examples/Kaleidoscope/toy.cpp

   1 #include "llvm/DerivedTypes.h"
   2 #include "llvm/ExecutionEngine/ExecutionEngine.h"
   3 #include "llvm/ExecutionEngine/Interpreter.h"
   4 #include "llvm/ExecutionEngine/JIT.h"
   5 #include "llvm/LLVMContext.h"
   6 #include "llvm/Module.h"
   7 #include "llvm/ModuleProvider.h"
   8 #include "llvm/PassManager.h"
   9 #include "llvm/Analysis/Verifier.h"
  10 #include "llvm/Target/TargetData.h"
  11 #include "llvm/Target/TargetSelect.h"
  12 #include "llvm/Transforms/Scalar.h"
  13 #include "llvm/Support/IRBuilder.h"
  14 #include <cstdio>
  15 #include <string>
  16 #include <map>
  17 #include <vector>
  18 using namespace llvm;
  19
  20 //===----------------------------------------------------------------------===//
  21 // Lexer
  22 //===----------------------------------------------------------------------===//
  23
  24 // The lexer returns tokens [0-255] if it is an unknown character, otherwise one
  25 // of these for known things.
  26 enum Token {
  27   tok_eof = -1,
  28
  29   // commands
  30   tok_def = -2, tok_extern = -3,
  31
  32   // primary
  33   tok_identifier = -4, tok_number = -5,
  34
  35   // control
  36   tok_if = -6, tok_then = -7, tok_else = -8,
  37   tok_for = -9, tok_in = -10,
  38
  39   // operators
  40   tok_binary = -11, tok_unary = -12,
  41
  42   // var definition
  43   tok_var = -13
  44 };
  45
  46 static std::string IdentifierStr;  // Filled in if tok_identifier
  47 static double NumVal;              // Filled in if tok_number
  48
  49 /// gettok - Return the next token from standard input.
  50 static int gettok() {
  51   static int LastChar = ' ';
  52
  53   // Skip any whitespace.
  54   while (isspace(LastChar))
  55     LastChar = getchar();
  56
  57   if (isalpha(LastChar)) { // identifier: [a-zA-Z][a-zA-Z0-9]*
  58     IdentifierStr = LastChar;
  59     while (isalnum((LastChar = getchar())))
  60       IdentifierStr += LastChar;
  61
  62     if (IdentifierStr == "def") return tok_def;
  63     if (IdentifierStr == "extern") return tok_extern;
  64     if (IdentifierStr == "if") return tok_if;
  65     if (IdentifierStr == "then") return tok_then;
  66     if (IdentifierStr == "else") return tok_else;
  67     if (IdentifierStr == "for") return tok_for;
  68     if (IdentifierStr == "in") return tok_in;
  69     if (IdentifierStr == "binary") return tok_binary;
  70     if (IdentifierStr == "unary") return tok_unary;
  71     if (IdentifierStr == "var") return tok_var;
  72     return tok_identifier;
  73   }
  74
  75   if (isdigit(LastChar) || LastChar == '.') {   // Number: [0-9.]+
  76     std::string NumStr;
  77     do {
  78       NumStr += LastChar;
  79       LastChar = getchar();
  80     } while (isdigit(LastChar) || LastChar == '.');
  81
  82     NumVal = strtod(NumStr.c_str(), 0);
  83     return tok_number;
  84   }
  85
  86   if (LastChar == '#') {
  87     // Comment until end of line.
  88     do LastChar = getchar();
  89     while (LastChar != EOF && LastChar != '\n' && LastChar != '\r');
  90
  91     if (LastChar != EOF)
  92       return gettok();
  93   }
  94
  95   // Check for end of file.  Don't eat the EOF.
  96   if (LastChar == EOF)
  97     return tok_eof;
  98
  99   // Otherwise, just return the character as its ascii value.
 100   int ThisChar = LastChar;
 101   LastChar = getchar();
 102   return ThisChar;
 103 }
 104
 105 //===----------------------------------------------------------------------===//
 106 // Abstract Syntax Tree (aka Parse Tree)
 107 //===----------------------------------------------------------------------===//
 108
 109 /// ExprAST - Base class for all expression nodes.
 110 class ExprAST {
 111 public:
 112   virtual ~ExprAST() {}
 113   virtual Value *Codegen() = 0;
 114 };
 115
 116 /// NumberExprAST - Expression class for numeric literals like "1.0".
 117 class NumberExprAST : public ExprAST {
 118   double Val;
 119 public:
 120   NumberExprAST(double val) : Val(val) {}
 121   virtual Value *Codegen();
 122 };
 123
 124 /// VariableExprAST - Expression class for referencing a variable, like "a".
 125 class VariableExprAST : public ExprAST {
 126   std::string Name;
 127 public:
 128   VariableExprAST(const std::string &name) : Name(name) {}
 129   const std::string &getName() const { return Name; }
 130   virtual Value *Codegen();
 131 };
 132
 133 /// UnaryExprAST - Expression class for a unary operator.
 134 class UnaryExprAST : public ExprAST {
 135   char Opcode;
 136   ExprAST *Operand;
 137 public:
 138   UnaryExprAST(char opcode, ExprAST *operand)
 139     : Opcode(opcode), Operand(operand) {}
 140   virtual Value *Codegen();
 141 };
 142
 143 /// BinaryExprAST - Expression class for a binary operator.
 144 class BinaryExprAST : public ExprAST {
 145   char Op;
 146   ExprAST *LHS, *RHS;
 147 public:
 148   BinaryExprAST(char op, ExprAST *lhs, ExprAST *rhs)
 149     : Op(op), LHS(lhs), RHS(rhs) {}
 150   virtual Value *Codegen();
 151 };
 152
 153 /// CallExprAST - Expression class for function calls.
 154 class CallExprAST : public ExprAST {
 155   std::string Callee;
 156   std::vector<ExprAST*> Args;
 157 public:
 158   CallExprAST(const std::string &callee, std::vector<ExprAST*> &args)
 159     : Callee(callee), Args(args) {}
 160   virtual Value *Codegen();
 161 };
 162
 163 /// IfExprAST - Expression class for if/then/else.
 164 class IfExprAST : public ExprAST {
 165   ExprAST *Cond, *Then, *Else;
 166 public:
 167   IfExprAST(ExprAST *cond, ExprAST *then, ExprAST *_else)
 168   : Cond(cond), Then(then), Else(_else) {}
 169   virtual Value *Codegen();
 170 };
 171
 172 /// ForExprAST - Expression class for for/in.
 173 class ForExprAST : public ExprAST {
 174   std::string VarName;
 175   ExprAST *Start, *End, *Step, *Body;
 176 public:
 177   ForExprAST(const std::string &varname, ExprAST *start, ExprAST *end,
 178              ExprAST *step, ExprAST *body)
 179     : VarName(varname), Start(start), End(end), Step(step), Body(body) {}
 180   virtual Value *Codegen();
 181 };
 182
 183 /// VarExprAST - Expression class for var/in
 184 class VarExprAST : public ExprAST {
 185   std::vector<std::pair<std::string, ExprAST*> > VarNames;
 186   ExprAST *Body;
 187 public:
 188   VarExprAST(const std::vector<std::pair<std::string, ExprAST*> > &varnames,
 189              ExprAST *body)
 190   : VarNames(varnames), Body(body) {}
 191
 192   virtual Value *Codegen();
 193 };
 194
 195 /// PrototypeAST - This class represents the "prototype" for a function,
 196 /// which captures its argument names as well as if it is an operator.
 197 class PrototypeAST {
 198   std::string Name;
 199   std::vector<std::string> Args;
 200   bool isOperator;
 201   unsigned Precedence;  // Precedence if a binary op.
 202 public:
 203   PrototypeAST(const std::string &name, const std::vector<std::string> &args,
 204                bool isoperator = false, unsigned prec = 0)
 205   : Name(name), Args(args), isOperator(isoperator), Precedence(prec) {}
 206
 207   bool isUnaryOp() const { return isOperator && Args.size() == 1; }
 208   bool isBinaryOp() const { return isOperator && Args.size() == 2; }
 209
 210   char getOperatorName() const {
 211     assert(isUnaryOp() || isBinaryOp());
 212     return Name[Name.size()-1];
 213   }
 214
 215   unsigned getBinaryPrecedence() const { return Precedence; }
 216
 217   Function *Codegen();
 218
 219   void CreateArgumentAllocas(Function *F);
 220 };
 221
 222 /// FunctionAST - This class represents a function definition itself.
 223 class FunctionAST {
 224   PrototypeAST *Proto;
 225   ExprAST *Body;
 226 public:
 227   FunctionAST(PrototypeAST *proto, ExprAST *body)
 228     : Proto(proto), Body(body) {}
 229
 230   Function *Codegen();
 231 };
 232
 233 //===----------------------------------------------------------------------===//
 234 // Parser
 235 //===----------------------------------------------------------------------===//
 236
 237 /// CurTok/getNextToken - Provide a simple token buffer.  CurTok is the current
 238 /// token the parser it looking at.  getNextToken reads another token from the
 239 /// lexer and updates CurTok with its results.
 240 static int CurTok;
 241 static int getNextToken() {
 242   return CurTok = gettok();
 243 }
 244
 245 /// BinopPrecedence - This holds the precedence for each binary operator that is
 246 /// defined.
 247 static std::map<char, int> BinopPrecedence;
 248
 249 /// GetTokPrecedence - Get the precedence of the pending binary operator token.
 250 static int GetTokPrecedence() {
 251   if (!isascii(CurTok))
 252     return -1;
 253
 254   // Make sure it's a declared binop.
 255   int TokPrec = BinopPrecedence[CurTok];
 256   if (TokPrec <= 0) return -1;
 257   return TokPrec;
 258 }
 259
 260 /// Error* - These are little helper functions for error handling.
 261 ExprAST *Error(const char *Str) { fprintf(stderr, "Error: %s\n", Str);return 0;}
 262 PrototypeAST *ErrorP(const char *Str) { Error(Str); return 0; }
 263 FunctionAST *ErrorF(const char *Str) { Error(Str); return 0; }
 264
 265 static ExprAST *ParseExpression();
 266
 267 /// identifierexpr
 268 ///   ::= identifier
 269 ///   ::= identifier '(' expression* ')'
 270 static ExprAST *ParseIdentifierExpr() {
 271   std::string IdName = IdentifierStr;
 272
 273   getNextToken();  // eat identifier.
 274
 275   if (CurTok != '(') // Simple variable ref.
 276     return new VariableExprAST(IdName);
 277
 278   // Call.
 279   getNextToken();  // eat (
 280   std::vector<ExprAST*> Args;
 281   if (CurTok != ')') {
 282     while (1) {
 283       ExprAST *Arg = ParseExpression();
 284       if (!Arg) return 0;
 285       Args.push_back(Arg);
 286
 287       if (CurTok == ')') break;
 288
 289       if (CurTok != ',')
 290         return Error("Expected ')' or ',' in argument list");
 291       getNextToken();
 292     }
 293   }
 294
 295   // Eat the ')'.
 296   getNextToken();
 297
 298   return new CallExprAST(IdName, Args);
 299 }
 300
 301 /// numberexpr ::= number
 302 static ExprAST *ParseNumberExpr() {
 303   ExprAST *Result = new NumberExprAST(NumVal);
 304   getNextToken(); // consume the number
 305   return Result;
 306 }
 307
 308 /// parenexpr ::= '(' expression ')'
 309 static ExprAST *ParseParenExpr() {
 310   getNextToken();  // eat (.
 311   ExprAST *V = ParseExpression();
 312   if (!V) return 0;
 313
 314   if (CurTok != ')')
 315     return Error("expected ')'");
 316   getNextToken();  // eat ).
 317   return V;
 318 }
 319
 320 /// ifexpr ::= 'if' expression 'then' expression 'else' expression
 321 static ExprAST *ParseIfExpr() {
 322   getNextToken();  // eat the if.
 323
 324   // condition.
 325   ExprAST *Cond = ParseExpression();
 326   if (!Cond) return 0;
 327
 328   if (CurTok != tok_then)
 329     return Error("expected then");
 330   getNextToken();  // eat the then
 331
 332   ExprAST *Then = ParseExpression();
 333   if (Then == 0) return 0;
 334
 335   if (CurTok != tok_else)
 336     return Error("expected else");
 337
 338   getNextToken();
 339
 340   ExprAST *Else = ParseExpression();
 341   if (!Else) return 0;
 342
 343   return new IfExprAST(Cond, Then, Else);
 344 }
 345
 346 /// forexpr ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression
 347 static ExprAST *ParseForExpr() {
 348   getNextToken();  // eat the for.
 349
 350   if (CurTok != tok_identifier)
 351     return Error("expected identifier after for");
 352
 353   std::string IdName = IdentifierStr;
 354   getNextToken();  // eat identifier.
 355
 356   if (CurTok != '=')
 357     return Error("expected '=' after for");
 358   getNextToken();  // eat '='.
 359
 360
 361   ExprAST *Start = ParseExpression();
 362   if (Start == 0) return 0;
 363   if (CurTok != ',')
 364     return Error("expected ',' after for start value");
 365   getNextToken();
 366
 367   ExprAST *End = ParseExpression();
 368   if (End == 0) return 0;
 369
 370   // The step value is optional.
 371   ExprAST *Step = 0;
 372   if (CurTok == ',') {
 373     getNextToken();
 374     Step = ParseExpression();
 375     if (Step == 0) return 0;
 376   }
 377
 378   if (CurTok != tok_in)
 379     return Error("expected 'in' after for");
 380   getNextToken();  // eat 'in'.
 381
 382   ExprAST *Body = ParseExpression();
 383   if (Body == 0) return 0;
 384
 385   return new ForExprAST(IdName, Start, End, Step, Body);
 386 }
 387
 388 /// varexpr ::= 'var' identifier ('=' expression)?
 389 //                    (',' identifier ('=' expression)?)* 'in' expression
 390 static ExprAST *ParseVarExpr() {
 391   getNextToken();  // eat the var.
 392
 393   std::vector<std::pair<std::string, ExprAST*> > VarNames;
 394
 395   // At least one variable name is required.
 396   if (CurTok != tok_identifier)
 397     return Error("expected identifier after var");
 398
 399   while (1) {
 400     std::string Name = IdentifierStr;
 401     getNextToken();  // eat identifier.
 402
 403     // Read the optional initializer.
 404     ExprAST *Init = 0;
 405     if (CurTok == '=') {
 406       getNextToken(); // eat the '='.
 407
 408       Init = ParseExpression();
 409       if (Init == 0) return 0;
 410     }
 411
 412     VarNames.push_back(std::make_pair(Name, Init));
 413
 414     // End of var list, exit loop.
 415     if (CurTok != ',') break;
 416     getNextToken(); // eat the ','.
 417
 418     if (CurTok != tok_identifier)
 419       return Error("expected identifier list after var");
 420   }
 421
 422   // At this point, we have to have 'in'.
 423   if (CurTok != tok_in)
 424     return Error("expected 'in' keyword after 'var'");
 425   getNextToken();  // eat 'in'.
 426
 427   ExprAST *Body = ParseExpression();
 428   if (Body == 0) return 0;
 429
 430   return new VarExprAST(VarNames, Body);
 431 }
 432
 433
 434 /// primary
 435 ///   ::= identifierexpr
 436 ///   ::= numberexpr
 437 ///   ::= parenexpr
 438 ///   ::= ifexpr
 439 ///   ::= forexpr
 440 ///   ::= varexpr
 441 static ExprAST *ParsePrimary() {
 442   switch (CurTok) {
 443   default: return Error("unknown token when expecting an expression");
 444   case tok_identifier: return ParseIdentifierExpr();
 445   case tok_number:     return ParseNumberExpr();
 446   case '(':            return ParseParenExpr();
 447   case tok_if:         return ParseIfExpr();
 448   case tok_for:        return ParseForExpr();
 449   case tok_var:        return ParseVarExpr();
 450   }
 451 }
 452
 453 /// unary
 454 ///   ::= primary
 455 ///   ::= '!' unary
 456 static ExprAST *ParseUnary() {
 457   // If the current token is not an operator, it must be a primary expr.
 458   if (!isascii(CurTok) || CurTok == '(' || CurTok == ',')
 459     return ParsePrimary();
 460
 461   // If this is a unary operator, read it.
 462   int Opc = CurTok;
 463   getNextToken();
 464   if (ExprAST *Operand = ParseUnary())
 465     return new UnaryExprAST(Opc, Operand);
 466   return 0;
 467 }
 468
 469 /// binoprhs
 470 ///   ::= ('+' unary)*
 471 static ExprAST *ParseBinOpRHS(int ExprPrec, ExprAST *LHS) {
 472   // If this is a binop, find its precedence.
 473   while (1) {
 474     int TokPrec = GetTokPrecedence();
 475
 476     // If this is a binop that binds at least as tightly as the current binop,
 477     // consume it, otherwise we are done.
 478     if (TokPrec < ExprPrec)
 479       return LHS;
 480
 481     // Okay, we know this is a binop.
 482     int BinOp = CurTok;
 483     getNextToken();  // eat binop
 484
 485     // Parse the unary expression after the binary operator.
 486     ExprAST *RHS = ParseUnary();
 487     if (!RHS) return 0;
 488
 489     // If BinOp binds less tightly with RHS than the operator after RHS, let
 490     // the pending operator take RHS as its LHS.
 491     int NextPrec = GetTokPrecedence();
 492     if (TokPrec < NextPrec) {
 493       RHS = ParseBinOpRHS(TokPrec+1, RHS);
 494       if (RHS == 0) return 0;
 495     }
 496
 497     // Merge LHS/RHS.
 498     LHS = new BinaryExprAST(BinOp, LHS, RHS);
 499   }
 500 }
 501
 502 /// expression
 503 ///   ::= unary binoprhs
 504 ///
 505 static ExprAST *ParseExpression() {
 506   ExprAST *LHS = ParseUnary();
 507   if (!LHS) return 0;
 508
 509   return ParseBinOpRHS(0, LHS);
 510 }
 511
 512 /// prototype
 513 ///   ::= id '(' id* ')'
 514 ///   ::= binary LETTER number? (id, id)
 515 ///   ::= unary LETTER (id)
 516 static PrototypeAST *ParsePrototype() {
 517   std::string FnName;
 518
 519   unsigned Kind = 0;            // 0 = identifier, 1 = unary, 2 = binary.
 520   unsigned BinaryPrecedence = 30;
 521
 522   switch (CurTok) {
 523   default:
 524     return ErrorP("Expected function name in prototype");
 525   case tok_identifier:
 526     FnName = IdentifierStr;
 527     Kind = 0;
 528     getNextToken();
 529     break;
 530   case tok_unary:
 531     getNextToken();
 532     if (!isascii(CurTok))
 533       return ErrorP("Expected unary operator");
 534     FnName = "unary";
 535     FnName += (char)CurTok;
 536     Kind = 1;
 537     getNextToken();
 538     break;
 539   case tok_binary:
 540     getNextToken();
 541     if (!isascii(CurTok))
 542       return ErrorP("Expected binary operator");
 543     FnName = "binary";
 544     FnName += (char)CurTok;
 545     Kind = 2;
 546     getNextToken();
 547
 548     // Read the precedence if present.
 549     if (CurTok == tok_number) {
 550       if (NumVal < 1 || NumVal > 100)
 551         return ErrorP("Invalid precedecnce: must be 1..100");
 552       BinaryPrecedence = (unsigned)NumVal;
 553       getNextToken();
 554     }
 555     break;
 556   }
 557
 558   if (CurTok != '(')
 559     return ErrorP("Expected '(' in prototype");
 560
 561   std::vector<std::string> ArgNames;
 562   while (getNextToken() == tok_identifier)
 563     ArgNames.push_back(IdentifierStr);
 564   if (CurTok != ')')
 565     return ErrorP("Expected ')' in prototype");
 566
 567   // success.
 568   getNextToken();  // eat ')'.
 569
 570   // Verify right number of names for operator.
 571   if (Kind && ArgNames.size() != Kind)
 572     return ErrorP("Invalid number of operands for operator");
 573
 574   return new PrototypeAST(FnName, ArgNames, Kind != 0, BinaryPrecedence);
 575 }
 576
 577 /// definition ::= 'def' prototype expression
 578 static FunctionAST *ParseDefinition() {
 579   getNextToken();  // eat def.
 580   PrototypeAST *Proto = ParsePrototype();
 581   if (Proto == 0) return 0;
 582
 583   if (ExprAST *E = ParseExpression())
 584     return new FunctionAST(Proto, E);
 585   return 0;
 586 }
 587
 588 /// toplevelexpr ::= expression
 589 static FunctionAST *ParseTopLevelExpr() {
 590   if (ExprAST *E = ParseExpression()) {
 591     // Make an anonymous proto.
 592     PrototypeAST *Proto = new PrototypeAST("", std::vector<std::string>());
 593     return new FunctionAST(Proto, E);
 594   }
 595   return 0;
 596 }
 597
 598 /// external ::= 'extern' prototype
 599 static PrototypeAST *ParseExtern() {
 600   getNextToken();  // eat extern.
 601   return ParsePrototype();
 602 }
 603
 604 //===----------------------------------------------------------------------===//
 605 // Code Generation
 606 //===----------------------------------------------------------------------===//
 607
 608 static Module *TheModule;
 609 static IRBuilder<> Builder(getGlobalContext());
 610 static std::map<std::string, AllocaInst*> NamedValues;
 611 static FunctionPassManager *TheFPM;
 612
 613 Value *ErrorV(const char *Str) { Error(Str); return 0; }
 614
 615 /// CreateEntryBlockAlloca - Create an alloca instruction in the entry block of
 616 /// the function.  This is used for mutable variables etc.
 617 static AllocaInst *CreateEntryBlockAlloca(Function *TheFunction,
 618                                           const std::string &VarName) {
 619   IRBuilder<> TmpB(&TheFunction->getEntryBlock(),
 620                  TheFunction->getEntryBlock().begin());
 621   return TmpB.CreateAlloca(Type::getDoubleTy(getGlobalContext()), 0,
 622                            VarName.c_str());
 623 }
 624
 625
 626 Value *NumberExprAST::Codegen() {
 627   return ConstantFP::get(getGlobalContext(), APFloat(Val));
 628 }
 629
 630 Value *VariableExprAST::Codegen() {
 631   // Look this variable up in the function.
 632   Value *V = NamedValues[Name];
 633   if (V == 0) return ErrorV("Unknown variable name");
 634
 635   // Load the value.
 636   return Builder.CreateLoad(V, Name.c_str());
 637 }
 638
 639 Value *UnaryExprAST::Codegen() {
 640   Value *OperandV = Operand->Codegen();
 641   if (OperandV == 0) return 0;
 642
 643   Function *F = TheModule->getFunction(std::string("unary")+Opcode);
 644   if (F == 0)
 645     return ErrorV("Unknown unary operator");
 646
 647   return Builder.CreateCall(F, OperandV, "unop");
 648 }
 649
 650
 651 Value *BinaryExprAST::Codegen() {
 652   // Special case '=' because we don't want to emit the LHS as an expression.
 653   if (Op == '=') {
 654     // Assignment requires the LHS to be an identifier.
 655     VariableExprAST *LHSE = dynamic_cast<VariableExprAST*>(LHS);
 656     if (!LHSE)
 657       return ErrorV("destination of '=' must be a variable");
 658     // Codegen the RHS.
 659     Value *Val = RHS->Codegen();
 660     if (Val == 0) return 0;
 661
 662     // Look up the name.
 663     Value *Variable = NamedValues[LHSE->getName()];
 664     if (Variable == 0) return ErrorV("Unknown variable name");
 665
 666     Builder.CreateStore(Val, Variable);
 667     return Val;
 668   }
 669
 670
 671   Value *L = LHS->Codegen();
 672   Value *R = RHS->Codegen();
 673   if (L == 0 || R == 0) return 0;
 674
 675   switch (Op) {
 676   case '+': return Builder.CreateAdd(L, R, "addtmp");
 677   case '-': return Builder.CreateSub(L, R, "subtmp");
 678   case '*': return Builder.CreateMul(L, R, "multmp");
 679   case '<':
 680     L = Builder.CreateFCmpULT(L, R, "cmptmp");
 681     // Convert bool 0/1 to double 0.0 or 1.0
 682     return Builder.CreateUIToFP(L, Type::getDoubleTy(getGlobalContext()),
 683                                 "booltmp");
 684   default: break;
 685   }
 686
 687   // If it wasn't a builtin binary operator, it must be a user defined one. Emit
 688   // a call to it.
 689   Function *F = TheModule->getFunction(std::string("binary")+Op);
 690   assert(F && "binary operator not found!");
 691
 692   Value *Ops[] = { L, R };
 693   return Builder.CreateCall(F, Ops, Ops+2, "binop");
 694 }
 695
 696 Value *CallExprAST::Codegen() {
 697   // Look up the name in the global module table.
 698   Function *CalleeF = TheModule->getFunction(Callee);
 699   if (CalleeF == 0)
 700     return ErrorV("Unknown function referenced");
 701
 702   // If argument mismatch error.
 703   if (CalleeF->arg_size() != Args.size())
 704     return ErrorV("Incorrect # arguments passed");
 705
 706   std::vector<Value*> ArgsV;
 707   for (unsigned i = 0, e = Args.size(); i != e; ++i) {
 708     ArgsV.push_back(Args[i]->Codegen());
 709     if (ArgsV.back() == 0) return 0;
 710   }
 711
 712   return Builder.CreateCall(CalleeF, ArgsV.begin(), ArgsV.end(), "calltmp");
 713 }
 714
 715 Value *IfExprAST::Codegen() {
 716   Value *CondV = Cond->Codegen();
 717   if (CondV == 0) return 0;
 718
 719   // Convert condition to a bool by comparing equal to 0.0.
 720   CondV = Builder.CreateFCmpONE(CondV,
 721                               ConstantFP::get(getGlobalContext(), APFloat(0.0)),
 722                                 "ifcond");
 723
 724   Function *TheFunction = Builder.GetInsertBlock()->getParent();
 725
 726   // Create blocks for the then and else cases.  Insert the 'then' block at the
 727   // end of the function.
 728   BasicBlock *ThenBB = BasicBlock::Create(getGlobalContext(), "then", TheFunction);
 729   BasicBlock *ElseBB = BasicBlock::Create(getGlobalContext(), "else");
 730   BasicBlock *MergeBB = BasicBlock::Create(getGlobalContext(), "ifcont");
 731
 732   Builder.CreateCondBr(CondV, ThenBB, ElseBB);
 733
 734   // Emit then value.
 735   Builder.SetInsertPoint(ThenBB);
 736
 737   Value *ThenV = Then->Codegen();
 738   if (ThenV == 0) return 0;
 739
 740   Builder.CreateBr(MergeBB);
 741   // Codegen of 'Then' can change the current block, update ThenBB for the PHI.
 742   ThenBB = Builder.GetInsertBlock();
 743
 744   // Emit else block.
 745   TheFunction->getBasicBlockList().push_back(ElseBB);
 746   Builder.SetInsertPoint(ElseBB);
 747
 748   Value *ElseV = Else->Codegen();
 749   if (ElseV == 0) return 0;
 750
 751   Builder.CreateBr(MergeBB);
 752   // Codegen of 'Else' can change the current block, update ElseBB for the PHI.
 753   ElseBB = Builder.GetInsertBlock();
 754
 755   // Emit merge block.
 756   TheFunction->getBasicBlockList().push_back(MergeBB);
 757   Builder.SetInsertPoint(MergeBB);
 758   PHINode *PN = Builder.CreatePHI(Type::getDoubleTy(getGlobalContext()),
 759                                   "iftmp");
 760
 761   PN->addIncoming(ThenV, ThenBB);
 762   PN->addIncoming(ElseV, ElseBB);
 763   return PN;
 764 }
 765
 766 Value *ForExprAST::Codegen() {
 767   // Output this as:
 768   //   var = alloca double
 769   //   ...
 770   //   start = startexpr
 771   //   store start -> var
 772   //   goto loop
 773   // loop:
 774   //   ...
 775   //   bodyexpr
 776   //   ...
 777   // loopend:
 778   //   step = stepexpr
 779   //   endcond = endexpr
 780   //
 781   //   curvar = load var
 782   //   nextvar = curvar + step
 783   //   store nextvar -> var
 784   //   br endcond, loop, endloop
 785   // outloop:
 786
 787   Function *TheFunction = Builder.GetInsertBlock()->getParent();
 788
 789   // Create an alloca for the variable in the entry block.
 790   AllocaInst *Alloca = CreateEntryBlockAlloca(TheFunction, VarName);
 791
 792   // Emit the start code first, without 'variable' in scope.
 793   Value *StartVal = Start->Codegen();
 794   if (StartVal == 0) return 0;
 795
 796   // Store the value into the alloca.
 797   Builder.CreateStore(StartVal, Alloca);
 798
 799   // Make the new basic block for the loop header, inserting after current
 800   // block.
 801   BasicBlock *LoopBB = BasicBlock::Create(getGlobalContext(), "loop", TheFunction);
 802
 803   // Insert an explicit fall through from the current block to the LoopBB.
 804   Builder.CreateBr(LoopBB);
 805
 806   // Start insertion in LoopBB.
 807   Builder.SetInsertPoint(LoopBB);
 808
 809   // Within the loop, the variable is defined equal to the PHI node.  If it
 810   // shadows an existing variable, we have to restore it, so save it now.
 811   AllocaInst *OldVal = NamedValues[VarName];
 812   NamedValues[VarName] = Alloca;
 813
 814   // Emit the body of the loop.  This, like any other expr, can change the
 815   // current BB.  Note that we ignore the value computed by the body, but don't
 816   // allow an error.
 817   if (Body->Codegen() == 0)
 818     return 0;
 819
 820   // Emit the step value.
 821   Value *StepVal;
 822   if (Step) {
 823     StepVal = Step->Codegen();
 824     if (StepVal == 0) return 0;
 825   } else {
 826     // If not specified, use 1.0.
 827     StepVal = ConstantFP::get(getGlobalContext(), APFloat(1.0));
 828   }
 829
 830   // Compute the end condition.
 831   Value *EndCond = End->Codegen();
 832   if (EndCond == 0) return EndCond;
 833
 834   // Reload, increment, and restore the alloca.  This handles the case where
 835   // the body of the loop mutates the variable.
 836   Value *CurVar = Builder.CreateLoad(Alloca, VarName.c_str());
 837   Value *NextVar = Builder.CreateAdd(CurVar, StepVal, "nextvar");
 838   Builder.CreateStore(NextVar, Alloca);
 839
 840   // Convert condition to a bool by comparing equal to 0.0.
 841   EndCond = Builder.CreateFCmpONE(EndCond,
 842                               ConstantFP::get(getGlobalContext(), APFloat(0.0)),
 843                                   "loopcond");
 844
 845   // Create the "after loop" block and insert it.
 846   BasicBlock *AfterBB = BasicBlock::Create(getGlobalContext(), "afterloop", TheFunction);
 847
 848   // Insert the conditional branch into the end of LoopEndBB.
 849   Builder.CreateCondBr(EndCond, LoopBB, AfterBB);
 850
 851   // Any new code will be inserted in AfterBB.
 852   Builder.SetInsertPoint(AfterBB);
 853
 854   // Restore the unshadowed variable.
 855   if (OldVal)
 856     NamedValues[VarName] = OldVal;
 857   else
 858     NamedValues.erase(VarName);
 859
 860
 861   // for expr always returns 0.0.
 862   return Constant::getNullValue(Type::getDoubleTy(getGlobalContext()));
 863 }
 864
 865 Value *VarExprAST::Codegen() {
 866   std::vector<AllocaInst *> OldBindings;
 867
 868   Function *TheFunction = Builder.GetInsertBlock()->getParent();
 869
 870   // Register all variables and emit their initializer.
 871   for (unsigned i = 0, e = VarNames.size(); i != e; ++i) {
 872     const std::string &VarName = VarNames[i].first;
 873     ExprAST *Init = VarNames[i].second;
 874
 875     // Emit the initializer before adding the variable to scope, this prevents
 876     // the initializer from referencing the variable itself, and permits stuff
 877     // like this:
 878     //  var a = 1 in
 879     //    var a = a in ...   # refers to outer 'a'.
 880     Value *InitVal;
 881     if (Init) {
 882       InitVal = Init->Codegen();
 883       if (InitVal == 0) return 0;
 884     } else { // If not specified, use 0.0.
 885       InitVal = ConstantFP::get(getGlobalContext(), APFloat(0.0));
 886     }
 887
 888     AllocaInst *Alloca = CreateEntryBlockAlloca(TheFunction, VarName);
 889     Builder.CreateStore(InitVal, Alloca);
 890
 891     // Remember the old variable binding so that we can restore the binding when
 892     // we unrecurse.
 893     OldBindings.push_back(NamedValues[VarName]);
 894
 895     // Remember this binding.
 896     NamedValues[VarName] = Alloca;
 897   }
 898
 899   // Codegen the body, now that all vars are in scope.
 900   Value *BodyVal = Body->Codegen();
 901   if (BodyVal == 0) return 0;
 902
 903   // Pop all our variables from scope.
 904   for (unsigned i = 0, e = VarNames.size(); i != e; ++i)
 905     NamedValues[VarNames[i].first] = OldBindings[i];
 906
 907   // Return the body computation.
 908   return BodyVal;
 909 }
 910
 911
 912 Function *PrototypeAST::Codegen() {
 913   // Make the function type:  double(double,double) etc.
 914   std::vector<const Type*> Doubles(Args.size(),
 915                                    Type::getDoubleTy(getGlobalContext()));
 916   FunctionType *FT = FunctionType::get(Type::getDoubleTy(getGlobalContext()),
 917                                        Doubles, false);
 918
 919   Function *F = Function::Create(FT, Function::ExternalLinkage, Name, TheModule);
 920
 921   // If F conflicted, there was already something named 'Name'.  If it has a
 922   // body, don't allow redefinition or reextern.
 923   if (F->getName() != Name) {
 924     // Delete the one we just made and get the existing one.
 925     F->eraseFromParent();
 926     F = TheModule->getFunction(Name);
 927
 928     // If F already has a body, reject this.
 929     if (!F->empty()) {
 930       ErrorF("redefinition of function");
 931       return 0;
 932     }
 933
 934     // If F took a different number of args, reject.
 935     if (F->arg_size() != Args.size()) {
 936       ErrorF("redefinition of function with different # args");
 937       return 0;
 938     }
 939   }
 940
 941   // Set names for all arguments.
 942   unsigned Idx = 0;
 943   for (Function::arg_iterator AI = F->arg_begin(); Idx != Args.size();
 944        ++AI, ++Idx)
 945     AI->setName(Args[Idx]);
 946
 947   return F;
 948 }
 949
 950 /// CreateArgumentAllocas - Create an alloca for each argument and register the
 951 /// argument in the symbol table so that references to it will succeed.
 952 void PrototypeAST::CreateArgumentAllocas(Function *F) {
 953   Function::arg_iterator AI = F->arg_begin();
 954   for (unsigned Idx = 0, e = Args.size(); Idx != e; ++Idx, ++AI) {
 955     // Create an alloca for this variable.
 956     AllocaInst *Alloca = CreateEntryBlockAlloca(F, Args[Idx]);
 957
 958     // Store the initial value into the alloca.
 959     Builder.CreateStore(AI, Alloca);
 960
 961     // Add arguments to variable symbol table.
 962     NamedValues[Args[Idx]] = Alloca;
 963   }
 964 }
 965
 966
 967 Function *FunctionAST::Codegen() {
 968   NamedValues.clear();
 969
 970   Function *TheFunction = Proto->Codegen();
 971   if (TheFunction == 0)
 972     return 0;
 973
 974   // If this is an operator, install it.
 975   if (Proto->isBinaryOp())
 976     BinopPrecedence[Proto->getOperatorName()] = Proto->getBinaryPrecedence();
 977
 978   // Create a new basic block to start insertion into.
 979   BasicBlock *BB = BasicBlock::Create(getGlobalContext(), "entry", TheFunction);
 980   Builder.SetInsertPoint(BB);
 981
 982   // Add all arguments to the symbol table and create their allocas.
 983   Proto->CreateArgumentAllocas(TheFunction);
 984
 985   if (Value *RetVal = Body->Codegen()) {
 986     // Finish off the function.
 987     Builder.CreateRet(RetVal);
 988
 989     // Validate the generated code, checking for consistency.
 990     verifyFunction(*TheFunction);
 991
 992     // Optimize the function.
 993     TheFPM->run(*TheFunction);
 994
 995     return TheFunction;
 996   }
 997
 998   // Error reading body, remove function.
 999   TheFunction->eraseFromParent();
1000
1001   if (Proto->isBinaryOp())
1002     BinopPrecedence.erase(Proto->getOperatorName());
1003   return 0;
1004 }
1005
1006 //===----------------------------------------------------------------------===//
1007 // Top-Level parsing and JIT Driver
1008 //===----------------------------------------------------------------------===//
1009
1010 static ExecutionEngine *TheExecutionEngine;
1011
1012 static void HandleDefinition() {
1013   if (FunctionAST *F = ParseDefinition()) {
1014     if (Function *LF = F->Codegen()) {
1015       fprintf(stderr, "Read function definition:");
1016       LF->dump();
1017     }
1018   } else {
1019     // Skip token for error recovery.
1020     getNextToken();
1021   }
1022 }
1023
1024 static void HandleExtern() {
1025   if (PrototypeAST *P = ParseExtern()) {
1026     if (Function *F = P->Codegen()) {
1027       fprintf(stderr, "Read extern: ");
1028       F->dump();
1029     }
1030   } else {
1031     // Skip token for error recovery.
1032     getNextToken();
1033   }
1034 }
1035
1036 static void HandleTopLevelExpression() {
1037   // Evaluate a top level expression into an anonymous function.
1038   if (FunctionAST *F = ParseTopLevelExpr()) {
1039     if (Function *LF = F->Codegen()) {
1040       // JIT the function, returning a function pointer.
1041       void *FPtr = TheExecutionEngine->getPointerToFunction(LF);
1042
1043       // Cast it to the right type (takes no arguments, returns a double) so we
1044       // can call it as a native function.
1045       double (*FP)() = (double (*)())(intptr_t)FPtr;
1046       fprintf(stderr, "Evaluated to %f\n", FP());
1047     }
1048   } else {
1049     // Skip token for error recovery.
1050     getNextToken();
1051   }
1052 }
1053
1054 /// top ::= definition | external | expression | ';'
1055 static void MainLoop() {
1056   while (1) {
1057     fprintf(stderr, "ready> ");
1058     switch (CurTok) {
1059     case tok_eof:    return;
1060     case ';':        getNextToken(); break;  // ignore top level semicolons.
1061     case tok_def:    HandleDefinition(); break;
1062     case tok_extern: HandleExtern(); break;
1063     default:         HandleTopLevelExpression(); break;
1064     }
1065   }
1066 }
1067
1068
1069
1070 //===----------------------------------------------------------------------===//
1071 // "Library" functions that can be "extern'd" from user code.
1072 //===----------------------------------------------------------------------===//
1073
1074 /// putchard - putchar that takes a double and returns 0.
1075 extern "C"
1076 double putchard(double X) {
1077   putchar((char)X);
1078   return 0;
1079 }
1080
1081 /// printd - printf that takes a double prints it as "%f\n", returning 0.
1082 extern "C"
1083 double printd(double X) {
1084   printf("%f\n", X);
1085   return 0;
1086 }
1087
1088 //===----------------------------------------------------------------------===//
1089 // Main driver code.
1090 //===----------------------------------------------------------------------===//
1091
1092 int main() {
1093   InitializeNativeTarget();
1094   LLVMContext &Context = getGlobalContext();
1095
1096   // Install standard binary operators.
1097   // 1 is lowest precedence.
1098   BinopPrecedence['='] = 2;
1099   BinopPrecedence['<'] = 10;
1100   BinopPrecedence['+'] = 20;
1101   BinopPrecedence['-'] = 20;
1102   BinopPrecedence['*'] = 40;  // highest.
1103
1104   // Prime the first token.
1105   fprintf(stderr, "ready> ");
1106   getNextToken();
1107
1108   // Make the module, which holds all the code.
1109   TheModule = new Module("my cool jit", Context);
1110
1111   ExistingModuleProvider *OurModuleProvider =
1112       new ExistingModuleProvider(TheModule);
1113
1114   // Create the JIT.  This takes ownership of the module and module provider.
1115   TheExecutionEngine = EngineBuilder(OurModuleProvider).create();
1116
1117   FunctionPassManager OurFPM(OurModuleProvider);
1118
1119   // Set up the optimizer pipeline.  Start with registering info about how the
1120   // target lays out data structures.
1121   OurFPM.add(new TargetData(*TheExecutionEngine->getTargetData()));
1122   // Promote allocas to registers.
1123   OurFPM.add(createPromoteMemoryToRegisterPass());
1124   // Do simple "peephole" optimizations and bit-twiddling optzns.
1125   OurFPM.add(createInstructionCombiningPass());
1126   // Reassociate expressions.
1127   OurFPM.add(createReassociatePass());
1128   // Eliminate Common SubExpressions.
1129   OurFPM.add(createGVNPass());
1130   // Simplify the control flow graph (deleting unreachable blocks, etc).
1131   OurFPM.add(createCFGSimplificationPass());
1132
1133   OurFPM.doInitialization();
1134
1135   // Set the global so the code gen can use this.
1136   TheFPM = &OurFPM;
1137
1138   // Run the main "interpreter loop" now.
1139   MainLoop();
1140
1141   TheFPM = 0;
1142
1143   // Print out all of the generated code.
1144   TheModule->dump();
1145
1146   return 0;
1147 }