examples/Kaleidoscope/toy.cpp

   1 #include "llvm/DerivedTypes.h"
   2 #include "llvm/ExecutionEngine/ExecutionEngine.h"
   3 #include "llvm/Module.h"
   4 #include "llvm/ModuleProvider.h"
   5 #include "llvm/PassManager.h"
   6 #include "llvm/Analysis/Verifier.h"
   7 #include "llvm/Target/TargetData.h"
   8 #include "llvm/Transforms/Scalar.h"
   9 #include "llvm/Support/IRBuilder.h"
  10 #include <cstdio>
  11 #include <string>
  12 #include <map>
  13 #include <vector>
  14 using namespace llvm;
  15
  16 //===----------------------------------------------------------------------===//
  17 // Lexer
  18 //===----------------------------------------------------------------------===//
  19
  20 // The lexer returns tokens [0-255] if it is an unknown character, otherwise one
  21 // of these for known things.
  22 enum Token {
  23   tok_eof = -1,
  24
  25   // commands
  26   tok_def = -2, tok_extern = -3,
  27
  28   // primary
  29   tok_identifier = -4, tok_number = -5,
  30
  31   // control
  32   tok_if = -6, tok_then = -7, tok_else = -8,
  33   tok_for = -9, tok_in = -10,
  34
  35   // operators
  36   tok_binary = -11, tok_unary = -12,
  37
  38   // var definition
  39   tok_var = -13
  40 };
  41
  42 static std::string IdentifierStr;  // Filled in if tok_identifier
  43 static double NumVal;              // Filled in if tok_number
  44
  45 /// gettok - Return the next token from standard input.
  46 static int gettok() {
  47   static int LastChar = ' ';
  48
  49   // Skip any whitespace.
  50   while (isspace(LastChar))
  51     LastChar = getchar();
  52
  53   if (isalpha(LastChar)) { // identifier: [a-zA-Z][a-zA-Z0-9]*
  54     IdentifierStr = LastChar;
  55     while (isalnum((LastChar = getchar())))
  56       IdentifierStr += LastChar;
  57
  58     if (IdentifierStr == "def") return tok_def;
  59     if (IdentifierStr == "extern") return tok_extern;
  60     if (IdentifierStr == "if") return tok_if;
  61     if (IdentifierStr == "then") return tok_then;
  62     if (IdentifierStr == "else") return tok_else;
  63     if (IdentifierStr == "for") return tok_for;
  64     if (IdentifierStr == "in") return tok_in;
  65     if (IdentifierStr == "binary") return tok_binary;
  66     if (IdentifierStr == "unary") return tok_unary;
  67     if (IdentifierStr == "var") return tok_var;
  68     return tok_identifier;
  69   }
  70
  71   if (isdigit(LastChar) || LastChar == '.') {   // Number: [0-9.]+
  72     std::string NumStr;
  73     do {
  74       NumStr += LastChar;
  75       LastChar = getchar();
  76     } while (isdigit(LastChar) || LastChar == '.');
  77
  78     NumVal = strtod(NumStr.c_str(), 0);
  79     return tok_number;
  80   }
  81
  82   if (LastChar == '#') {
  83     // Comment until end of line.
  84     do LastChar = getchar();
  85     while (LastChar != EOF && LastChar != '\n' && LastChar != '\r');
  86
  87     if (LastChar != EOF)
  88       return gettok();
  89   }
  90
  91   // Check for end of file.  Don't eat the EOF.
  92   if (LastChar == EOF)
  93     return tok_eof;
  94
  95   // Otherwise, just return the character as its ascii value.
  96   int ThisChar = LastChar;
  97   LastChar = getchar();
  98   return ThisChar;
  99 }
 100
 101 //===----------------------------------------------------------------------===//
 102 // Abstract Syntax Tree (aka Parse Tree)
 103 //===----------------------------------------------------------------------===//
 104
 105 /// ExprAST - Base class for all expression nodes.
 106 class ExprAST {
 107 public:
 108   virtual ~ExprAST() {}
 109   virtual Value *Codegen() = 0;
 110 };
 111
 112 /// NumberExprAST - Expression class for numeric literals like "1.0".
 113 class NumberExprAST : public ExprAST {
 114   double Val;
 115 public:
 116   NumberExprAST(double val) : Val(val) {}
 117   virtual Value *Codegen();
 118 };
 119
 120 /// VariableExprAST - Expression class for referencing a variable, like "a".
 121 class VariableExprAST : public ExprAST {
 122   std::string Name;
 123 public:
 124   VariableExprAST(const std::string &name) : Name(name) {}
 125   const std::string &getName() const { return Name; }
 126   virtual Value *Codegen();
 127 };
 128
 129 /// UnaryExprAST - Expression class for a unary operator.
 130 class UnaryExprAST : public ExprAST {
 131   char Opcode;
 132   ExprAST *Operand;
 133 public:
 134   UnaryExprAST(char opcode, ExprAST *operand)
 135     : Opcode(opcode), Operand(operand) {}
 136   virtual Value *Codegen();
 137 };
 138
 139 /// BinaryExprAST - Expression class for a binary operator.
 140 class BinaryExprAST : public ExprAST {
 141   char Op;
 142   ExprAST *LHS, *RHS;
 143 public:
 144   BinaryExprAST(char op, ExprAST *lhs, ExprAST *rhs)
 145     : Op(op), LHS(lhs), RHS(rhs) {}
 146   virtual Value *Codegen();
 147 };
 148
 149 /// CallExprAST - Expression class for function calls.
 150 class CallExprAST : public ExprAST {
 151   std::string Callee;
 152   std::vector<ExprAST*> Args;
 153 public:
 154   CallExprAST(const std::string &callee, std::vector<ExprAST*> &args)
 155     : Callee(callee), Args(args) {}
 156   virtual Value *Codegen();
 157 };
 158
 159 /// IfExprAST - Expression class for if/then/else.
 160 class IfExprAST : public ExprAST {
 161   ExprAST *Cond, *Then, *Else;
 162 public:
 163   IfExprAST(ExprAST *cond, ExprAST *then, ExprAST *_else)
 164   : Cond(cond), Then(then), Else(_else) {}
 165   virtual Value *Codegen();
 166 };
 167
 168 /// ForExprAST - Expression class for for/in.
 169 class ForExprAST : public ExprAST {
 170   std::string VarName;
 171   ExprAST *Start, *End, *Step, *Body;
 172 public:
 173   ForExprAST(const std::string &varname, ExprAST *start, ExprAST *end,
 174              ExprAST *step, ExprAST *body)
 175     : VarName(varname), Start(start), End(end), Step(step), Body(body) {}
 176   virtual Value *Codegen();
 177 };
 178
 179 /// VarExprAST - Expression class for var/in
 180 class VarExprAST : public ExprAST {
 181   std::vector<std::pair<std::string, ExprAST*> > VarNames;
 182   ExprAST *Body;
 183 public:
 184   VarExprAST(const std::vector<std::pair<std::string, ExprAST*> > &varnames,
 185              ExprAST *body)
 186   : VarNames(varnames), Body(body) {}
 187
 188   virtual Value *Codegen();
 189 };
 190
 191 /// PrototypeAST - This class represents the "prototype" for a function,
 192 /// which captures its argument names as well as if it is an operator.
 193 class PrototypeAST {
 194   std::string Name;
 195   std::vector<std::string> Args;
 196   bool isOperator;
 197   unsigned Precedence;  // Precedence if a binary op.
 198 public:
 199   PrototypeAST(const std::string &name, const std::vector<std::string> &args,
 200                bool isoperator = false, unsigned prec = 0)
 201   : Name(name), Args(args), isOperator(isoperator), Precedence(prec) {}
 202
 203   bool isUnaryOp() const { return isOperator && Args.size() == 1; }
 204   bool isBinaryOp() const { return isOperator && Args.size() == 2; }
 205
 206   char getOperatorName() const {
 207     assert(isUnaryOp() || isBinaryOp());
 208     return Name[Name.size()-1];
 209   }
 210
 211   unsigned getBinaryPrecedence() const { return Precedence; }
 212
 213   Function *Codegen();
 214
 215   void CreateArgumentAllocas(Function *F);
 216 };
 217
 218 /// FunctionAST - This class represents a function definition itself.
 219 class FunctionAST {
 220   PrototypeAST *Proto;
 221   ExprAST *Body;
 222 public:
 223   FunctionAST(PrototypeAST *proto, ExprAST *body)
 224     : Proto(proto), Body(body) {}
 225
 226   Function *Codegen();
 227 };
 228
 229 //===----------------------------------------------------------------------===//
 230 // Parser
 231 //===----------------------------------------------------------------------===//
 232
 233 /// CurTok/getNextToken - Provide a simple token buffer.  CurTok is the current
 234 /// token the parser it looking at.  getNextToken reads another token from the
 235 /// lexer and updates CurTok with its results.
 236 static int CurTok;
 237 static int getNextToken() {
 238   return CurTok = gettok();
 239 }
 240
 241 /// BinopPrecedence - This holds the precedence for each binary operator that is
 242 /// defined.
 243 static std::map<char, int> BinopPrecedence;
 244
 245 /// GetTokPrecedence - Get the precedence of the pending binary operator token.
 246 static int GetTokPrecedence() {
 247   if (!isascii(CurTok))
 248     return -1;
 249
 250   // Make sure it's a declared binop.
 251   int TokPrec = BinopPrecedence[CurTok];
 252   if (TokPrec <= 0) return -1;
 253   return TokPrec;
 254 }
 255
 256 /// Error* - These are little helper functions for error handling.
 257 ExprAST *Error(const char *Str) { fprintf(stderr, "Error: %s\n", Str);return 0;}
 258 PrototypeAST *ErrorP(const char *Str) { Error(Str); return 0; }
 259 FunctionAST *ErrorF(const char *Str) { Error(Str); return 0; }
 260
 261 static ExprAST *ParseExpression();
 262
 263 /// identifierexpr
 264 ///   ::= identifier
 265 ///   ::= identifier '(' expression* ')'
 266 static ExprAST *ParseIdentifierExpr() {
 267   std::string IdName = IdentifierStr;
 268
 269   getNextToken();  // eat identifier.
 270
 271   if (CurTok != '(') // Simple variable ref.
 272     return new VariableExprAST(IdName);
 273
 274   // Call.
 275   getNextToken();  // eat (
 276   std::vector<ExprAST*> Args;
 277   if (CurTok != ')') {
 278     while (1) {
 279       ExprAST *Arg = ParseExpression();
 280       if (!Arg) return 0;
 281       Args.push_back(Arg);
 282
 283       if (CurTok == ')') break;
 284
 285       if (CurTok != ',')
 286         return Error("Expected ')' or ',' in argument list");
 287       getNextToken();
 288     }
 289   }
 290
 291   // Eat the ')'.
 292   getNextToken();
 293
 294   return new CallExprAST(IdName, Args);
 295 }
 296
 297 /// numberexpr ::= number
 298 static ExprAST *ParseNumberExpr() {
 299   ExprAST *Result = new NumberExprAST(NumVal);
 300   getNextToken(); // consume the number
 301   return Result;
 302 }
 303
 304 /// parenexpr ::= '(' expression ')'
 305 static ExprAST *ParseParenExpr() {
 306   getNextToken();  // eat (.
 307   ExprAST *V = ParseExpression();
 308   if (!V) return 0;
 309
 310   if (CurTok != ')')
 311     return Error("expected ')'");
 312   getNextToken();  // eat ).
 313   return V;
 314 }
 315
 316 /// ifexpr ::= 'if' expression 'then' expression 'else' expression
 317 static ExprAST *ParseIfExpr() {
 318   getNextToken();  // eat the if.
 319
 320   // condition.
 321   ExprAST *Cond = ParseExpression();
 322   if (!Cond) return 0;
 323
 324   if (CurTok != tok_then)
 325     return Error("expected then");
 326   getNextToken();  // eat the then
 327
 328   ExprAST *Then = ParseExpression();
 329   if (Then == 0) return 0;
 330
 331   if (CurTok != tok_else)
 332     return Error("expected else");
 333
 334   getNextToken();
 335
 336   ExprAST *Else = ParseExpression();
 337   if (!Else) return 0;
 338
 339   return new IfExprAST(Cond, Then, Else);
 340 }
 341
 342 /// forexpr ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression
 343 static ExprAST *ParseForExpr() {
 344   getNextToken();  // eat the for.
 345
 346   if (CurTok != tok_identifier)
 347     return Error("expected identifier after for");
 348
 349   std::string IdName = IdentifierStr;
 350   getNextToken();  // eat identifier.
 351
 352   if (CurTok != '=')
 353     return Error("expected '=' after for");
 354   getNextToken();  // eat '='.
 355
 356
 357   ExprAST *Start = ParseExpression();
 358   if (Start == 0) return 0;
 359   if (CurTok != ',')
 360     return Error("expected ',' after for start value");
 361   getNextToken();
 362
 363   ExprAST *End = ParseExpression();
 364   if (End == 0) return 0;
 365
 366   // The step value is optional.
 367   ExprAST *Step = 0;
 368   if (CurTok == ',') {
 369     getNextToken();
 370     Step = ParseExpression();
 371     if (Step == 0) return 0;
 372   }
 373
 374   if (CurTok != tok_in)
 375     return Error("expected 'in' after for");
 376   getNextToken();  // eat 'in'.
 377
 378   ExprAST *Body = ParseExpression();
 379   if (Body == 0) return 0;
 380
 381   return new ForExprAST(IdName, Start, End, Step, Body);
 382 }
 383
 384 /// varexpr ::= 'var' identifier ('=' expression)?
 385 //                    (',' identifier ('=' expression)?)* 'in' expression
 386 static ExprAST *ParseVarExpr() {
 387   getNextToken();  // eat the var.
 388
 389   std::vector<std::pair<std::string, ExprAST*> > VarNames;
 390
 391   // At least one variable name is required.
 392   if (CurTok != tok_identifier)
 393     return Error("expected identifier after var");
 394
 395   while (1) {
 396     std::string Name = IdentifierStr;
 397     getNextToken();  // eat identifier.
 398
 399     // Read the optional initializer.
 400     ExprAST *Init = 0;
 401     if (CurTok == '=') {
 402       getNextToken(); // eat the '='.
 403
 404       Init = ParseExpression();
 405       if (Init == 0) return 0;
 406     }
 407
 408     VarNames.push_back(std::make_pair(Name, Init));
 409
 410     // End of var list, exit loop.
 411     if (CurTok != ',') break;
 412     getNextToken(); // eat the ','.
 413
 414     if (CurTok != tok_identifier)
 415       return Error("expected identifier list after var");
 416   }
 417
 418   // At this point, we have to have 'in'.
 419   if (CurTok != tok_in)
 420     return Error("expected 'in' keyword after 'var'");
 421   getNextToken();  // eat 'in'.
 422
 423   ExprAST *Body = ParseExpression();
 424   if (Body == 0) return 0;
 425
 426   return new VarExprAST(VarNames, Body);
 427 }
 428
 429
 430 /// primary
 431 ///   ::= identifierexpr
 432 ///   ::= numberexpr
 433 ///   ::= parenexpr
 434 ///   ::= ifexpr
 435 ///   ::= forexpr
 436 ///   ::= varexpr
 437 static ExprAST *ParsePrimary() {
 438   switch (CurTok) {
 439   default: return Error("unknown token when expecting an expression");
 440   case tok_identifier: return ParseIdentifierExpr();
 441   case tok_number:     return ParseNumberExpr();
 442   case '(':            return ParseParenExpr();
 443   case tok_if:         return ParseIfExpr();
 444   case tok_for:        return ParseForExpr();
 445   case tok_var:        return ParseVarExpr();
 446   }
 447 }
 448
 449 /// unary
 450 ///   ::= primary
 451 ///   ::= '!' unary
 452 static ExprAST *ParseUnary() {
 453   // If the current token is not an operator, it must be a primary expr.
 454   if (!isascii(CurTok) || CurTok == '(' || CurTok == ',')
 455     return ParsePrimary();
 456
 457   // If this is a unary operator, read it.
 458   int Opc = CurTok;
 459   getNextToken();
 460   if (ExprAST *Operand = ParseUnary())
 461     return new UnaryExprAST(Opc, Operand);
 462   return 0;
 463 }
 464
 465 /// binoprhs
 466 ///   ::= ('+' unary)*
 467 static ExprAST *ParseBinOpRHS(int ExprPrec, ExprAST *LHS) {
 468   // If this is a binop, find its precedence.
 469   while (1) {
 470     int TokPrec = GetTokPrecedence();
 471
 472     // If this is a binop that binds at least as tightly as the current binop,
 473     // consume it, otherwise we are done.
 474     if (TokPrec < ExprPrec)
 475       return LHS;
 476
 477     // Okay, we know this is a binop.
 478     int BinOp = CurTok;
 479     getNextToken();  // eat binop
 480
 481     // Parse the unary expression after the binary operator.
 482     ExprAST *RHS = ParseUnary();
 483     if (!RHS) return 0;
 484
 485     // If BinOp binds less tightly with RHS than the operator after RHS, let
 486     // the pending operator take RHS as its LHS.
 487     int NextPrec = GetTokPrecedence();
 488     if (TokPrec < NextPrec) {
 489       RHS = ParseBinOpRHS(TokPrec+1, RHS);
 490       if (RHS == 0) return 0;
 491     }
 492
 493     // Merge LHS/RHS.
 494     LHS = new BinaryExprAST(BinOp, LHS, RHS);
 495   }
 496 }
 497
 498 /// expression
 499 ///   ::= unary binoprhs
 500 ///
 501 static ExprAST *ParseExpression() {
 502   ExprAST *LHS = ParseUnary();
 503   if (!LHS) return 0;
 504
 505   return ParseBinOpRHS(0, LHS);
 506 }
 507
 508 /// prototype
 509 ///   ::= id '(' id* ')'
 510 ///   ::= binary LETTER number? (id, id)
 511 ///   ::= unary LETTER (id)
 512 static PrototypeAST *ParsePrototype() {
 513   std::string FnName;
 514
 515   unsigned Kind = 0;            // 0 = identifier, 1 = unary, 2 = binary.
 516   unsigned BinaryPrecedence = 30;
 517
 518   switch (CurTok) {
 519   default:
 520     return ErrorP("Expected function name in prototype");
 521   case tok_identifier:
 522     FnName = IdentifierStr;
 523     Kind = 0;
 524     getNextToken();
 525     break;
 526   case tok_unary:
 527     getNextToken();
 528     if (!isascii(CurTok))
 529       return ErrorP("Expected unary operator");
 530     FnName = "unary";
 531     FnName += (char)CurTok;
 532     Kind = 1;
 533     getNextToken();
 534     break;
 535   case tok_binary:
 536     getNextToken();
 537     if (!isascii(CurTok))
 538       return ErrorP("Expected binary operator");
 539     FnName = "binary";
 540     FnName += (char)CurTok;
 541     Kind = 2;
 542     getNextToken();
 543
 544     // Read the precedence if present.
 545     if (CurTok == tok_number) {
 546       if (NumVal < 1 || NumVal > 100)
 547         return ErrorP("Invalid precedecnce: must be 1..100");
 548       BinaryPrecedence = (unsigned)NumVal;
 549       getNextToken();
 550     }
 551     break;
 552   }
 553
 554   if (CurTok != '(')
 555     return ErrorP("Expected '(' in prototype");
 556
 557   std::vector<std::string> ArgNames;
 558   while (getNextToken() == tok_identifier)
 559     ArgNames.push_back(IdentifierStr);
 560   if (CurTok != ')')
 561     return ErrorP("Expected ')' in prototype");
 562
 563   // success.
 564   getNextToken();  // eat ')'.
 565
 566   // Verify right number of names for operator.
 567   if (Kind && ArgNames.size() != Kind)
 568     return ErrorP("Invalid number of operands for operator");
 569
 570   return new PrototypeAST(FnName, ArgNames, Kind != 0, BinaryPrecedence);
 571 }
 572
 573 /// definition ::= 'def' prototype expression
 574 static FunctionAST *ParseDefinition() {
 575   getNextToken();  // eat def.
 576   PrototypeAST *Proto = ParsePrototype();
 577   if (Proto == 0) return 0;
 578
 579   if (ExprAST *E = ParseExpression())
 580     return new FunctionAST(Proto, E);
 581   return 0;
 582 }
 583
 584 /// toplevelexpr ::= expression
 585 static FunctionAST *ParseTopLevelExpr() {
 586   if (ExprAST *E = ParseExpression()) {
 587     // Make an anonymous proto.
 588     PrototypeAST *Proto = new PrototypeAST("", std::vector<std::string>());
 589     return new FunctionAST(Proto, E);
 590   }
 591   return 0;
 592 }
 593
 594 /// external ::= 'extern' prototype
 595 static PrototypeAST *ParseExtern() {
 596   getNextToken();  // eat extern.
 597   return ParsePrototype();
 598 }
 599
 600 //===----------------------------------------------------------------------===//
 601 // Code Generation
 602 //===----------------------------------------------------------------------===//
 603
 604 static Module *TheModule;
 605 static IRBuilder<> Builder;
 606 static std::map<std::string, AllocaInst*> NamedValues;
 607 static FunctionPassManager *TheFPM;
 608
 609 Value *ErrorV(const char *Str) { Error(Str); return 0; }
 610
 611 /// CreateEntryBlockAlloca - Create an alloca instruction in the entry block of
 612 /// the function.  This is used for mutable variables etc.
 613 static AllocaInst *CreateEntryBlockAlloca(Function *TheFunction,
 614                                           const std::string &VarName) {
 615   IRBuilder<> TmpB(&TheFunction->getEntryBlock(),
 616                  TheFunction->getEntryBlock().begin());
 617   return TmpB.CreateAlloca(Type::DoubleTy, 0, VarName.c_str());
 618 }
 619
 620
 621 Value *NumberExprAST::Codegen() {
 622   return ConstantFP::get(APFloat(Val));
 623 }
 624
 625 Value *VariableExprAST::Codegen() {
 626   // Look this variable up in the function.
 627   Value *V = NamedValues[Name];
 628   if (V == 0) return ErrorV("Unknown variable name");
 629
 630   // Load the value.
 631   return Builder.CreateLoad(V, Name.c_str());
 632 }
 633
 634 Value *UnaryExprAST::Codegen() {
 635   Value *OperandV = Operand->Codegen();
 636   if (OperandV == 0) return 0;
 637
 638   Function *F = TheModule->getFunction(std::string("unary")+Opcode);
 639   if (F == 0)
 640     return ErrorV("Unknown unary operator");
 641
 642   return Builder.CreateCall(F, OperandV, "unop");
 643 }
 644
 645
 646 Value *BinaryExprAST::Codegen() {
 647   // Special case '=' because we don't want to emit the LHS as an expression.
 648   if (Op == '=') {
 649     // Assignment requires the LHS to be an identifier.
 650     VariableExprAST *LHSE = dynamic_cast<VariableExprAST*>(LHS);
 651     if (!LHSE)
 652       return ErrorV("destination of '=' must be a variable");
 653     // Codegen the RHS.
 654     Value *Val = RHS->Codegen();
 655     if (Val == 0) return 0;
 656
 657     // Look up the name.
 658     Value *Variable = NamedValues[LHSE->getName()];
 659     if (Variable == 0) return ErrorV("Unknown variable name");
 660
 661     Builder.CreateStore(Val, Variable);
 662     return Val;
 663   }
 664
 665
 666   Value *L = LHS->Codegen();
 667   Value *R = RHS->Codegen();
 668   if (L == 0 || R == 0) return 0;
 669
 670   switch (Op) {
 671   case '+': return Builder.CreateAdd(L, R, "addtmp");
 672   case '-': return Builder.CreateSub(L, R, "subtmp");
 673   case '*': return Builder.CreateMul(L, R, "multmp");
 674   case '<':
 675     L = Builder.CreateFCmpULT(L, R, "cmptmp");
 676     // Convert bool 0/1 to double 0.0 or 1.0
 677     return Builder.CreateUIToFP(L, Type::DoubleTy, "booltmp");
 678   default: break;
 679   }
 680
 681   // If it wasn't a builtin binary operator, it must be a user defined one. Emit
 682   // a call to it.
 683   Function *F = TheModule->getFunction(std::string("binary")+Op);
 684   assert(F && "binary operator not found!");
 685
 686   Value *Ops[] = { L, R };
 687   return Builder.CreateCall(F, Ops, Ops+2, "binop");
 688 }
 689
 690 Value *CallExprAST::Codegen() {
 691   // Look up the name in the global module table.
 692   Function *CalleeF = TheModule->getFunction(Callee);
 693   if (CalleeF == 0)
 694     return ErrorV("Unknown function referenced");
 695
 696   // If argument mismatch error.
 697   if (CalleeF->arg_size() != Args.size())
 698     return ErrorV("Incorrect # arguments passed");
 699
 700   std::vector<Value*> ArgsV;
 701   for (unsigned i = 0, e = Args.size(); i != e; ++i) {
 702     ArgsV.push_back(Args[i]->Codegen());
 703     if (ArgsV.back() == 0) return 0;
 704   }
 705
 706   return Builder.CreateCall(CalleeF, ArgsV.begin(), ArgsV.end(), "calltmp");
 707 }
 708
 709 Value *IfExprAST::Codegen() {
 710   Value *CondV = Cond->Codegen();
 711   if (CondV == 0) return 0;
 712
 713   // Convert condition to a bool by comparing equal to 0.0.
 714   CondV = Builder.CreateFCmpONE(CondV,
 715                                 ConstantFP::get(APFloat(0.0)),
 716                                 "ifcond");
 717
 718   Function *TheFunction = Builder.GetInsertBlock()->getParent();
 719
 720   // Create blocks for the then and else cases.  Insert the 'then' block at the
 721   // end of the function.
 722   BasicBlock *ThenBB = BasicBlock::Create("then", TheFunction);
 723   BasicBlock *ElseBB = BasicBlock::Create("else");
 724   BasicBlock *MergeBB = BasicBlock::Create("ifcont");
 725
 726   Builder.CreateCondBr(CondV, ThenBB, ElseBB);
 727
 728   // Emit then value.
 729   Builder.SetInsertPoint(ThenBB);
 730
 731   Value *ThenV = Then->Codegen();
 732   if (ThenV == 0) return 0;
 733
 734   Builder.CreateBr(MergeBB);
 735   // Codegen of 'Then' can change the current block, update ThenBB for the PHI.
 736   ThenBB = Builder.GetInsertBlock();
 737
 738   // Emit else block.
 739   TheFunction->getBasicBlockList().push_back(ElseBB);
 740   Builder.SetInsertPoint(ElseBB);
 741
 742   Value *ElseV = Else->Codegen();
 743   if (ElseV == 0) return 0;
 744
 745   Builder.CreateBr(MergeBB);
 746   // Codegen of 'Else' can change the current block, update ElseBB for the PHI.
 747   ElseBB = Builder.GetInsertBlock();
 748
 749   // Emit merge block.
 750   TheFunction->getBasicBlockList().push_back(MergeBB);
 751   Builder.SetInsertPoint(MergeBB);
 752   PHINode *PN = Builder.CreatePHI(Type::DoubleTy, "iftmp");
 753
 754   PN->addIncoming(ThenV, ThenBB);
 755   PN->addIncoming(ElseV, ElseBB);
 756   return PN;
 757 }
 758
 759 Value *ForExprAST::Codegen() {
 760   // Output this as:
 761   //   var = alloca double
 762   //   ...
 763   //   start = startexpr
 764   //   store start -> var
 765   //   goto loop
 766   // loop:
 767   //   ...
 768   //   bodyexpr
 769   //   ...
 770   // loopend:
 771   //   step = stepexpr
 772   //   endcond = endexpr
 773   //
 774   //   curvar = load var
 775   //   nextvar = curvar + step
 776   //   store nextvar -> var
 777   //   br endcond, loop, endloop
 778   // outloop:
 779
 780   Function *TheFunction = Builder.GetInsertBlock()->getParent();
 781
 782   // Create an alloca for the variable in the entry block.
 783   AllocaInst *Alloca = CreateEntryBlockAlloca(TheFunction, VarName);
 784
 785   // Emit the start code first, without 'variable' in scope.
 786   Value *StartVal = Start->Codegen();
 787   if (StartVal == 0) return 0;
 788
 789   // Store the value into the alloca.
 790   Builder.CreateStore(StartVal, Alloca);
 791
 792   // Make the new basic block for the loop header, inserting after current
 793   // block.
 794   BasicBlock *LoopBB = BasicBlock::Create("loop", TheFunction);
 795
 796   // Insert an explicit fall through from the current block to the LoopBB.
 797   Builder.CreateBr(LoopBB);
 798
 799   // Start insertion in LoopBB.
 800   Builder.SetInsertPoint(LoopBB);
 801
 802   // Within the loop, the variable is defined equal to the PHI node.  If it
 803   // shadows an existing variable, we have to restore it, so save it now.
 804   AllocaInst *OldVal = NamedValues[VarName];
 805   NamedValues[VarName] = Alloca;
 806
 807   // Emit the body of the loop.  This, like any other expr, can change the
 808   // current BB.  Note that we ignore the value computed by the body, but don't
 809   // allow an error.
 810   if (Body->Codegen() == 0)
 811     return 0;
 812
 813   // Emit the step value.
 814   Value *StepVal;
 815   if (Step) {
 816     StepVal = Step->Codegen();
 817     if (StepVal == 0) return 0;
 818   } else {
 819     // If not specified, use 1.0.
 820     StepVal = ConstantFP::get(APFloat(1.0));
 821   }
 822
 823   // Compute the end condition.
 824   Value *EndCond = End->Codegen();
 825   if (EndCond == 0) return EndCond;
 826
 827   // Reload, increment, and restore the alloca.  This handles the case where
 828   // the body of the loop mutates the variable.
 829   Value *CurVar = Builder.CreateLoad(Alloca, VarName.c_str());
 830   Value *NextVar = Builder.CreateAdd(CurVar, StepVal, "nextvar");
 831   Builder.CreateStore(NextVar, Alloca);
 832
 833   // Convert condition to a bool by comparing equal to 0.0.
 834   EndCond = Builder.CreateFCmpONE(EndCond,
 835                                   ConstantFP::get(APFloat(0.0)),
 836                                   "loopcond");
 837
 838   // Create the "after loop" block and insert it.
 839   BasicBlock *AfterBB = BasicBlock::Create("afterloop", TheFunction);
 840
 841   // Insert the conditional branch into the end of LoopEndBB.
 842   Builder.CreateCondBr(EndCond, LoopBB, AfterBB);
 843
 844   // Any new code will be inserted in AfterBB.
 845   Builder.SetInsertPoint(AfterBB);
 846
 847   // Restore the unshadowed variable.
 848   if (OldVal)
 849     NamedValues[VarName] = OldVal;
 850   else
 851     NamedValues.erase(VarName);
 852
 853
 854   // for expr always returns 0.0.
 855   return Constant::getNullValue(Type::DoubleTy);
 856 }
 857
 858 Value *VarExprAST::Codegen() {
 859   std::vector<AllocaInst *> OldBindings;
 860
 861   Function *TheFunction = Builder.GetInsertBlock()->getParent();
 862
 863   // Register all variables and emit their initializer.
 864   for (unsigned i = 0, e = VarNames.size(); i != e; ++i) {
 865     const std::string &VarName = VarNames[i].first;
 866     ExprAST *Init = VarNames[i].second;
 867
 868     // Emit the initializer before adding the variable to scope, this prevents
 869     // the initializer from referencing the variable itself, and permits stuff
 870     // like this:
 871     //  var a = 1 in
 872     //    var a = a in ...   # refers to outer 'a'.
 873     Value *InitVal;
 874     if (Init) {
 875       InitVal = Init->Codegen();
 876       if (InitVal == 0) return 0;
 877     } else { // If not specified, use 0.0.
 878       InitVal = ConstantFP::get(APFloat(0.0));
 879     }
 880
 881     AllocaInst *Alloca = CreateEntryBlockAlloca(TheFunction, VarName);
 882     Builder.CreateStore(InitVal, Alloca);
 883
 884     // Remember the old variable binding so that we can restore the binding when
 885     // we unrecurse.
 886     OldBindings.push_back(NamedValues[VarName]);
 887
 888     // Remember this binding.
 889     NamedValues[VarName] = Alloca;
 890   }
 891
 892   // Codegen the body, now that all vars are in scope.
 893   Value *BodyVal = Body->Codegen();
 894   if (BodyVal == 0) return 0;
 895
 896   // Pop all our variables from scope.
 897   for (unsigned i = 0, e = VarNames.size(); i != e; ++i)
 898     NamedValues[VarNames[i].first] = OldBindings[i];
 899
 900   // Return the body computation.
 901   return BodyVal;
 902 }
 903
 904
 905 Function *PrototypeAST::Codegen() {
 906   // Make the function type:  double(double,double) etc.
 907   std::vector<const Type*> Doubles(Args.size(), Type::DoubleTy);
 908   FunctionType *FT = FunctionType::get(Type::DoubleTy, Doubles, false);
 909
 910   Function *F = Function::Create(FT, Function::ExternalLinkage, Name, TheModule);
 911
 912   // If F conflicted, there was already something named 'Name'.  If it has a
 913   // body, don't allow redefinition or reextern.
 914   if (F->getName() != Name) {
 915     // Delete the one we just made and get the existing one.
 916     F->eraseFromParent();
 917     F = TheModule->getFunction(Name);
 918
 919     // If F already has a body, reject this.
 920     if (!F->empty()) {
 921       ErrorF("redefinition of function");
 922       return 0;
 923     }
 924
 925     // If F took a different number of args, reject.
 926     if (F->arg_size() != Args.size()) {
 927       ErrorF("redefinition of function with different # args");
 928       return 0;
 929     }
 930   }
 931
 932   // Set names for all arguments.
 933   unsigned Idx = 0;
 934   for (Function::arg_iterator AI = F->arg_begin(); Idx != Args.size();
 935        ++AI, ++Idx)
 936     AI->setName(Args[Idx]);
 937
 938   return F;
 939 }
 940
 941 /// CreateArgumentAllocas - Create an alloca for each argument and register the
 942 /// argument in the symbol table so that references to it will succeed.
 943 void PrototypeAST::CreateArgumentAllocas(Function *F) {
 944   Function::arg_iterator AI = F->arg_begin();
 945   for (unsigned Idx = 0, e = Args.size(); Idx != e; ++Idx, ++AI) {
 946     // Create an alloca for this variable.
 947     AllocaInst *Alloca = CreateEntryBlockAlloca(F, Args[Idx]);
 948
 949     // Store the initial value into the alloca.
 950     Builder.CreateStore(AI, Alloca);
 951
 952     // Add arguments to variable symbol table.
 953     NamedValues[Args[Idx]] = Alloca;
 954   }
 955 }
 956
 957
 958 Function *FunctionAST::Codegen() {
 959   NamedValues.clear();
 960
 961   Function *TheFunction = Proto->Codegen();
 962   if (TheFunction == 0)
 963     return 0;
 964
 965   // If this is an operator, install it.
 966   if (Proto->isBinaryOp())
 967     BinopPrecedence[Proto->getOperatorName()] = Proto->getBinaryPrecedence();
 968
 969   // Create a new basic block to start insertion into.
 970   BasicBlock *BB = BasicBlock::Create("entry", TheFunction);
 971   Builder.SetInsertPoint(BB);
 972
 973   // Add all arguments to the symbol table and create their allocas.
 974   Proto->CreateArgumentAllocas(TheFunction);
 975
 976   if (Value *RetVal = Body->Codegen()) {
 977     // Finish off the function.
 978     Builder.CreateRet(RetVal);
 979
 980     // Validate the generated code, checking for consistency.
 981     verifyFunction(*TheFunction);
 982
 983     // Optimize the function.
 984     TheFPM->run(*TheFunction);
 985
 986     return TheFunction;
 987   }
 988
 989   // Error reading body, remove function.
 990   TheFunction->eraseFromParent();
 991
 992   if (Proto->isBinaryOp())
 993     BinopPrecedence.erase(Proto->getOperatorName());
 994   return 0;
 995 }
 996
 997 //===----------------------------------------------------------------------===//
 998 // Top-Level parsing and JIT Driver
 999 //===----------------------------------------------------------------------===//
1000
1001 static ExecutionEngine *TheExecutionEngine;
1002
1003 static void HandleDefinition() {
1004   if (FunctionAST *F = ParseDefinition()) {
1005     if (Function *LF = F->Codegen()) {
1006       fprintf(stderr, "Read function definition:");
1007       LF->dump();
1008     }
1009   } else {
1010     // Skip token for error recovery.
1011     getNextToken();
1012   }
1013 }
1014
1015 static void HandleExtern() {
1016   if (PrototypeAST *P = ParseExtern()) {
1017     if (Function *F = P->Codegen()) {
1018       fprintf(stderr, "Read extern: ");
1019       F->dump();
1020     }
1021   } else {
1022     // Skip token for error recovery.
1023     getNextToken();
1024   }
1025 }
1026
1027 static void HandleTopLevelExpression() {
1028   // Evaluate a top level expression into an anonymous function.
1029   if (FunctionAST *F = ParseTopLevelExpr()) {
1030     if (Function *LF = F->Codegen()) {
1031       // JIT the function, returning a function pointer.
1032       void *FPtr = TheExecutionEngine->getPointerToFunction(LF);
1033
1034       // Cast it to the right type (takes no arguments, returns a double) so we
1035       // can call it as a native function.
1036       double (*FP)() = (double (*)())(intptr_t)FPtr;
1037       fprintf(stderr, "Evaluated to %f\n", FP());
1038     }
1039   } else {
1040     // Skip token for error recovery.
1041     getNextToken();
1042   }
1043 }
1044
1045 /// top ::= definition | external | expression | ';'
1046 static void MainLoop() {
1047   while (1) {
1048     fprintf(stderr, "ready> ");
1049     switch (CurTok) {
1050     case tok_eof:    return;
1051     case ';':        getNextToken(); break;  // ignore top level semicolons.
1052     case tok_def:    HandleDefinition(); break;
1053     case tok_extern: HandleExtern(); break;
1054     default:         HandleTopLevelExpression(); break;
1055     }
1056   }
1057 }
1058
1059
1060
1061 //===----------------------------------------------------------------------===//
1062 // "Library" functions that can be "extern'd" from user code.
1063 //===----------------------------------------------------------------------===//
1064
1065 /// putchard - putchar that takes a double and returns 0.
1066 extern "C"
1067 double putchard(double X) {
1068   putchar((char)X);
1069   return 0;
1070 }
1071
1072 /// printd - printf that takes a double prints it as "%f\n", returning 0.
1073 extern "C"
1074 double printd(double X) {
1075   printf("%f\n", X);
1076   return 0;
1077 }
1078
1079 //===----------------------------------------------------------------------===//
1080 // Main driver code.
1081 //===----------------------------------------------------------------------===//
1082
1083 int main() {
1084   // Install standard binary operators.
1085   // 1 is lowest precedence.
1086   BinopPrecedence['='] = 2;
1087   BinopPrecedence['<'] = 10;
1088   BinopPrecedence['+'] = 20;
1089   BinopPrecedence['-'] = 20;
1090   BinopPrecedence['*'] = 40;  // highest.
1091
1092   // Prime the first token.
1093   fprintf(stderr, "ready> ");
1094   getNextToken();
1095
1096   // Make the module, which holds all the code.
1097   TheModule = new Module("my cool jit");
1098
1099   // Create the JIT.
1100   TheExecutionEngine = ExecutionEngine::create(TheModule);
1101
1102   {
1103     ExistingModuleProvider OurModuleProvider(TheModule);
1104     FunctionPassManager OurFPM(&OurModuleProvider);
1105
1106     // Set up the optimizer pipeline.  Start with registering info about how the
1107     // target lays out data structures.
1108     OurFPM.add(new TargetData(*TheExecutionEngine->getTargetData()));
1109     // Promote allocas to registers.
1110     OurFPM.add(createPromoteMemoryToRegisterPass());
1111     // Do simple "peephole" optimizations and bit-twiddling optzns.
1112     OurFPM.add(createInstructionCombiningPass());
1113     // Reassociate expressions.
1114     OurFPM.add(createReassociatePass());
1115     // Eliminate Common SubExpressions.
1116     OurFPM.add(createGVNPass());
1117     // Simplify the control flow graph (deleting unreachable blocks, etc).
1118     OurFPM.add(createCFGSimplificationPass());
1119
1120     // Set the global so the code gen can use this.
1121     TheFPM = &OurFPM;
1122
1123     // Run the main "interpreter loop" now.
1124     MainLoop();
1125
1126     TheFPM = 0;
1127
1128     // Print out all of the generated code.
1129     TheModule->dump();
1130
1131   }  // Free module provider (and thus the module) and pass manager.
1132
1133   return 0;
1134 }
1135