ }
+
+ if (isByRef) {
+ if (auto *Var = dynamic_cast<VariableExprAST*>(ArgExpr)) {
+ Value *Ptr = NamedValues[Var->getName()];
+ if (!Ptr) {
+ fprintf(stderr, "Error: Unknown variable %s in BYREF call\n", Var->getName().c_str());
+ return nullptr;
+ }
+ Args.push_back(Ptr);
+ } else {
+ fprintf(stderr, "Error: BYREF argument must be a variable.\n");
+ return nullptr;
+ }
+ } else {
+ Args.push_back(emitExpr(ArgExpr));
+ }
}
return FuncGen->emitCallExpr(Call, Args);
}
@@ -203,7 +228,7 @@ void CodeGen::emitForStmt(ForStmtAST *Stmt) {
Value *StartVal = emitExpr(Stmt->getStart());
if (!StartVal) return;
- AllocaInst *Alloca = NamedValues[VarName];
+ Value *Alloca = NamedValues[VarName];
if (!Alloca) {
fprintf(stderr, "Error: Unknown variable in FOR loop %s\n", VarName.c_str());
return;
@@ -285,9 +310,34 @@ void CodeGen::emitStmt(StmtAST *Stmt) {
}
else if (auto *Call = dynamic_cast<CallStmtAST*>(Stmt)) {
+ Function *CalleeF = TheModule->getFunction(Call->getCallee());
std::vector<Value*> Args;
- for (const auto &ArgExpr : Call->getArgs()) {
- Args.push_back(emitExpr(ArgExpr.get()));
+
+ for (unsigned i = 0; i < Call->getArgs().size(); ++i) {
+ ExprAST *ArgExpr = Call->getArgs()[i].get();
+ bool isByRef = false;
+
+ if (CalleeF && i < CalleeF->arg_size()) {
+ if (CalleeF->getArg(i)->getType()->isPointerTy()) {
+ isByRef = true;
+ }
+ }
+
+ if (isByRef) {
+ if (auto *Var = dynamic_cast<VariableExprAST*>(ArgExpr)) {
+ Value *Ptr = NamedValues[Var->getName()];
+ if (!Ptr) {
+ fprintf(stderr, "Error: Unknown variable %s in BYREF call\n", Var->getName().c_str());
+ return;
+ }
+ Args.push_back(Ptr);
+ } else {
+ fprintf(stderr, "Error: BYREF argument must be a variable.\n");
+ return;
+ }
+ } else {
+ Args.push_back(emitExpr(ArgExpr));
+ }
}
FuncGen->emitCallStmt(Call, Args);
return;
@@ -311,7 +361,7 @@ void CodeGen::emitStmt(StmtAST *Stmt) {
else if (auto *Assign = dynamic_cast<AssignStmtAST*>(Stmt)) {
Value *Val = emitExpr(Assign->getExpr());
if (Val) {
- AllocaInst *Alloca = NamedValues[Assign->getName()];
+ Value *Alloca = NamedValues[Assign->getName()];
if (!Alloca) {
fprintf(stderr, "Error: Unknown variable name %s\n", Assign->getName().c_str());
return;
@@ -320,7 +370,7 @@ void CodeGen::emitStmt(StmtAST *Stmt) {
}
}
else if (auto *In = dynamic_cast<InputStmtAST*>(Stmt)) {
- AllocaInst *Alloca = NamedValues[In->getName()];
+ Value *Alloca = NamedValues[In->getName()];
if (Alloca) {
std::vector<Value*> Args;
Args.push_back(ScanfFormatStr);
diff --git a/lib/CodeGen/FunctionGen.cc b/lib/CodeGen/FunctionGen.cc
index b2adc08..a929943 100644
--- a/lib/CodeGen/FunctionGen.cc
+++ b/lib/CodeGen/FunctionGen.cc
@@ -13,27 +13,37 @@ Type *FunctionGen::getLLVMType(const std::string &TypeName) {
return Type::getInt64Ty(Context);
}
-void FunctionGen::createArgumentAllocas(Function *F, const std::vector<std::pair<std::string, std::string>> &Args) {
+void FunctionGen::createArgumentAllocas(Function *F, const std::vector<std::tuple<std::string, std::string, bool>> &Args) {
Function::arg_iterator AI = F->arg_begin();
for (unsigned Idx = 0, E = Args.size(); Idx != E; ++Idx, ++AI) {
- std::string ArgName = Args[Idx].first;
- std::string ArgTypeStr = Args[Idx].second;
-
- Type *ArgType = getLLVMType(ArgTypeStr);
- IRBuilder<> TmpB(&F->getEntryBlock(), F->getEntryBlock().begin());
- AllocaInst *Alloca = TmpB.CreateAlloca(ArgType, nullptr, ArgName);
+ std::string ArgName = std::get<0>(Args[Idx]);
+ std::string ArgTypeStr = std::get<1>(Args[Idx]);
+ bool IsRef = std::get<2>(Args[Idx]);
Value *ArgVal = &(*AI);
- Builder.CreateStore(ArgVal, Alloca);
-
- NamedValues[ArgName] = Alloca;
+ ArgVal->setName(ArgName);
+
+ if (IsRef) {
+ NamedValues[ArgName] = ArgVal;
+ } else {
+ Type *ArgType = getLLVMType(ArgTypeStr);
+ IRBuilder<> TmpB(&F->getEntryBlock(), F->getEntryBlock().begin());
+ AllocaInst *Alloca = TmpB.CreateAlloca(ArgType, nullptr, ArgName);
+
+ Builder.CreateStore(ArgVal, Alloca);
+ NamedValues[ArgName] = Alloca;
+ }
}
}
Function *FunctionGen::emitPrototype(PrototypeAST *Proto) {
std::vector<Type*> ArgTypes;
for (const auto &Arg : Proto->getArgs()) {
- ArgTypes.push_back(getLLVMType(Arg.second));
+ Type *T = getLLVMType(std::get<1>(Arg));
+ if (std::get<2>(Arg)) {
+ T = T->getPointerTo();
+ }
+ ArgTypes.push_back(T);
}
Type *RetType = getLLVMType(Proto->getReturnType());
@@ -43,7 +53,7 @@ Function *FunctionGen::emitPrototype(PrototypeAST *Proto) {
unsigned Idx = 0;
for (auto &Arg : F->args()) {
if (Idx < Proto->getArgs().size())
- Arg.setName(Proto->getArgs()[Idx++].first);
+ Arg.setName(std::get<0>(Proto->getArgs()[Idx++]));
}
return F;
}
@@ -65,7 +75,7 @@ Function *FunctionGen::emitFunctionDef(FunctionDefAST *FuncAST,
BasicBlock *BB = BasicBlock::Create(Context, "entry", TheFunction);
Builder.SetInsertPoint(BB);
- std::map<std::string, llvm::AllocaInst*> OldNamedValues = NamedValues;
+ std::map<std::string, llvm::Value*> OldNamedValues = NamedValues;
NamedValues.clear();
createArgumentAllocas(TheFunction, Proto->getArgs());
diff --git a/lib/Lexer/Lexer.cc b/lib/Lexer/Lexer.cc
index 6854cc9..da2e47f 100644
--- a/lib/Lexer/Lexer.cc
+++ b/lib/Lexer/Lexer.cc
@@ -54,6 +54,9 @@ int Lexer::gettok() {
if (IdentifierStr == "RETURN") return tok_return;
if (IdentifierStr == "RETURNS") return tok_returns;
if (IdentifierStr == "CALL") return tok_call;
+
+ if (IdentifierStr == "BYREF") return tok_byref;
+ if (IdentifierStr == "BYVAL") return tok_byval;
return tok_identifier;
}
diff --git a/lib/Parser/Parser_Function.cc b/lib/Parser/ParserFunction.cc
index bdb4ee7..d689aa6 100644
--- a/lib/Parser/Parser_Function.cc
+++ b/lib/Parser/ParserFunction.cc
@@ -1,16 +1,26 @@
#include "cps/Parser.h"
#include "cps/FunctionAST.h"
+#include <tuple>
using namespace cps;
-std::vector<std::pair<std::string, std::string>> Parser::ParsePrototypeArgs() {
- std::vector<std::pair<std::string, std::string>> Args;
+std::vector<std::tuple<std::string, std::string, bool>> Parser::ParsePrototypeArgs() {
+ std::vector<std::tuple<std::string, std::string, bool>> Args;
if (CurTok != '(') return Args;
getNextToken();
while (CurTok != ')') {
if (CurTok == -1000 || CurTok == -1001) getNextToken();
+ bool IsRef = false;
+ if (CurTok == tok_byref) {
+ IsRef = true;
+ getNextToken();
+ } else if (CurTok == tok_byval) {
+ IsRef = false;
+ getNextToken();
+ }
+
if (CurTok != tok_identifier) {
fprintf(stderr, "Error: Expected argument name\n");
return Args;
@@ -35,7 +45,7 @@ std::vector<std::pair<std::string, std::string>> Parser::ParsePrototypeArgs() {
return Args;
}
- Args.push_back({Name, Type});
+ Args.emplace_back(Name, Type, IsRef);
if (CurTok == ')') break;
if (CurTok != ',') {